Add deprecation warning

Signed-off-by: Yuri Shkuro <yurishkuro@users.noreply.github.com>
Use jaeger storage integration helpers from upstream (#119 )
2024-12-16 23:48:22 -04:00 · 2022-10-12 19:42:03 +13:00 · 2022-10-10 21:39:01 +13:00 · 2022-10-10 21:37:28 +13:00 · 2022-09-14 21:18:29 +12:00 · 2022-08-23 19:16:16 +12:00
67 changed files with 6929 additions and 962 deletions
--- a/.github/workflows/build.yaml
+++ b/.github/workflows/build.yaml
@ -1,4 +1,4 @@
-name: Build and format
+name: Build, test, format and lint

 on:
  push:
@ -7,7 +7,18 @@ on:
 jobs:
  build-binaries:
    runs-on: ubuntu-latest
-    name: Build binaries
+    name: Build binary for ${{ matrix.platform.name }}
+    strategy:
+      matrix:
+        platform:
+        - name: linux on amd64
+          task: build-linux-amd64
+        - name: linux on arm64
+          task: build-linux-arm64
+        - name: osx on amd64
+          task: build-darwin-amd64
+        - name: osx on arm64
+          task: build-darwin-arm64
    steps:
      - uses: actions/checkout@v2.3.4
        with:
@ -15,14 +26,14 @@ jobs:

      - uses: actions/setup-go@v2
        with:
-          go-version: ^1.16
+          go-version: ^1.19

      - name: Build binaries
-        run: make build
+        run: make ${{ matrix.platform.task }}

-  format:
+  format-lint:
    runs-on: ubuntu-latest
-    name: Format
+    name: Format and lint
    steps:
      - uses: actions/checkout@v2.3.4
        with:
@ -30,10 +41,61 @@ jobs:

      - uses: actions/setup-go@v2
        with:
-          go-version: ^1.16
+          go-version: ^1.19

      - name: Install tools
        run: make install-tools

      - name: Format
        run: make fmt && git diff --quiet
+
+      - name: Lint
+        run: make lint
+
+  e2e-test:
+    runs-on: ubuntu-latest
+    name: E2E Test
+    steps:
+      - uses: actions/checkout@v2.3.4
+        with:
+          submodules: true
+
+      - uses: actions/setup-go@v2
+        with:
+          go-version: ^1.19
+
+      - name: Run e2e test
+        run: make e2e-tests
+
+  unit-tests:
+    runs-on: ubuntu-latest
+    name: Unit tests
+    steps:
+      - uses: actions/checkout@v2.3.4
+        with:
+          submodules: true
+
+      - uses: actions/setup-go@v2
+        with:
+          go-version: ^1.19
+
+      - name: Run unit test
+        run: make test
+
+  integration-tests:
+    runs-on: ubuntu-latest
+    name: Integration tests
+    steps:
+      - uses: actions/checkout@v2.3.4
+        with:
+          submodules: true
+
+      - uses: actions/setup-go@v2
+        with:
+          go-version: ^1.19
+
+      - name: Setup database
+        run: docker run --rm -d -p9000:9000 --name test-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server:22
+
+      - name: Run integration tests
+        run: make integration-test
--- a/.github/workflows/release.yaml
+++ b/.github/workflows/release.yaml
@ -15,10 +15,22 @@ jobs:

      - uses: actions/setup-go@v2
        with:
-          go-version: ^1.16
+          go-version: ^1.19

      - name: Create release distribution
-        run: make build tar
+        run: make build-all-platforms tar-all-platforms
+
+      - name: Log in to the Container registry
+        uses: docker/login-action@f054a8b539a109f9f41c372932f1ae047eff08c9
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Build docker image
+        run: |
+          export DOCKER_TAG=${GITHUB_REF##*/}
+          make docker docker-push

      - name: Create Github release
        run: |
--- a/.gitignore
+++ b/.gitignore
@ -11,6 +11,9 @@
 # Output of the go coverage tool, specifically when used with LiteIDE
 *.out

+# IDE
+.idea/
+
 # Dependency directories (remove the comment below to include it)
 # vendor/

--- a/.golangci.yml
+++ b/.golangci.yml
@ -0,0 +1,146 @@
+# options for analysis running
+run:
+  # default concurrency is a available CPU number
+  concurrency: 4
+
+  # timeout for analysis, e.g. 30s, 5m, default is 1m
+  timeout: 10m
+
+  # exit code when at least one issue was found, default is 1
+  issues-exit-code: 1
+
+  # include test files or not, default is true
+  tests: true
+
+  # which dirs to skip: issues from them won't be reported;
+  # can use regexp here: generated.*, regexp is applied on full path;
+  # default value is empty list, but default dirs are skipped independently
+  # from this option's value (see skip-dirs-use-default).
+  skip-dirs:
+
+  # default is true. Enables skipping of directories:
+  #   vendor$, third_party$, testdata$, examples$, Godeps$, builtin$
+  skip-dirs-use-default: false
+
+  # which files to skip: they will be analyzed, but issues from them
+  # won't be reported. Default value is empty list, but there is
+  # no need to include all autogenerated files, we confidently recognize
+  # autogenerated files. If it's not please let us know.
+  skip-files:
+
+  # by default isn't set. If set we pass it to "go list -mod={option}". From "go help modules":
+  # If invoked with -mod=readonly, the go command is disallowed from the implicit
+  # automatic updating of go.mod described above. Instead, it fails when any changes
+  # to go.mod are needed. This setting is most useful to check that go.mod does
+  # not need updates, such as in a continuous integration and testing system.
+  # If invoked with -mod=vendor, the go command assumes that the vendor
+  # directory holds the correct copies of dependencies and ignores
+  # the dependency descriptions in go.mod.
+  modules-download-mode: readonly
+
+# output configuration options
+output:
+  # colored-line-number|line-number|json|tab|checkstyle|code-climate, default is "colored-line-number"
+  format: colored-line-number
+
+  # print lines of code with issue, default is true
+  print-issued-lines: true
+
+  # print linter name in the end of issue text, default is true
+  print-linter-name: true
+
+# all available settings of specific linters
+linters-settings:
+  govet:
+    # report about shadowed variables
+    check-shadowing: true
+
+    # settings per analyzer
+    settings:
+      printf: # analyzer name, run `go tool vet help` to see all analyzers
+        funcs: # run `go tool vet help printf` to see available settings for `printf` analyzer
+          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Infof
+          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Warnf
+          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Errorf
+          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Fatalf
+
+    enable-all: true
+    # TODO: Enable this and fix the alignment issues.
+    disable:
+      - fieldalignment
+
+  revive:
+    # minimal confidence for issues, default is 0.8
+    min-confidence: 0.8
+
+  gofmt:
+    # simplify code: gofmt with `-s` option, true by default
+    simplify: true
+
+  goimports:
+    # put imports beginning with prefix after 3rd-party packages;
+    # it's a comma-separated list of prefixes
+    local-prefixes: go.opentelemetry.io/collector
+
+  misspell:
+    # Correct spellings using locale preferences for US or UK.
+    # Default is to use a neutral variety of English.
+    # Setting locale to US will correct the British spelling of 'colour' to 'color'.
+    locale: US
+    ignore-words:
+      - cancelled
+      - metre
+      - meter
+      - metres
+      - kilometre
+      - kilometres
+
+linters:
+  disable:
+    - errcheck
+  enable:
+    - exportloopref
+    - gocritic
+    - gofmt
+    - goimports
+    - gosec
+    - govet
+    - misspell
+    - revive
+    - staticcheck
+    - unconvert
+    - unparam
+
+issues:
+  # Excluding configuration per-path, per-linter, per-text and per-source
+  exclude-rules:
+    # Exclude some linters from running on tests files.
+    - path: otlp_test.go
+      linters:
+        # See https://github.com/golangci/golangci-lint/issues/537#issuecomment-545170007
+        - structcheck
+    - text: "G404:"
+      linters:
+        - gosec
+    - text: "G402:"
+      linters:
+        - gosec
+    - path: grpc_test.go
+      linters:
+        # See https://github.com/golangci/golangci-lint/issues/2286
+        - typecheck
+
+  # The list of ids of default excludes to include or disable. By default it's empty.
+  # See the list of default excludes here https://golangci-lint.run/usage/configuration.
+  include:
+    - EXC0001
+    - EXC0002
+    - EXC0003
+    - EXC0004
+    - EXC0005
+    - EXC0006
+    - EXC0007
+    # - EXC0008 - Duplicated errcheck checks
+    - EXC0009
+    - EXC0010
+    - EXC0011
--- a/8
+++ b/8
@ -0,0 +1,8 @@
+FROM docker.io/library/alpine:3.16
+
+ADD jaeger-clickhouse-linux-amd64 /go/bin/jaeger-clickhouse
+
+RUN mkdir /plugin
+
+# /plugin/ location is defined in jaeger-operator
+CMD ["cp", "/go/bin/jaeger-clickhouse", "/plugin/jaeger-clickhouse"]
--- a/82
+++ b/82
@ -1,28 +1,102 @@
 GOOS ?= $(shell go env GOOS)
 GOARCH ?= $(shell go env GOARCH)
+GOBUILD=CGO_ENABLED=0 installsuffix=cgo go build -trimpath

 TOOLS_MOD_DIR = ./internal/tools
+JAEGER_VERSION ?= 1.32.0

-JAEGER_ALL_IN_ONE ?= ${HOME}/projects/jaegertracing/jaeger/cmd/all-in-one/all-in-one-linux-amd64
+DOCKER_REPO ?= ghcr.io/jaegertracing/jaeger-clickhouse
+DOCKER_TAG ?= latest

 .PHONY: build
 build:
-	go build -o jaeger-clickhouse-$(GOOS)-$(GOARCH) ./cmd/jaeger-clickhouse/main.go
+	${GOBUILD} -o jaeger-clickhouse-$(GOOS)-$(GOARCH) ./cmd/jaeger-clickhouse/main.go
+
+.PHONY: build-linux-amd64
+build-linux-amd64:
+	GOOS=linux GOARCH=amd64 $(MAKE) build
+
+.PHONY: build-linux-arm64
+build-linux-arm64:
+	GOOS=linux GOARCH=arm64 $(MAKE) build
+
+.PHONY: build-darwin-amd64
+build-darwin-amd64:
+	GOOS=darwin GOARCH=amd64 $(MAKE) build
+
+.PHONY: build-darwin-arm64
+build-darwin-arm64:
+	GOOS=darwin GOARCH=arm64 $(MAKE) build
+
+.PHONY: build-all-platforms
+build-all-platforms: build-linux-amd64 build-linux-arm64 build-darwin-amd64 build-darwin-arm64
+
+.PHONY: e2e-tests
+e2e-tests:
+	GOOS=linux GOARCH=amd64 $(MAKE) build
+	E2E_TEST=true go test ./e2etests... -v

 .PHONY: run
 run:
-	SPAN_STORAGE_TYPE=grpc-plugin ${JAEGER_ALL_IN_ONE} --grpc-storage-plugin.binary=./jaeger-clickhouse-$(GOOS)-$(GOARCH) --grpc-storage-plugin.configuration-file=./config.yaml
+	docker run --rm --name jaeger -e JAEGER_DISABLED=false --link some-clickhouse-server -it -u ${shell id -u} -p16686:16686 -p14250:14250 -p14268:14268 -p6831:6831/udp -v "${PWD}:/data" -e SPAN_STORAGE_TYPE=grpc-plugin jaegertracing/all-in-one:${JAEGER_VERSION} --query.ui-config=/data/jaeger-ui.json --grpc-storage-plugin.binary=/data/jaeger-clickhouse-$(GOOS)-$(GOARCH) --grpc-storage-plugin.configuration-file=/data/config.yaml --grpc-storage-plugin.log-level=debug
+
+.PHONY: run-hotrod
+run-hotrod:
+	docker run --rm --link jaeger --env JAEGER_AGENT_HOST=jaeger --env JAEGER_AGENT_PORT=6831 -p8080:8080 jaegertracing/example-hotrod:${JAEGER_VERSION} all

 .PHONY: fmt
 fmt:
 	go fmt ./...
-	goimports -w  -local github.com/pavolloffay/jaeger-clickhouse ./
+	goimports -w -local github.com/jaegertracing/jaeger-clickhouse ./
+
+.PHONY: lint
+lint:
+	golangci-lint -v run --allow-parallel-runners ./...
+
+.PHONY: test
+test:
+	go test ./...
+
+.PHONY: integration-test
+integration-test: build
+	STORAGE=grpc-plugin \
+	PLUGIN_BINARY_PATH=$(PWD)/jaeger-clickhouse-linux-amd64 \
+	PLUGIN_CONFIG_PATH=$(PWD)/integration/config-local.yaml \
+	go test ./integration

 .PHONY: tar
 tar:
 	tar -czvf jaeger-clickhouse-$(GOOS)-$(GOARCH).tar.gz  jaeger-clickhouse-$(GOOS)-$(GOARCH) config.yaml

+.PHONY: tar-linux-amd64
+tar-linux-amd64:
+	GOOS=linux GOARCH=amd64 $(MAKE) tar
+
+.PHONY: tar-linux-arm64
+tar-linux-arm64:
+	GOOS=linux GOARCH=arm64 $(MAKE) tar
+
+.PHONY: tar-darwin-amd64
+tar-darwin-amd64:
+	GOOS=darwin GOARCH=amd64 $(MAKE) tar
+
+.PHONY: tar-darwin-arm64
+tar-darwin-arm64:
+	GOOS=darwin GOARCH=arm64 $(MAKE) tar
+
+.PHONY: tar-all-platforms
+tar-all-platforms: tar-linux-amd64 tar-linux-arm64 tar-darwin-amd64 tar-darwin-arm64
+
+.PHONY: docker
+docker: build
+	docker build -t ${DOCKER_REPO}:${DOCKER_TAG} -f Dockerfile .
+
+.PHONY: docker-push
+docker-push: build
+	docker push ${DOCKER_REPO}:${DOCKER_TAG}
+
 .PHONY: install-tools
 install-tools:
 	cd $(TOOLS_MOD_DIR) && go install golang.org/x/tools/cmd/goimports
+	cd $(TOOLS_MOD_DIR) && go install github.com/golangci/golangci-lint/cmd/golangci-lint

--- a/README.md
+++ b/README.md
@ -1,22 +1,65 @@
-# Jaeger ClickHouse
+# Jaeger ClickHouse (experimental)

-Jaeger ClickHouse gRPC [storage plugin](https://github.com/jaegertracing/jaeger/tree/master/plugin/storage/grpc).
+⚠️ This module only implements grpc-plugin API that has been deprecated in Jaeger (https://github.com/jaegertracing/jaeger/issues/4647).

-This is WIP and it is based on https://github.com/bobrik/jaeger/tree/ivan/clickhouse/plugin/storage/clickhouse. 
-See as well [jaegertracing/jaeger/issues/1438](https://github.com/jaegertracing/jaeger/issues/1438) for ClickHouse plugin.
+This is a [Jaeger gRPC storage plugin](https://github.com/jaegertracing/jaeger/tree/master/plugin/storage/grpc) implementation for storing traces in ClickHouse.

-## Build
+## Project status
+
+This is a community-driven project, and we would love to hear your issues and feature requests.
+Pull requests are also greatly appreciated.
+
+## Why use ClickHouse for Jaeger?
+
+[ClickHouse](https://clickhouse.com) is an analytical column-oriented database management system.
+It is designed to analyze streams of events which are kind of resemblant to spans.
+It's open-source, optimized for performance, and actively developed.
+
+## How it works
+
+Jaeger spans are stored in 2 tables. The first contains the whole span encoded either in JSON or Protobuf.
+The second stores key information about spans for searching. This table is indexed by span duration and tags.
+Also, info about operations is stored in the materialized view. There are not indexes for archived spans.
+Storing data in replicated local tables with distributed global tables is natively supported. Spans are bufferized.
+Span buffers are flushed to DB either by timer or after reaching max batch size. Timer interval and batch size can be
+set in [config file](./config.yaml).
+
+Database schema generated by JetBrains DataGrip
+![Picture of tables](./pictures/tables.png)
+
+# How to start using Jaeger over ClickHouse
+
+## Documentation
+
+Refer to the [config.yaml](./config.yaml) for all supported configuration options.
+
+* [Kubernetes deployment](./guide-kubernetes.md)
+* [Sharding and replication](./guide-sharding-and-replication.md)
+* [Multi-tenancy](./guide-multitenancy.md)
+
+## Build & Run
+
+### Docker database example
+
+```bash
+docker run --rm -it -p9000:9000 --name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server:22
+GOOS=linux make build run
+make run-hotrod
+```
+
+Open [localhost:16686](http://localhost:16686) and [localhost:8080](http://localhost:8080).
+
+### Custom database
+
+You need to specify connection options in `config.yaml`, then you can run

 ```bash
 make build
+SPAN_STORAGE_TYPE=grpc-plugin {Jaeger binary adress} --query.ui-config=jaeger-ui.json --grpc-storage-plugin.binary=./{name of built binary} --grpc-storage-plugin.configuration-file=config.yaml --grpc-storage-plugin.log-level=debug
 ```

-## Run 
+## Credits

-```bash
-docker run --rm -it -p9000:9000 --name some-clickhouse-server --ulimit nofile=262144:262144 yandex/clickhouse-server:21
-# download Jaeger all-in-one
-JAEGER_ALL_IN_ONE=<path to all-in-one> make run
-```
+This project is originally based on [this clickhouse plugin implementation](https://github.com/bobrik/jaeger/tree/ivan/clickhouse/plugin/storage/clickhouse).

-After each Jaeger run the ClickHouse server has to be restarted or changed `sql_scripts_dir` to a directory with no SLQ scripts.
+See also [jaegertracing/jaeger/issues/1438](https://github.com/jaegertracing/jaeger/issues/1438) for historical discussion regarding the implementation of a ClickHouse plugin.
--- a/blog/post1.md
+++ b/blog/post1.md
@ -0,0 +1,96 @@
+# Jaeger ClickHouse
+This is an implementation of Jaeger's [storage plugin](https://github.com/jaegertracing/jaeger/tree/master/plugin/storage/grpc) for ClickHouse.
+See as well [jaegertracing/jaeger/issues/1438](https://github.com/jaegertracing/jaeger/issues/1438) for historical discussion regarding Clickhouse plugin.
+
+## Project status
+
+Jaeger ClickHouse is a community-driven project, we would love to hear your feature requests.
+Pull requests also will be greatly appreciated.
+
+## Why use ClickHouse for Jaeger?
+
+[ClickHouse](https://github.com/clickhouse/clickhouse) is an analytical column-oriented database management system. It is designed to analyze streams of clicks which are kind of resemblant to spans. It's open-source, optimized for performance, and actively developed.
+
+## How does it work?
+
+Jaeger spans are stored in 2 tables. First one contains whole span encoded either in JSON or Protobuf.
+Second stores key information about spans for searching. This table is indexed by span duration and tags.
+Also, info about operations is stored in the materialized view. There are no indexes for archived spans.
+Storing data in replicated local tables with distributed global tables is natively supported. Spans are buffered.
+Span buffers are flushed to DB either by timer or after reaching max batch size.
+Timer interval and batch size can be set in [config file](../config.yaml).
+
+![Picture of tables](post1_pics/tables.png)
+
+## Benchmarks
+
+10^8 traces were flushed using [jaeger-tracegen](https://www.jaegertracing.io/docs/1.25/tools/) to Clickhouse and ElasticSearch servers.
+Clickhouse server consisted of 3 shards, 2 hosts in each, and 3 Zookeeper hosts. Elasticsearch server consisted of 6 hosts,
+with 5 shards for primary index and 1 replica. All hosts were equal(8 vCPU, 32 GiB RAM, 20 GiB SSD).
+
+### General stats
+
+Cpu usage, [% of 1 host CPU]
+![CPU usage](post1_pics/cpu-usage.png)
+
+Memory usage, [bytes]
+![Memory usage](post1_pics/memory-usage.png)
+
+IO write, [operations]
+![IO write](post1_pics/io-write.png)
+
+Disk usage, [bytes]
+![Disk usage](post1_pics/disk-usage.png)
+
+### Recorded
+
+#### ClickHouse
+
+```sql
+SELECT count()
+FROM jaeger_index
+WHERE service = 'tracegen'
+
+┌──count()─┐
+│ 57026426 │
+└──────────┘
+```
+
+#### Elasticsearch
+
+![Elasticsearch span count](post1_pics/elastic-count.png?raw=true)
+
+# How to start using Jaeger over ClickHouse
+
+## Documentation
+
+Refer to the [config.yaml](../config.yaml) for all supported configuration options.
+
+* [Kubernetes deployment](../guide-kubernetes.md)
+* [Sharding and replication](../guide-sharding-and-replication.md)
+* [Multi-tenancy](../guide-multitenancy.md)
+
+## Build & Run
+
+### Docker database example
+
+```bash
+docker run --rm -it -p9000:9000 --name some-clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-server:22
+GOOS=linux make build run
+make run-hotrod
+```
+
+Open [localhost:16686](http://localhost:16686) and [localhost:8080](http://localhost:8080).
+
+### Custom database
+
+You need to specify connection options in config.yaml file, then you can run
+
+```bash
+make build
+SPAN_STORAGE_TYPE=grpc-plugin {Jaeger binary adress} --query.ui-config=jaeger-ui.json --grpc-storage-plugin.binary=./{name of built binary} --grpc-storage-plugin.configuration-file=config.yaml --grpc-storage-plugin.log-level=debug
+```
+
+## Credits
+
+This project is based on https://github.com/bobrik/jaeger/tree/ivan/clickhouse/plugin/storage/clickhouse.
--- a/blog/post1_pics/cpu-usage.png
+++ b/blog/post1_pics/cpu-usage.png
--- a/blog/post1_pics/disk-usage.png
+++ b/blog/post1_pics/disk-usage.png
--- a/blog/post1_pics/elastic-count.png
+++ b/blog/post1_pics/elastic-count.png
--- a/blog/post1_pics/io-write.png
+++ b/blog/post1_pics/io-write.png
--- a/blog/post1_pics/memory-usage.png
+++ b/blog/post1_pics/memory-usage.png
--- a/blog/post1_pics/tables.png
+++ b/blog/post1_pics/tables.png
--- a/cmd/jaeger-clickhouse/main.go
+++ b/cmd/jaeger-clickhouse/main.go
@ -2,16 +2,20 @@ package main

 import (
 	"flag"
-	"io/ioutil"
+	"net/http"
 	"os"
+	"path/filepath"

-	"github.com/hashicorp/go-hclog"
+	// Package contains time zone info for connecting to ClickHouse servers with non-UTC time zone
+	_ "time/tzdata"
+
+	hclog "github.com/hashicorp/go-hclog"
 	"github.com/jaegertracing/jaeger/plugin/storage/grpc"
 	"github.com/jaegertracing/jaeger/plugin/storage/grpc/shared"
-	"gopkg.in/yaml.v3"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
+	yaml "gopkg.in/yaml.v3"

-	jaegerclickhouse "github.com/pavolloffay/jaeger-clickhouse"
-	"github.com/pavolloffay/jaeger-clickhouse/storage"
+	"github.com/jaegertracing/jaeger-clickhouse/storage"
 )

 func main() {
@ -20,31 +24,44 @@ func main() {
 	flag.Parse()

 	logger := hclog.New(&hclog.LoggerOptions{
-		Name:  "jaeger-clickhouse",
-		Level: hclog.Warn, // Jaeger only captures >= Warn, so don't bother logging below Warn
+		Name: "jaeger-clickhouse",
+		// If this is set to e.g. Warn, the debug logs are never sent to Jaeger even despite
+		// --grpc-storage-plugin.log-level=debug
+		Level:      hclog.Trace,
+		JSONFormat: true,
 	})

-	cfgFile, err := ioutil.ReadFile(configPath)
+	cfgFile, err := os.ReadFile(filepath.Clean(configPath))
 	if err != nil {
-		logger.Error("Could not read config file: %q: %q", configPath, err)
+		logger.Error("Could not read config file", "config", configPath, "error", err)
 		os.Exit(1)
 	}
 	var cfg storage.Configuration
 	err = yaml.Unmarshal(cfgFile, &cfg)
 	if err != nil {
-		logger.Error("Could not parse config file: %q", err)
+		logger.Error("Could not parse config file", "error", err)
 	}

-	var store shared.PluginServices
-	s, err := storage.NewStore(logger, cfg, jaegerclickhouse.EmbeddedFiles)
+	go func() {
+		http.Handle("/metrics", promhttp.Handler())
+		err = http.ListenAndServe(cfg.MetricsEndpoint, nil)
+		if err != nil {
+			logger.Error("Failed to listen for metrics endpoint", "error", err)
+		}
+	}()
+
+	var pluginServices shared.PluginServices
+	store, err := storage.NewStore(logger, cfg)
 	if err != nil {
-		logger.Error("Failed to crate storage", err)
+		logger.Error("Failed to create a storage", err)
 		os.Exit(1)
 	}
-	store.Store = s
+	pluginServices.Store = store
+	pluginServices.ArchiveStore = store
+	pluginServices.StreamingSpanWriter = store

-	grpc.Serve(&store)
-	if err = s.Close(); err != nil {
+	grpc.Serve(&pluginServices)
+	if err = store.Close(); err != nil {
 		logger.Error("Failed to close store", "error", err)
 		os.Exit(1)
 	}
--- a/config.yaml
+++ b/config.yaml
@ -1,3 +1,49 @@
-address: tcp://localhost:9000
-# When empty the embedded scripts from sqlscripts directory are used
+address: some-clickhouse-server:9000
+# Directory with .sql files to run at plugin startup, mainly for integration tests.
+# Depending on the value of "init_tables", this can be run as a
+# replacement or supplement to creating default tables for span storage.
+# If init_tables is also enabled, the scripts in this directory will be run first.
 init_sql_scripts_dir:
+# Whether to automatically attempt to create tables in ClickHouse.
+# By default, this is enabled if init_sql_scripts_dir is empty,
+# or disabled if init_sql_scripts_dir is provided.
+init_tables:
+# Maximal amount of spans that can be pending writes at a time.
+# New spans exceeding this limit will be discarded,
+# keeping memory in check if there are issues writing to ClickHouse.
+# Check the "jaeger_clickhouse_discarded_spans" metric to keep track of discards.
+# If 0, no limit is set. Default 10_000_000.
+max_span_count:
+# Batch write size. Default 10_000.
+batch_write_size:
+# Batch flush interval. Default 5s.
+batch_flush_interval:
+# Encoding of stored data. Either json or protobuf. Default json.
+encoding:
+# Path to CA TLS certificate.
+ca_file:
+# Username for connection to ClickHouse. Default is "default".
+username:
+# Password for connection to ClickHouse.
+password:
+# ClickHouse database name. The database must be created manually before Jaeger starts. Default is "default".
+database:
+# If non-empty, enables a tenant column in tables, and uses the provided tenant name for this instance.
+# Default is empty. See guide-multitenancy.md for more information.
+tenant:
+# Endpoint for serving prometheus metrics. Default localhost:9090.
+metrics_endpoint: localhost:9090
+# Whether to use sql scripts supporting replication and sharding.
+# Replication can be used only on database with Atomic engine.
+# Default false.
+replication:
+# Table with spans. Default "jaeger_spans_local" or "jaeger_spans" when replication is enabled.
+spans_table:
+# Span index table. Default "jaeger_index_local" or "jaeger_index" when replication is enabled.
+spans_index_table:
+# Operations table. Default "jaeger_operations_local" or "jaeger_operations" when replication is enabled.
+operations_table:
+# TTL for data in tables in days. If 0, no TTL is set. Default 0.
+ttl:
+# The maximum number of spans to fetch per trace. If 0, no limit is set. Default 0.
+max_num_spans:
--- a/e2etests/clickhouse-replicated.xml
+++ b/e2etests/clickhouse-replicated.xml
@ -0,0 +1,47 @@
+<!-- Minimal configuration to enable cluster mode in a single clickhouse process -->
+<yandex>
+    <macros>
+        <installation>cluster</installation>
+        <all-sharded-shard>0</all-sharded-shard>
+        <cluster>cluster</cluster>
+        <shard>0</shard>
+        <replica>cluster-0-0</replica>
+    </macros>
+
+    <remote_servers>
+        <cluster>
+            <shard>
+                <internal_replication>true</internal_replication>
+                <replica>
+                    <host>localhost</host>
+                    <port>9000</port>
+                </replica>
+            </shard>
+        </cluster>
+    </remote_servers>
+
+    <keeper_server>
+        <tcp_port>2181</tcp_port>
+        <server_id>0</server_id>
+        <log_storage_path>/var/log/clickhouse-server/coordination/log</log_storage_path>
+        <snapshot_storage_path>/var/lib/clickhouse/coordination/snapshots</snapshot_storage_path>
+        <raft_configuration>
+            <server>
+                <id>0</id>
+                <hostname>localhost</hostname>
+                <port>9444</port>
+            </server>
+        </raft_configuration>
+    </keeper_server>
+
+    <zookeeper>
+        <!-- Clickhouse Keeper -->
+        <node>
+            <host>localhost</host>
+            <port>2181</port>
+        </node>
+    </zookeeper>
+    <distributed_ddl>
+        <path>/clickhouse/cluster/task_queue/ddl</path>
+    </distributed_ddl>
+</yandex>
--- a/e2etests/config-local-multi1.yaml
+++ b/e2etests/config-local-multi1.yaml
@ -0,0 +1,4 @@
+address: chi:9000
+tenant: multi1
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/config-local-multi2.yaml
+++ b/e2etests/config-local-multi2.yaml
@ -0,0 +1,4 @@
+address: chi:9000
+tenant: multi2
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/config-local-single.yaml
+++ b/e2etests/config-local-single.yaml
@ -0,0 +1,3 @@
+address: chi:9000
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/config-replication-multi1.yaml
+++ b/e2etests/config-replication-multi1.yaml
@ -0,0 +1,5 @@
+address: chi:9000
+replication: true
+tenant: multi1
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/config-replication-multi2.yaml
+++ b/e2etests/config-replication-multi2.yaml
@ -0,0 +1,5 @@
+address: chi:9000
+replication: true
+tenant: multi2
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/config-replication-single.yaml
+++ b/e2etests/config-replication-single.yaml
@ -0,0 +1,4 @@
+address: chi:9000
+replication: true
+# For test purposes flush on every write
+batch_write_size: 1
--- a/e2etests/e2e_test.go
+++ b/e2etests/e2e_test.go
@ -0,0 +1,203 @@
+package e2etests
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"io/ioutil"
+	"net/http"
+	"os"
+	"testing"
+	"time"
+
+	clickhouse "github.com/ClickHouse/clickhouse-go/v2"
+	"github.com/ecodia/golang-awaitility/awaitility"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	testcontainers "github.com/testcontainers/testcontainers-go"
+	"github.com/testcontainers/testcontainers-go/wait"
+)
+
+const (
+	clickHouseImage = "clickhouse/clickhouse-server:22"
+	jaegerImage     = "jaegertracing/all-in-one:1.32.0"
+
+	networkName     = "chi-jaeger-test"
+	clickhousePort  = "9000/tcp"
+	jaegerQueryPort = "16686/tcp"
+	jaegerAdminPort = "14269/tcp"
+)
+
+type testCase struct {
+	configs []string
+	chiconf *string
+}
+
+func TestE2E(t *testing.T) {
+	if os.Getenv("E2E_TEST") == "" {
+		t.Skip("Set E2E_TEST=true to run the test")
+	}
+
+	// Minimal additional configuration (config.d) to enable cluster mode
+	chireplconf := "clickhouse-replicated.xml"
+
+	tests := map[string]testCase{
+		"local-single": {
+			configs: []string{"config-local-single.yaml"},
+			chiconf: nil,
+		},
+		"local-multi": {
+			configs: []string{"config-local-multi1.yaml", "config-local-multi2.yaml"},
+			chiconf: nil,
+		},
+		"replication-single": {
+			configs: []string{"config-replication-single.yaml"},
+			chiconf: &chireplconf,
+		},
+		"replication-multi": {
+			configs: []string{"config-replication-multi1.yaml", "config-replication-multi2.yaml"},
+			chiconf: &chireplconf,
+		},
+	}
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			testE2E(t, test)
+		})
+	}
+}
+
+func testE2E(t *testing.T, test testCase) {
+	ctx := context.Background()
+	workingDir, err := os.Getwd()
+	require.NoError(t, err)
+
+	network, err := testcontainers.GenericNetwork(ctx, testcontainers.GenericNetworkRequest{
+		NetworkRequest: testcontainers.NetworkRequest{Name: networkName},
+	})
+	require.NoError(t, err)
+	defer network.Remove(ctx)
+
+	var bindMounts map[string]string
+	if test.chiconf != nil {
+		bindMounts = map[string]string{
+			fmt.Sprintf("%s/%s", workingDir, *test.chiconf): "/etc/clickhouse-server/config.d/testconf.xml",
+		}
+	} else {
+		bindMounts = map[string]string{}
+	}
+	chReq := testcontainers.ContainerRequest{
+		Image:        clickHouseImage,
+		ExposedPorts: []string{clickhousePort},
+		WaitingFor:   &clickhouseWaitStrategy{test: t, pollInterval: time.Millisecond * 200, startupTimeout: time.Minute},
+		Networks:     []string{networkName},
+		Hostname:     "chi",
+		BindMounts:   bindMounts,
+	}
+	chContainer, err := testcontainers.GenericContainer(ctx, testcontainers.GenericContainerRequest{
+		ContainerRequest: chReq,
+		Started:          true,
+	})
+	require.NoError(t, err)
+	defer chContainer.Terminate(ctx)
+
+	jaegerContainers := make([]testcontainers.Container, 0)
+	for _, pluginConfig := range test.configs {
+		jaegerReq := testcontainers.ContainerRequest{
+			Image:        jaegerImage,
+			ExposedPorts: []string{jaegerQueryPort, jaegerAdminPort},
+			WaitingFor:   wait.ForHTTP("/").WithPort(jaegerAdminPort).WithStartupTimeout(time.Second * 10),
+			Env: map[string]string{
+				"SPAN_STORAGE_TYPE": "grpc-plugin",
+			},
+			Cmd: []string{
+				"--grpc-storage-plugin.binary=/project-dir/jaeger-clickhouse-linux-amd64",
+				fmt.Sprintf("--grpc-storage-plugin.configuration-file=/project-dir/e2etests/%s", pluginConfig),
+				"--grpc-storage-plugin.log-level=debug",
+			},
+			BindMounts: map[string]string{
+				workingDir + "/..": "/project-dir",
+			},
+			Networks: []string{networkName},
+		}
+		// Call Start() manually here so that if it fails then we can still access the logs.
+		jaegerContainer, err := testcontainers.GenericContainer(ctx, testcontainers.GenericContainerRequest{
+			ContainerRequest: jaegerReq,
+		})
+		require.NoError(t, err)
+		defer func() {
+			logs, errLogs := jaegerContainer.Logs(ctx)
+			require.NoError(t, errLogs)
+			all, errLogs := ioutil.ReadAll(logs)
+			require.NoError(t, errLogs)
+			fmt.Printf("Jaeger logs:\n---->\n%s<----\n\n", string(all))
+			jaegerContainer.Terminate(ctx)
+		}()
+		err = jaegerContainer.Start(ctx)
+		require.NoError(t, err)
+
+		jaegerContainers = append(jaegerContainers, jaegerContainer)
+	}
+
+	for _, jaegerContainer := range jaegerContainers {
+		jaegerQueryPort, err := jaegerContainer.MappedPort(ctx, jaegerQueryPort)
+		require.NoError(t, err)
+
+		err = awaitility.Await(100*time.Millisecond, time.Second*3, func() bool {
+			// Jaeger traces itself so this request generates some spans
+			response, errHTTP := http.Get(fmt.Sprintf("http://localhost:%d/api/services", jaegerQueryPort.Int()))
+			require.NoError(t, errHTTP)
+			body, errHTTP := ioutil.ReadAll(response.Body)
+			require.NoError(t, errHTTP)
+			var r result
+			errHTTP = json.Unmarshal(body, &r)
+			require.NoError(t, errHTTP)
+			return len(r.Data) == 1 && r.Data[0] == "jaeger-query"
+		})
+		assert.NoError(t, err)
+	}
+}
+
+type result struct {
+	Data []string `json:"data"`
+}
+
+type clickhouseWaitStrategy struct {
+	test           *testing.T
+	pollInterval   time.Duration
+	startupTimeout time.Duration
+}
+
+var _ wait.Strategy = (*clickhouseWaitStrategy)(nil)
+
+func (c *clickhouseWaitStrategy) WaitUntilReady(ctx context.Context, target wait.StrategyTarget) error {
+	ctx, cancelContext := context.WithTimeout(ctx, c.startupTimeout)
+	defer cancelContext()
+
+	port, err := target.MappedPort(ctx, clickhousePort)
+	require.NoError(c.test, err)
+
+	db := clickhouse.OpenDB(&clickhouse.Options{
+		Addr: []string{
+			fmt.Sprintf("localhost:%d", port.Int()),
+		},
+		Auth: clickhouse.Auth{
+			Database: "default",
+		},
+		Compression: &clickhouse.Compression{
+			Method: clickhouse.CompressionLZ4,
+		},
+	})
+	require.NoError(c.test, err)
+
+	for {
+		select {
+		case <-ctx.Done():
+			return ctx.Err()
+		case <-time.After(c.pollInterval):
+			if err := db.Ping(); err != nil {
+				continue
+			}
+			return nil
+		}
+	}
+}
--- a/embed.go
+++ b/embed.go
@ -1,6 +1,6 @@
-package jaeger_clickhouse
+package jaegerclickhouse

 import "embed"

 //go:embed sqlscripts/*
-var EmbeddedFiles embed.FS
+var SQLScripts embed.FS
--- a/go.mod
+++ b/go.mod
@ -1,13 +1,102 @@
-module github.com/pavolloffay/jaeger-clickhouse
+module github.com/jaegertracing/jaeger-clickhouse

-go 1.16
+go 1.19

 require (
-	github.com/ClickHouse/clickhouse-go v1.4.5
+	github.com/ClickHouse/clickhouse-go/v2 v2.3.0
+	github.com/DATA-DOG/go-sqlmock v1.5.0
+	github.com/ecodia/golang-awaitility v0.0.0-20180710094957-fb55e59708c7
 	github.com/gogo/protobuf v1.3.2
-	github.com/hashicorp/go-hclog v0.16.1
-	github.com/jaegertracing/jaeger v1.24.0
+	github.com/hashicorp/go-hclog v1.3.1
+	github.com/jaegertracing/jaeger v1.38.2-0.20221007043206-b4c88ddf6cdd
 	github.com/opentracing/opentracing-go v1.2.0
-	go.uber.org/zap v1.18.1
-	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b
+	github.com/prometheus/client_golang v1.13.0
+	github.com/stretchr/testify v1.8.0
+	github.com/testcontainers/testcontainers-go v0.11.1
+	go.uber.org/zap v1.23.0
+	gopkg.in/yaml.v3 v3.0.1
+)
+
+require (
+	github.com/Azure/go-ansiterm v0.0.0-20170929234023-d6e3b3328b78 // indirect
+	github.com/ClickHouse/ch-go v0.47.3 // indirect
+	github.com/Microsoft/go-winio v0.4.17-0.20210211115548-6eac466e5fa3 // indirect
+	github.com/Microsoft/hcsshim v0.8.16 // indirect
+	github.com/andybalholm/brotli v1.0.4 // indirect
+	github.com/benbjohnson/clock v1.3.0 // indirect
+	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/cenkalti/backoff v2.2.1+incompatible // indirect
+	github.com/cespare/xxhash/v2 v2.1.2 // indirect
+	github.com/containerd/cgroups v0.0.0-20210114181951-8a68de567b68 // indirect
+	github.com/containerd/containerd v1.5.0-beta.4 // indirect
+	github.com/davecgh/go-spew v1.1.1 // indirect
+	github.com/docker/distribution v2.7.1+incompatible // indirect
+	github.com/docker/docker v20.10.7+incompatible // indirect
+	github.com/docker/go-connections v0.4.0 // indirect
+	github.com/docker/go-units v0.4.0 // indirect
+	github.com/fatih/color v1.13.0 // indirect
+	github.com/fsnotify/fsnotify v1.5.4 // indirect
+	github.com/go-faster/city v1.0.1 // indirect
+	github.com/go-faster/errors v0.6.1 // indirect
+	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
+	github.com/golang/protobuf v1.5.2 // indirect
+	github.com/google/uuid v1.3.0 // indirect
+	github.com/grpc-ecosystem/grpc-opentracing v0.0.0-20180507213350-8e809c8a8645 // indirect
+	github.com/hashicorp/go-plugin v1.4.5 // indirect
+	github.com/hashicorp/hcl v1.0.0 // indirect
+	github.com/hashicorp/yamux v0.0.0-20211028200310-0bc27b27de87 // indirect
+	github.com/inconshreveable/mousetrap v1.0.0 // indirect
+	github.com/klauspost/compress v1.15.10 // indirect
+	github.com/kr/pretty v0.3.0 // indirect
+	github.com/kr/text v0.2.0 // indirect
+	github.com/magiconair/properties v1.8.6 // indirect
+	github.com/mattn/go-colorable v0.1.12 // indirect
+	github.com/mattn/go-isatty v0.0.14 // indirect
+	github.com/matttproud/golang_protobuf_extensions v1.0.2-0.20181231171920-c182affec369 // indirect
+	github.com/mitchellh/go-testing-interface v1.14.1 // indirect
+	github.com/mitchellh/mapstructure v1.5.0 // indirect
+	github.com/moby/sys/mount v0.2.0 // indirect
+	github.com/moby/sys/mountinfo v0.4.1 // indirect
+	github.com/moby/term v0.0.0-20201216013528-df9cb8a40635 // indirect
+	github.com/morikuni/aec v0.0.0-20170113033406-39771216ff4c // indirect
+	github.com/oklog/run v1.1.0 // indirect
+	github.com/opencontainers/go-digest v1.0.0 // indirect
+	github.com/opencontainers/image-spec v1.0.1 // indirect
+	github.com/opencontainers/runc v1.0.0-rc93 // indirect
+	github.com/paulmach/orb v0.7.1 // indirect
+	github.com/pelletier/go-toml v1.9.5 // indirect
+	github.com/pelletier/go-toml/v2 v2.0.5 // indirect
+	github.com/pierrec/lz4/v4 v4.1.15 // indirect
+	github.com/pkg/errors v0.9.1 // indirect
+	github.com/pmezard/go-difflib v1.0.0 // indirect
+	github.com/prometheus/client_model v0.2.0 // indirect
+	github.com/prometheus/common v0.37.0 // indirect
+	github.com/prometheus/procfs v0.8.0 // indirect
+	github.com/rogpeppe/go-internal v1.8.1 // indirect
+	github.com/segmentio/asm v1.2.0 // indirect
+	github.com/shopspring/decimal v1.3.1 // indirect
+	github.com/sirupsen/logrus v1.8.1 // indirect
+	github.com/spf13/afero v1.8.2 // indirect
+	github.com/spf13/cast v1.5.0 // indirect
+	github.com/spf13/cobra v1.5.0 // indirect
+	github.com/spf13/jwalterweatherman v1.1.0 // indirect
+	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/spf13/viper v1.13.0 // indirect
+	github.com/subosito/gotenv v1.4.1 // indirect
+	github.com/uber/jaeger-client-go v2.30.0+incompatible // indirect
+	github.com/uber/jaeger-lib v2.4.1+incompatible // indirect
+	go.opencensus.io v0.23.0 // indirect
+	go.opentelemetry.io/otel v1.10.0 // indirect
+	go.opentelemetry.io/otel/trace v1.10.0 // indirect
+	go.uber.org/atomic v1.10.0 // indirect
+	go.uber.org/multierr v1.8.0 // indirect
+	golang.org/x/net v0.0.0-20221002022538-bcab6841153b // indirect
+	golang.org/x/sys v0.0.0-20220928140112-f11e5e49a4ec // indirect
+	golang.org/x/text v0.3.7 // indirect
+	google.golang.org/genproto v0.0.0-20220822174746-9e6da59bd2fc // indirect
+	google.golang.org/grpc v1.50.0 // indirect
+	google.golang.org/protobuf v1.28.1 // indirect
+	gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c // indirect
+	gopkg.in/ini.v1 v1.67.0 // indirect
+	gopkg.in/yaml.v2 v2.4.0 // indirect
 )
--- a/go.sum
+++ b/go.sum
--- a/guide-kubernetes.md
+++ b/guide-kubernetes.md
@ -0,0 +1,87 @@
+# Kubernetes Deployment
+
+This is a guide to deploy Jaeger with Clickhouse storage on Kubernetes.
+
+## Prerequisites
+
+1. Deploy [Jaeger operator](https://github.com/jaegertracing/jaeger-operator). Note that `grpc-plugin` storage type is supported since version 1.25.0.
+2. Deploy [Clickhouse operator](https://github.com/Altinity/clickhouse-operator)
+3. Deploy [Zookeeper](https://github.com/Altinity/clickhouse-operator/blob/master/docs/replication_setup.md) (if replication is used)
+
+## Deploy
+
+Deploy Clickhouse:
+
+```yaml
+cat <<EOF | kubectl apply -f -
+apiVersion: clickhouse.altinity.com/v1
+kind: ClickHouseInstallation
+metadata:
+  name: jaeger
+  labels:
+    jaeger-clickhouse: demo
+spec:
+  configuration:
+    clusters:
+      - name: cluster1
+        layout:
+          shardsCount: 1
+EOF
+```
+
+Create config map for Jaeger Clickhouse plugin:
+
+```yaml
+cat <<EOF | kubectl apply -f -
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: jaeger-clickhouse
+  labels:
+    jaeger-clickhouse: demo
+data:
+  config.yaml: |
+    address: clickhouse-jaeger:9000
+    username: clickhouse_operator
+    password: clickhouse_operator_password
+    spans_table:
+    spans_index_table:
+    operations_table:
+EOF
+```
+
+Deploy Jaeger:
+
+```yaml
+cat <<EOF | kubectl apply -f -
+apiVersion: jaegertracing.io/v1
+kind: Jaeger
+metadata:
+  name: jaeger-clickhouse
+  labels:
+    jaeger-clickhouse: demo
+spec:
+  storage:
+    type: grpc-plugin
+    grpcPlugin:
+      image: ghcr.io/jaegertracing/jaeger-clickhouse:0.7.0
+    options:
+      grpc-storage-plugin:
+        binary: /plugin/jaeger-clickhouse
+        configuration-file: /plugin-config/config.yaml
+        log-level: debug
+  volumeMounts:
+    - name: plugin-config
+      mountPath: /plugin-config
+  volumes:
+    - name: plugin-config
+      configMap:
+        name: jaeger-clickhouse
+EOF
+```
+
+## Delete all
+
+```bash
+kubectl delete jaeger,cm,chi -l jaeger-clickhouse=demo
+```
--- a/guide-multitenancy.md
+++ b/guide-multitenancy.md
@ -0,0 +1,78 @@
+# Multi-tenant deployment
+
+It may be desirable to share a common ClickHouse instance across multiple Jaeger instances.
+There are two ways of doing this, depending on whether spanning the tenants across separate databases is preferable.
+
+## Shared database/tables
+
+If you wish to reuse the same ClickHouse database/tables across all tenants, you can specify a different `tenant: "<name>"` in each jaeger-clickhouse instance config.
+
+When a non-empty `tenant` is specified, all tables will be created with a `tenant` column, and all reads/writes for a given Jaeger instance will be applied against the configured tenant name for that instance.
+
+1. Create a shared database:
+    ```sql
+    CREATE DATABASE shared ENGINE=Atomic
+    ```
+2. Configure the per-tenant jaeger-clickhouse clients to specify tenant names:
+    ```yaml
+    database: shared
+    tenant: tenant_1
+    ```
+    ```yaml
+    database: shared
+    tenant: tenant_2
+    ```
+
+Multitenant mode must be enabled when the deployment is first created and cannot be toggled later, except perhaps by manually adding/removing the `tenant` column from all tables.
+Multitenant/singletenant instances must not be mixed within the same database - the two modes are mutually exclusive of each other.
+
+## Separate databases
+
+If you wish to keep instances fully separate, you can configure one ClickHouse database per tenant.
+This may be useful when different per-database configuration across tenants is desirable.
+
+1. Create a database for each tenant:
+    ```sql
+    CREATE DATABASE tenant_1 ENGINE=Atomic;
+    CREATE DATABASE tenant_2 ENGINE=Atomic;
+    ```
+2. Configure the per-tenant jaeger-clickhouse plugins matching databases:
+    ```yaml
+    database: tenant_1
+    ```
+    ```yaml
+    database: tenant_2
+    ```
+
+## Mixing methods in the same ClickHouse instance
+
+Each of the methods applies on a per-database basis. The methods require different schemas and must not be mixed in a single database, but it is possible to have different databases using different methods in the same ClickHouse instance.
+
+For example, there could be a `shared` database where multiple tenants are sharing the same tables:
+
+    ```sql
+    CREATE DATABASE shared ENGINE=Atomic
+    ```
+    ```yaml
+    database: shared
+    tenant: tenant_1
+    ```
+    ```yaml
+    database: shared
+    tenant: tenant_2
+    ```
+
+Then there could be separate `isolated_x` databases for tenants that should be provided with their own dedicated tables, enabling e.g. better ACL isolation:
+
+    ```sql
+    CREATE DATABASE isolated_1 ENGINE=Atomic
+    CREATE DATABASE isolated_2 ENGINE=Atomic
+    ```
+    ```yaml
+    database: isolated_1
+    ```
+    ```yaml
+    database: isolated_2
+    ```
+
+
--- a/guide-sharding-and-replication.md
+++ b/guide-sharding-and-replication.md
@ -0,0 +1,208 @@
+# Sharding and Replication
+
+This is a guide how to setup sharding and replication for Jaeger data.
+This guide uses [clickhouse-operator](https://github.com/Altinity/clickhouse-operator) to deploy
+the storage.
+
+Note that the Jaeger ClickHouse plugin supports creating replicated schema out-of-the-box. Therefore,
+this guide is not necessary for setting up default replicated deployment. Also note that the
+ClickHouse operator uses by default `Ordinary` database engine, which does not work with the
+embedded replication scripts in Jaeger.
+Refer to the `config.yaml` how to setup replicated deployment.
+
+## Sharding
+
+Sharding is a feature that allows splitting the data into multiple Clickhouse nodes to
+increase throughput and decrease latency.
+The sharding feature uses `Distributed` engine that is backed by local tables.
+The distributed engine is a "virtual" table that does not store any data. It is used as
+an interface to insert and query data.
+
+To setup sharding run the following statements on all nodes in the cluster.
+The "local" tables have to be created on the nodes before the distributed table.
+
+```sql
+CREATE DATABASE jaeger ENGINE=Atomic;
+USE jaeger;
+
+CREATE TABLE IF NOT EXISTS jaeger_spans AS jaeger_spans_local ENGINE = Distributed('{cluster}', default, jaeger_spans_local, cityHash64(traceID));
+CREATE TABLE IF NOT EXISTS jaeger_index AS jaeger_index_local ENGINE = Distributed('{cluster}', default, jaeger_index_local, cityHash64(traceID));
+CREATE TABLE IF NOT EXISTS jaeger_operations AS jaeger_operations_local ENGINE = Distributed('{cluster}', default, jaeger_operations_local, rand());
+```
+
+* The `AS <table-name>` statement creates table with the same schema as the specified one.
+* The `Distributed` engine takes as parameters cluster , database, table name and sharding key.
+
+If the distributed table is not created on all Clickhouse nodes the Jaeger query fails to get the data from the storage.
+
+### Deploy Clickhouse
+
+Deploy Clickhouse with 2 shards:
+
+```yaml
+cat <<EOF | kubectl apply -f -
+apiVersion: clickhouse.altinity.com/v1
+kind: ClickHouseInstallation
+metadata:
+  name: jaeger
+spec:
+  configuration:
+    clusters:
+      - name: cluster1
+        layout:
+          shardsCount: 2
+EOF
+```
+
+Use the following command to run `clickhouse-client` on Clickhouse nodes and create the distributed tables:
+```bash
+kubectl exec -it statefulset.apps/chi-jaeger-cluster1-0-0 -- clickhouse-client
+```
+
+### Plugin configuration
+
+The plugin has to be configured to write and read that from the global tables:
+
+```yaml
+address: clickhouse-jaeger:9000
+# database: jaeger
+spans_table: jaeger_spans
+spans_index_table: jaeger_index
+operations_table: jaeger_operations
+```
+
+## Replication
+
+Replication as the name suggest automatically replicates the data across multiple Clickhouse nodes.
+It is used to accomplish high availability, load scaling and migration/updates.
+
+The replication uses Zookeeper. Refer to the Clickhouse operator how to deploy Zookeeper.
+
+Zookeeper allows us to use `ON CLUSTER` to automatically replicate table creation on all nodes.
+Therefore the following command can be run only on a single Clickhouse node:
+
+```sql
+CREATE DATABASE IF NOT EXISTS jaeger ON CLUSTER '{cluster}' ENGINE=Atomic;
+USE jaeger;
+
+CREATE TABLE IF NOT EXISTS jaeger_spans_local ON CLUSTER '{cluster}' (
+    timestamp DateTime CODEC(Delta, ZSTD(1)),
+    traceID String CODEC(ZSTD(1)),
+    model String CODEC(ZSTD(3))
+) ENGINE ReplicatedMergeTree
+PARTITION BY toDate(timestamp)
+ORDER BY traceID
+SETTINGS index_granularity=1024;
+
+CREATE TABLE IF NOT EXISTS jaeger_index_local ON CLUSTER '{cluster}' (
+    timestamp DateTime CODEC(Delta, ZSTD(1)),
+    traceID String CODEC(ZSTD(1)),
+    service LowCardinality(String) CODEC(ZSTD(1)),
+    operation LowCardinality(String) CODEC(ZSTD(1)),
+    durationUs UInt64 CODEC(ZSTD(1)),
+    tags Array(String) CODEC(ZSTD(1)),
+    INDEX idx_tags tags TYPE bloom_filter(0.01) GRANULARITY 64,
+    INDEX idx_duration durationUs TYPE minmax GRANULARITY 1
+) ENGINE ReplicatedMergeTree
+PARTITION BY toDate(timestamp)
+ORDER BY (service, -toUnixTimestamp(timestamp))
+SETTINGS index_granularity=1024;
+
+CREATE MATERIALIZED VIEW IF NOT EXISTS jaeger_operations_local ON CLUSTER '{cluster}'
+ENGINE ReplicatedMergeTree
+PARTITION BY toYYYYMM(date) ORDER BY (date, service, operation)
+SETTINGS index_granularity=32
+POPULATE
+AS SELECT
+    toDate(timestamp) AS date,
+    service,
+    operation,
+count() as count
+FROM jaeger.jaeger_index_local
+GROUP BY date, service, operation;
+
+
+CREATE TABLE IF NOT EXISTS jaeger_spans ON CLUSTER '{cluster}' AS jaeger.jaeger_spans_local ENGINE = Distributed('{cluster}', jaeger, jaeger_spans_local, cityHash64(traceID));
+CREATE TABLE IF NOT EXISTS jaeger_index ON CLUSTER '{cluster}' AS jaeger.jaeger_index_local ENGINE = Distributed('{cluster}', jaeger, jaeger_index_local, cityHash64(traceID));
+CREATE TABLE IF NOT EXISTS jaeger_operations on CLUSTER '{cluster}' AS jaeger.jaeger_operations_local ENGINE = Distributed('{cluster}', jaeger, jaeger_operations_local, rand());
+```
+
+### Deploy Clickhouse
+
+Before deploying Clickhouse make sure Zookeeper is running in `zoo1ns` namespace.
+
+Deploy Clickhouse with 3 shards and 2 replicas. In total Clickhouse operator will deploy 6 pods:
+
+```yaml
+cat <<EOF | kubectl apply -f -
+apiVersion: clickhouse.altinity.com/v1
+kind: ClickHouseInstallation
+metadata:
+  name: jaeger
+spec:
+  defaults:
+    templates:
+      dataVolumeClaimTemplate: data-volume-template
+      logVolumeClaimTemplate: log-volume-template
+  configuration:
+    zookeeper:
+      nodes:
+        - host: zookeeper.zoo1ns
+    clusters:
+      - name: cluster1
+        layout:
+          shardsCount: 3
+          replicasCount: 2
+  templates:
+    volumeClaimTemplates:
+      - name: data-volume-template
+        spec:
+          accessModes:
+            - ReadWriteOnce
+          resources:
+            requests:
+              storage: 1Gi
+      - name: log-volume-template
+        spec:
+          accessModes:
+            - ReadWriteOnce
+          resources:
+            requests:
+              storage: 100Mi
+EOF
+```
+
+The Clickhouse deployment will look like this:
+```bash
+k get statefulsets
+NAME                      READY   AGE
+chi-jaeger-cluster1-0-0   1/1     17m    # shard 0
+chi-jaeger-cluster1-0-1   1/1     17m    # shard 0, replica 1
+chi-jaeger-cluster1-1-0   1/1     16m    # shard 1
+chi-jaeger-cluster1-1-1   1/1     16m    # shard 1, replica 1
+chi-jaeger-cluster1-2-0   1/1     7m43s  # shard 2
+chi-jaeger-cluster1-2-1   1/1     7m26s  # shard 2, replica 1
+```
+
+#### Scaling up
+
+Just increase `shardsCount` number and new Clickhouse node will come up. It will have initialized Jaeger tables so
+no other steps are required. Note that the old data are not re-balanced, only new writes take into the account
+the new node.
+
+## Useful Commands
+
+### SQL
+
+```sql
+show tables;
+select count() from jaeger_spans;
+```
+
+### Kubectl
+
+```bash
+kubectl get chi -o wide
+kubectl port-forward service/clickhouse-jaeger 9000:9000
+kubectl delete chi jaeger
+```
--- a/integration/config-local.yaml
+++ b/integration/config-local.yaml
@ -0,0 +1,3 @@
+address: localhost:9000
+init_sql_scripts_dir: init_sql_scripts
+init_tables: true
--- a/integration/grpc_test.go
+++ b/integration/grpc_test.go
@ -0,0 +1,107 @@
+// Copyright (c) 2019 The Jaeger Authors.
+// Copyright (c) 2018 Uber Technologies, Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package integration
+
+import (
+	"os"
+	"testing"
+
+	"github.com/jaegertracing/jaeger/pkg/config"
+	"github.com/jaegertracing/jaeger/pkg/metrics"
+	"github.com/jaegertracing/jaeger/pkg/testutils"
+	"github.com/jaegertracing/jaeger/plugin/storage/grpc"
+	"github.com/jaegertracing/jaeger/plugin/storage/integration"
+	"github.com/stretchr/testify/require"
+	"go.uber.org/zap"
+)
+
+const defaultPluginBinaryPath = "../../../examples/memstore-plugin/memstore-plugin"
+
+type GRPCStorageIntegrationTestSuite struct {
+	integration.StorageIntegration
+	logger           *zap.Logger
+	pluginBinaryPath string
+	pluginConfigPath string
+}
+
+func (s *GRPCStorageIntegrationTestSuite) initialize() error {
+	s.logger, _ = testutils.NewLogger()
+
+	f := grpc.NewFactory()
+	v, command := config.Viperize(f.AddFlags)
+	flags := []string{
+		"--grpc-storage-plugin.binary",
+		s.pluginBinaryPath,
+		"--grpc-storage-plugin.log-level",
+		"debug",
+	}
+	if s.pluginConfigPath != "" {
+		flags = append(flags,
+			"--grpc-storage-plugin.configuration-file",
+			s.pluginConfigPath,
+		)
+	}
+	err := command.ParseFlags(flags)
+	if err != nil {
+		return err
+	}
+	f.InitFromViper(v, zap.NewNop())
+	if err = f.Initialize(metrics.NullFactory, s.logger); err != nil {
+		return err
+	}
+
+	if s.SpanWriter, err = f.CreateSpanWriter(); err != nil {
+		return err
+	}
+	if s.SpanReader, err = f.CreateSpanReader(); err != nil {
+		return err
+	}
+
+	// TODO DependencyWriter is not implemented in grpc store
+
+	s.Refresh = s.refresh
+	s.CleanUp = s.cleanUp
+	return nil
+}
+
+func (s *GRPCStorageIntegrationTestSuite) refresh() error {
+	return nil
+}
+
+func (s *GRPCStorageIntegrationTestSuite) cleanUp() error {
+	return s.initialize()
+}
+
+func TestGRPCStorage(t *testing.T) {
+	if os.Getenv("STORAGE") != "grpc-plugin" {
+		t.Skip("Integration test against grpc skipped; set STORAGE env var to grpc-plugin to run this")
+	}
+	binaryPath := os.Getenv("PLUGIN_BINARY_PATH")
+	if binaryPath == "" {
+		t.Logf("PLUGIN_BINARY_PATH env var not set, using %s", defaultPluginBinaryPath)
+		binaryPath = defaultPluginBinaryPath
+	}
+	configPath := os.Getenv("PLUGIN_CONFIG_PATH")
+	if configPath == "" {
+		t.Log("PLUGIN_CONFIG_PATH env var not set")
+	}
+	s := &GRPCStorageIntegrationTestSuite{
+		pluginBinaryPath: binaryPath,
+		pluginConfigPath: configPath,
+	}
+	require.NoError(t, s.initialize())
+	s.IntegrationTestAll(t)
+}
--- a/integration/init_sql_scripts/0001-drop-database.sql
+++ b/integration/init_sql_scripts/0001-drop-database.sql
@ -0,0 +1 @@
+DROP DATABASE IF EXISTS default;
--- a/integration/init_sql_scripts/0002-create-database.sql
+++ b/integration/init_sql_scripts/0002-create-database.sql
@ -0,0 +1 @@
+CREATE DATABASE IF NOT EXISTS default;
--- a/internal/tools/go.mod
+++ b/internal/tools/go.mod
@ -1,5 +1,145 @@
-module github.com/pavolloffay/jaeger-clickhouse/internal/tools
+module github.com/jaegertracing/jaeger-clickhouse/internal/tools

-go 1.16
+go 1.19

-require golang.org/x/tools v0.1.5
+require (
+	github.com/golangci/golangci-lint v1.41.1
+	golang.org/x/tools v0.1.5
+)
+
+require (
+	4d63.com/gochecknoglobals v0.0.0-20201008074935-acfc0b28355a // indirect
+	github.com/BurntSushi/toml v0.3.1 // indirect
+	github.com/Djarvur/go-err113 v0.0.0-20210108212216-aea10b59be24 // indirect
+	github.com/Masterminds/semver v1.5.0 // indirect
+	github.com/OpenPeeDeeP/depguard v1.0.1 // indirect
+	github.com/alexkohler/prealloc v1.0.0 // indirect
+	github.com/ashanbrown/forbidigo v1.2.0 // indirect
+	github.com/ashanbrown/makezero v0.0.0-20210520155254-b6261585ddde // indirect
+	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/bkielbasa/cyclop v1.2.0 // indirect
+	github.com/bombsimon/wsl/v3 v3.3.0 // indirect
+	github.com/cespare/xxhash/v2 v2.1.1 // indirect
+	github.com/charithe/durationcheck v0.0.8 // indirect
+	github.com/chavacava/garif v0.0.0-20210405164556-e8a0a408d6af // indirect
+	github.com/daixiang0/gci v0.2.8 // indirect
+	github.com/davecgh/go-spew v1.1.1 // indirect
+	github.com/denis-tingajkin/go-header v0.4.2 // indirect
+	github.com/esimonov/ifshort v1.0.2 // indirect
+	github.com/ettle/strcase v0.1.1 // indirect
+	github.com/fatih/color v1.12.0 // indirect
+	github.com/fatih/structtag v1.2.0 // indirect
+	github.com/fsnotify/fsnotify v1.4.9 // indirect
+	github.com/fzipp/gocyclo v0.3.1 // indirect
+	github.com/go-critic/go-critic v0.5.6 // indirect
+	github.com/go-toolsmith/astcast v1.0.0 // indirect
+	github.com/go-toolsmith/astcopy v1.0.0 // indirect
+	github.com/go-toolsmith/astequal v1.0.0 // indirect
+	github.com/go-toolsmith/astfmt v1.0.0 // indirect
+	github.com/go-toolsmith/astp v1.0.0 // indirect
+	github.com/go-toolsmith/strparse v1.0.0 // indirect
+	github.com/go-toolsmith/typep v1.0.2 // indirect
+	github.com/go-xmlfmt/xmlfmt v0.0.0-20191208150333-d5b6f63a941b // indirect
+	github.com/gobwas/glob v0.2.3 // indirect
+	github.com/gofrs/flock v0.8.0 // indirect
+	github.com/golang/protobuf v1.4.3 // indirect
+	github.com/golangci/check v0.0.0-20180506172741-cfe4005ccda2 // indirect
+	github.com/golangci/dupl v0.0.0-20180902072040-3e9179ac440a // indirect
+	github.com/golangci/go-misc v0.0.0-20180628070357-927a3d87b613 // indirect
+	github.com/golangci/gofmt v0.0.0-20190930125516-244bba706f1a // indirect
+	github.com/golangci/lint-1 v0.0.0-20191013205115-297bf364a8e0 // indirect
+	github.com/golangci/maligned v0.0.0-20180506175553-b1d89398deca // indirect
+	github.com/golangci/misspell v0.3.5 // indirect
+	github.com/golangci/revgrep v0.0.0-20210208091834-cd28932614b5 // indirect
+	github.com/golangci/unconvert v0.0.0-20180507085042-28b1c447d1f4 // indirect
+	github.com/google/go-cmp v0.5.4 // indirect
+	github.com/gordonklaus/ineffassign v0.0.0-20210225214923-2e10b2664254 // indirect
+	github.com/gostaticanalysis/analysisutil v0.4.1 // indirect
+	github.com/gostaticanalysis/comment v1.4.1 // indirect
+	github.com/gostaticanalysis/forcetypeassert v0.0.0-20200621232751-01d4955beaa5 // indirect
+	github.com/gostaticanalysis/nilerr v0.1.1 // indirect
+	github.com/hashicorp/errwrap v1.0.0 // indirect
+	github.com/hashicorp/go-multierror v1.1.1 // indirect
+	github.com/hashicorp/hcl v1.0.0 // indirect
+	github.com/inconshreveable/mousetrap v1.0.0 // indirect
+	github.com/jgautheron/goconst v1.5.1 // indirect
+	github.com/jingyugao/rowserrcheck v1.1.0 // indirect
+	github.com/jirfag/go-printf-func-name v0.0.0-20200119135958-7558a9eaa5af // indirect
+	github.com/julz/importas v0.0.0-20210419104244-841f0c0fe66d // indirect
+	github.com/kisielk/errcheck v1.6.0 // indirect
+	github.com/kisielk/gotool v1.0.0 // indirect
+	github.com/kulti/thelper v0.4.0 // indirect
+	github.com/kunwardeep/paralleltest v1.0.2 // indirect
+	github.com/kyoh86/exportloopref v0.1.8 // indirect
+	github.com/ldez/gomoddirectives v0.2.1 // indirect
+	github.com/ldez/tagliatelle v0.2.0 // indirect
+	github.com/magiconair/properties v1.8.1 // indirect
+	github.com/maratori/testpackage v1.0.1 // indirect
+	github.com/matoous/godox v0.0.0-20210227103229-6504466cf951 // indirect
+	github.com/mattn/go-colorable v0.1.8 // indirect
+	github.com/mattn/go-isatty v0.0.12 // indirect
+	github.com/mattn/go-runewidth v0.0.9 // indirect
+	github.com/matttproud/golang_protobuf_extensions v1.0.1 // indirect
+	github.com/mbilski/exhaustivestruct v1.2.0 // indirect
+	github.com/mgechev/dots v0.0.0-20190921121421-c36f7dcfbb81 // indirect
+	github.com/mgechev/revive v1.0.7 // indirect
+	github.com/mitchellh/go-homedir v1.1.0 // indirect
+	github.com/mitchellh/mapstructure v1.1.2 // indirect
+	github.com/moricho/tparallel v0.2.1 // indirect
+	github.com/nakabonne/nestif v0.3.0 // indirect
+	github.com/nbutton23/zxcvbn-go v0.0.0-20210217022336-fa2cb2858354 // indirect
+	github.com/nishanths/exhaustive v0.1.0 // indirect
+	github.com/nishanths/predeclared v0.2.1 // indirect
+	github.com/olekukonko/tablewriter v0.0.5 // indirect
+	github.com/pelletier/go-toml v1.2.0 // indirect
+	github.com/phayes/checkstyle v0.0.0-20170904204023-bfd46e6a821d // indirect
+	github.com/pkg/errors v0.9.1 // indirect
+	github.com/pmezard/go-difflib v1.0.0 // indirect
+	github.com/polyfloyd/go-errorlint v0.0.0-20210510181950-ab96adb96fea // indirect
+	github.com/prometheus/client_golang v1.7.1 // indirect
+	github.com/prometheus/client_model v0.2.0 // indirect
+	github.com/prometheus/common v0.10.0 // indirect
+	github.com/prometheus/procfs v0.1.3 // indirect
+	github.com/quasilyte/go-ruleguard v0.3.4 // indirect
+	github.com/quasilyte/regex/syntax v0.0.0-20200407221936-30656e2c4a95 // indirect
+	github.com/ryancurrah/gomodguard v1.2.2 // indirect
+	github.com/ryanrolds/sqlclosecheck v0.3.0 // indirect
+	github.com/sanposhiho/wastedassign/v2 v2.0.6 // indirect
+	github.com/securego/gosec/v2 v2.8.0 // indirect
+	github.com/shazow/go-diff v0.0.0-20160112020656-b6b7b6733b8c // indirect
+	github.com/sirupsen/logrus v1.8.1 // indirect
+	github.com/sonatard/noctx v0.0.1 // indirect
+	github.com/sourcegraph/go-diff v0.6.1 // indirect
+	github.com/spf13/afero v1.1.2 // indirect
+	github.com/spf13/cast v1.3.0 // indirect
+	github.com/spf13/cobra v1.1.3 // indirect
+	github.com/spf13/jwalterweatherman v1.0.0 // indirect
+	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/spf13/viper v1.7.1 // indirect
+	github.com/ssgreg/nlreturn/v2 v2.1.0 // indirect
+	github.com/stretchr/objx v0.1.1 // indirect
+	github.com/stretchr/testify v1.7.0 // indirect
+	github.com/subosito/gotenv v1.2.0 // indirect
+	github.com/tdakkota/asciicheck v0.0.0-20200416200610-e657995f937b // indirect
+	github.com/tetafro/godot v1.4.7 // indirect
+	github.com/timakin/bodyclose v0.0.0-20200424151742-cb6215831a94 // indirect
+	github.com/tomarrell/wrapcheck/v2 v2.1.0 // indirect
+	github.com/tommy-muehle/go-mnd/v2 v2.4.0 // indirect
+	github.com/ultraware/funlen v0.0.3 // indirect
+	github.com/ultraware/whitespace v0.0.4 // indirect
+	github.com/uudashr/gocognit v1.0.1 // indirect
+	github.com/yeya24/promlinter v0.1.0 // indirect
+	golang.org/x/mod v0.4.2 // indirect
+	golang.org/x/sys v0.0.0-20210510120138-977fb7262007 // indirect
+	golang.org/x/text v0.3.5 // indirect
+	golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 // indirect
+	google.golang.org/protobuf v1.25.0 // indirect
+	gopkg.in/ini.v1 v1.51.0 // indirect
+	gopkg.in/yaml.v2 v2.4.0 // indirect
+	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b // indirect
+	honnef.co/go/tools v0.2.0 // indirect
+	mvdan.cc/gofumpt v0.1.1 // indirect
+	mvdan.cc/interfacer v0.0.0-20180901003855-c20040233aed // indirect
+	mvdan.cc/lint v0.0.0-20170908181259-adc824a0674b // indirect
+	mvdan.cc/unparam v0.0.0-20210104141923-aac4ce9116a7 // indirect
+)
--- a/internal/tools/go.sum
+++ b/internal/tools/go.sum
--- a/internal/tools/tools.go
+++ b/internal/tools/tools.go
@ -1,5 +1,6 @@
 package tools

 import (
+	_ "github.com/golangci/golangci-lint/cmd/golangci-lint"
 	_ "golang.org/x/tools/cmd/goimports"
 )
--- a/jaeger-ui.json
+++ b/jaeger-ui.json
@ -0,0 +1,8 @@
+{
+  "dependencies": {
+    "dagMaxNumServices": 200,
+    "menuEnabled": true
+  },
+  "archiveEnabled": true
+}
+
--- a/pictures/tables.png
+++ b/pictures/tables.png
--- a/sqlscripts/0001-jaeger-index.sql
+++ b/sqlscripts/0001-jaeger-index.sql
@ -1,13 +0,0 @@
-CREATE TABLE IF NOT EXISTS jaeger_index_v2 (
-                                 timestamp DateTime CODEC(Delta, ZSTD(1)),
-                                 traceID String CODEC(ZSTD(1)),
-                                 service LowCardinality(String) CODEC(ZSTD(1)),
-                                 operation LowCardinality(String) CODEC(ZSTD(1)),
-                                 durationUs UInt64 CODEC(ZSTD(1)),
-                                 tags Array(String) CODEC(ZSTD(1)),
-                                 INDEX idx_tags tags TYPE bloom_filter(0.01) GRANULARITY 64,
-                                 INDEX idx_duration durationUs TYPE minmax GRANULARITY 1
-) ENGINE MergeTree()
-PARTITION BY toDate(timestamp)
-ORDER BY (service, -toUnixTimestamp(timestamp))
-SETTINGS index_granularity=1024
--- a/sqlscripts/0002-jaeger-spans.sql
+++ b/sqlscripts/0002-jaeger-spans.sql
@ -1,8 +0,0 @@
-CREATE TABLE IF NOT EXISTS jaeger_spans_v2 (
-                                 timestamp DateTime CODEC(Delta, ZSTD(1)),
-                                 traceID String CODEC(ZSTD(1)),
-                                 model String CODEC(ZSTD(3))
-) ENGINE MergeTree()
-PARTITION BY toDate(timestamp)
-ORDER BY traceID
-SETTINGS index_granularity=1024
--- a/sqlscripts/0003-jaeger-operations.sql
+++ b/sqlscripts/0003-jaeger-operations.sql
@ -1,12 +0,0 @@
-CREATE MATERIALIZED VIEW IF NOT EXISTS jaeger_operations_v2
-ENGINE SummingMergeTree
-PARTITION BY toYYYYMM(date) ORDER BY (date, service, operation)
-SETTINGS index_granularity=32
-POPULATE
-AS SELECT
-              toDate(timestamp) AS date,
-  service,
-  operation,
-  count() as count
-   FROM jaeger_index_v2
-   GROUP BY date, service, operation
--- a/sqlscripts/distributed-table.tmpl.sql
+++ b/sqlscripts/distributed-table.tmpl.sql
@ -0,0 +1,3 @@
+CREATE TABLE IF NOT EXISTS {{.Table}}
+    ON CLUSTER '{cluster}' AS {{.Database}}.{{.Table}}_local
+    ENGINE = Distributed('{cluster}', {{.Database}}, {{.Table}}_local, {{.Hash}})
--- a/sqlscripts/jaeger-index.tmpl.sql
+++ b/sqlscripts/jaeger-index.tmpl.sql
@ -0,0 +1,28 @@
+CREATE TABLE IF NOT EXISTS {{.SpansIndexTable}}
+{{if .Replication}}ON CLUSTER '{cluster}'{{end}}
+(
+    {{if .Multitenant -}}
+    tenant     LowCardinality(String) CODEC (ZSTD(1)),
+    {{- end -}}
+    timestamp  DateTime CODEC (Delta, ZSTD(1)),
+    traceID    String CODEC (ZSTD(1)),
+    service    LowCardinality(String) CODEC (ZSTD(1)),
+    operation  LowCardinality(String) CODEC (ZSTD(1)),
+    durationUs UInt64 CODEC (ZSTD(1)),
+    tags Nested
+    (
+        key LowCardinality(String),
+        value String
+    ) CODEC (ZSTD(1)),
+    INDEX idx_tag_keys tags.key TYPE bloom_filter(0.01) GRANULARITY 64,
+    INDEX idx_duration durationUs TYPE minmax GRANULARITY 1
+) ENGINE {{if .Replication}}ReplicatedMergeTree{{else}}MergeTree(){{end}}
+    {{.TTLTimestamp}}
+    PARTITION BY (
+        {{if .Multitenant -}}
+        tenant,
+        {{- end -}}
+        toDate(timestamp)
+    )
+    ORDER BY (service, -toUnixTimestamp(timestamp))
+    SETTINGS index_granularity = 1024
--- a/sqlscripts/jaeger-operations.tmpl.sql
+++ b/sqlscripts/jaeger-operations.tmpl.sql
@ -0,0 +1,43 @@
+CREATE MATERIALIZED VIEW IF NOT EXISTS {{.OperationsTable}}
+{{if .Replication}}ON CLUSTER '{cluster}'{{end}}
+    ENGINE {{if .Replication}}ReplicatedSummingMergeTree{{else}}SummingMergeTree{{end}}
+    {{.TTLDate}}
+    PARTITION BY (
+        {{if .Multitenant -}}
+        tenant,
+        {{- end -}}
+        toYYYYMM(date)
+    )
+    ORDER BY (
+        {{if .Multitenant -}}
+        tenant,
+        {{- end -}}
+        date,
+        service,
+        operation
+    )
+    SETTINGS index_granularity = 32
+    POPULATE
+AS SELECT
+    {{if .Multitenant -}}
+    tenant,
+    {{- end -}}
+    toDate(timestamp) AS date,
+    service,
+    operation,
+    count() AS count,
+    if(
+        has(tags.key, 'span.kind'),
+        tags.value[indexOf(tags.key, 'span.kind')],
+        ''
+    ) AS spankind
+FROM {{.Database}}.{{.SpansIndexTable}}
+GROUP BY
+    {{if .Multitenant -}}
+    tenant,
+    {{- end -}}
+    date,
+    service,
+    operation,
+    tags.key,
+    tags.value
--- a/sqlscripts/jaeger-spans-archive.tmpl.sql
+++ b/sqlscripts/jaeger-spans-archive.tmpl.sql
@ -0,0 +1,19 @@
+CREATE TABLE IF NOT EXISTS {{.SpansArchiveTable}}
+{{if .Replication}}ON CLUSTER '{cluster}'{{end}}
+(
+    {{if .Multitenant -}}
+    tenant    LowCardinality(String) CODEC (ZSTD(1)),
+    {{- end -}}
+    timestamp DateTime CODEC (Delta, ZSTD(1)),
+    traceID   String CODEC (ZSTD(1)),
+    model     String CODEC (ZSTD(3))
+) ENGINE {{if .Replication}}ReplicatedMergeTree{{else}}MergeTree(){{end}}
+    {{.TTLTimestamp}}
+    PARTITION BY (
+        {{if .Multitenant -}}
+        tenant,
+        {{- end -}}
+        toYYYYMM(timestamp)
+    )
+    ORDER BY traceID
+    SETTINGS index_granularity = 1024
--- a/sqlscripts/jaeger-spans.tmpl.sql
+++ b/sqlscripts/jaeger-spans.tmpl.sql
@ -0,0 +1,19 @@
+CREATE TABLE IF NOT EXISTS {{.SpansTable}}
+{{if .Replication}}ON CLUSTER '{cluster}'{{end}}
+(
+    {{if .Multitenant -}}
+    tenant    LowCardinality(String) CODEC (ZSTD(1)),
+    {{- end -}}
+    timestamp DateTime CODEC (Delta, ZSTD(1)),
+    traceID   String CODEC (ZSTD(1)),
+    model     String CODEC (ZSTD(3))
+) ENGINE {{if .Replication}}ReplicatedMergeTree{{else}}MergeTree(){{end}}
+    {{.TTLTimestamp}}
+    PARTITION BY (
+        {{if .Multitenant -}}
+        tenant,
+        {{- end -}}
+        toDate(timestamp)
+    )
+    ORDER BY traceID
+    SETTINGS index_granularity = 1024
--- a/storage/clickhousedependencystore/dependencystore_test.go
+++ b/storage/clickhousedependencystore/dependencystore_test.go
@ -0,0 +1,18 @@
+package clickhousedependencystore
+
+import (
+	"context"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestDependencyStore_GetDependencies(t *testing.T) {
+	dependencyStore := NewDependencyStore()
+
+	dependencies, err := dependencyStore.GetDependencies(context.Background(), time.Now(), time.Hour)
+
+	assert.EqualError(t, err, errNotImplemented.Error())
+	assert.Nil(t, dependencies)
+}
--- a/storage/clickhousespanstore/heap.go
+++ b/storage/clickhousespanstore/heap.go
@ -0,0 +1,85 @@
+package clickhousespanstore
+
+import (
+	"container/heap"
+	"fmt"
+	"time"
+)
+
+var (
+	_                 heap.Interface = workerHeap{}
+	errWorkerNotFound                = fmt.Errorf("worker not found in heap")
+)
+
+type heapItem struct {
+	pushTime time.Time
+	worker   *WriteWorker
+}
+
+// workerHeap is a heap for WriteWorkers where worker's push time is the key.
+type workerHeap struct {
+	elems   *[]*heapItem
+	indexes map[*WriteWorker]int
+}
+
+func newWorkerHeap(cap int) workerHeap {
+	elems := make([]*heapItem, 0, cap)
+	return workerHeap{
+		elems:   &elems,
+		indexes: make(map[*WriteWorker]int),
+	}
+}
+
+func (workerHeap workerHeap) AddWorker(worker *WriteWorker) {
+	heap.Push(workerHeap, heapItem{
+		worker:   worker,
+		pushTime: time.Now(),
+	})
+}
+
+func (workerHeap *workerHeap) RemoveWorker(worker *WriteWorker) error {
+	idx, ok := workerHeap.indexes[worker]
+	if !ok {
+		return errWorkerNotFound
+	}
+	heap.Remove(workerHeap, idx)
+	return nil
+}
+
+func (workerHeap *workerHeap) CloseWorkers() {
+	for _, item := range *workerHeap.elems {
+		item.worker.Close()
+	}
+}
+
+func (workerHeap workerHeap) Len() int {
+	return len(*workerHeap.elems)
+}
+
+func (workerHeap workerHeap) Less(i, j int) bool {
+	return (*workerHeap.elems)[i].pushTime.Before((*workerHeap.elems)[j].pushTime)
+}
+
+func (workerHeap workerHeap) Swap(i, j int) {
+	(*workerHeap.elems)[i], (*workerHeap.elems)[j] = (*workerHeap.elems)[j], (*workerHeap.elems)[i]
+	workerHeap.indexes[(*workerHeap.elems)[i].worker] = i
+	workerHeap.indexes[(*workerHeap.elems)[j].worker] = j
+}
+
+func (workerHeap workerHeap) Push(x interface{}) {
+	switch t := x.(type) {
+	case heapItem:
+		*workerHeap.elems = append(*workerHeap.elems, &t)
+		workerHeap.indexes[t.worker] = len(*workerHeap.elems) - 1
+	default:
+		panic("Unknown type")
+	}
+}
+
+func (workerHeap workerHeap) Pop() interface{} {
+	lastInd := len(*workerHeap.elems) - 1
+	last := (*workerHeap.elems)[lastInd]
+	delete(workerHeap.indexes, last.worker)
+	*workerHeap.elems = (*workerHeap.elems)[:lastInd]
+	return last.worker
+}
--- a/storage/clickhousespanstore/mocks/converter.go
+++ b/storage/clickhousespanstore/mocks/converter.go
@ -0,0 +1,40 @@
+package mocks
+
+import (
+	"database/sql/driver"
+	"fmt"
+	"time"
+
+	"github.com/jaegertracing/jaeger/model"
+)
+
+var _ driver.ValueConverter = ConverterMock{}
+
+type ConverterMock struct{}
+
+func (conv ConverterMock) ConvertValue(v interface{}) (driver.Value, error) {
+	switch t := v.(type) {
+	case model.TraceID:
+		return driver.Value(t.String()), nil
+	case time.Time:
+		return driver.Value(t), nil
+	case time.Duration:
+		return driver.Value(t.Nanoseconds()), nil
+	case model.SpanID:
+		return driver.Value(t), nil
+	case string:
+		return driver.Value(t), nil
+	case []uint8:
+		return driver.Value(t), nil
+	case int64:
+		return driver.Value(t), nil
+	case uint64:
+		return driver.Value(t), nil
+	case int:
+		return driver.Value(t), nil
+	case []string:
+		return driver.Value(fmt.Sprint(t)), nil
+	default:
+		return nil, fmt.Errorf("unknown type %T", t)
+	}
+}
--- a/storage/clickhousespanstore/mocks/converter_test.go
+++ b/storage/clickhousespanstore/mocks/converter_test.go
@ -0,0 +1,63 @@
+package mocks
+
+import (
+	"database/sql/driver"
+	"testing"
+	"time"
+
+	"github.com/jaegertracing/jaeger/model"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestConverterMock_ConvertValue(t *testing.T) {
+	converter := ConverterMock{}
+
+	testCases := map[string]struct {
+		valueToConvert interface{}
+		expectedResult driver.Value
+	}{
+		"string value":       {valueToConvert: "some string value", expectedResult: driver.Value("some string value")},
+		"string slice value": {valueToConvert: []string{"some", "slice", "of", "strings"}, expectedResult: driver.Value("[some slice of strings]")},
+		"time value": {
+			valueToConvert: time.Date(2002, time.February, 19, 14, 43, 51, 0, time.UTC),
+			expectedResult: driver.Value(time.Date(2002, time.February, 19, 14, 43, 51, 0, time.UTC)),
+		},
+		"duration value": {
+			valueToConvert: time.Unix(12340, 123456789).Sub(time.Unix(0, 0)),
+			expectedResult: driver.Value(int64(12340123456789)),
+		},
+		"int64 value":         {valueToConvert: int64(1823), expectedResult: driver.Value(int64(1823))},
+		"int value":           {valueToConvert: 1823, expectedResult: driver.Value(1823)},
+		"model.SpanID value":  {valueToConvert: model.SpanID(318148), expectedResult: driver.Value(model.SpanID(318148))},
+		"model.TraceID value": {valueToConvert: model.TraceID{Low: 0xabd5, High: 0xa31}, expectedResult: driver.Value("0000000000000a31000000000000abd5")},
+		"uint8 slice value":   {valueToConvert: []uint8("asdkja"), expectedResult: driver.Value([]uint8{0x61, 0x73, 0x64, 0x6b, 0x6a, 0x61})},
+	}
+
+	for name, test := range testCases {
+		t.Run(name, func(t *testing.T) {
+			converted, err := converter.ConvertValue(test.valueToConvert)
+			assert.NoError(t, err)
+			assert.Equal(t, test.expectedResult, converted)
+		})
+	}
+}
+
+func TestConverterMock_Fail(t *testing.T) {
+	converter := ConverterMock{}
+
+	tests := map[string]struct {
+		valueToConvert   interface{}
+		expectedErrorMsg string
+	}{
+		"float64 value": {valueToConvert: float64(1e-4), expectedErrorMsg: "unknown type float64"},
+		"int32 value":   {valueToConvert: int32(12831), expectedErrorMsg: "unknown type int32"},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			val, err := converter.ConvertValue(test.valueToConvert)
+			assert.Equal(t, nil, val)
+			assert.EqualError(t, err, test.expectedErrorMsg)
+		})
+	}
+}
--- a/storage/clickhousespanstore/mocks/db.go
+++ b/storage/clickhousespanstore/mocks/db.go
@ -0,0 +1,14 @@
+package mocks
+
+import (
+	"database/sql"
+
+	sqlmock "github.com/DATA-DOG/go-sqlmock"
+)
+
+func GetDbMock() (*sql.DB, sqlmock.Sqlmock, error) {
+	return sqlmock.New(
+		sqlmock.QueryMatcherOption(sqlmock.QueryMatcherEqual),
+		sqlmock.ValueConverterOption(ConverterMock{}),
+	)
+}
--- a/storage/clickhousespanstore/mocks/spylogger.go
+++ b/storage/clickhousespanstore/mocks/spylogger.go
@ -0,0 +1,115 @@
+package mocks
+
+import (
+	"io"
+	"log"
+	"testing"
+
+	hclog "github.com/hashicorp/go-hclog"
+	"github.com/stretchr/testify/assert"
+)
+
+const levelCount = 5
+
+var _ hclog.Logger = SpyLogger{}
+
+type LogMock struct {
+	Msg  string
+	Args []interface{}
+}
+
+type SpyLogger struct {
+	logs [][]LogMock
+}
+
+func NewSpyLogger() SpyLogger {
+	return SpyLogger{logs: make([][]LogMock, levelCount)}
+}
+
+func (logger *SpyLogger) AssertLogsOfLevelEqual(t *testing.T, level hclog.Level, want []LogMock) {
+	assert.Equal(t, want, logger.getLogs(level))
+}
+
+func (logger *SpyLogger) getLogs(level hclog.Level) []LogMock {
+	return logger.logs[level-1]
+}
+
+func (logger *SpyLogger) AssertLogsEmpty(t *testing.T) {
+	assert.Equal(t, logger.logs, make([][]LogMock, levelCount))
+}
+
+func (logger SpyLogger) Log(level hclog.Level, msg string, args ...interface{}) {
+	logger.logs[level-1] = append(logger.getLogs(level), LogMock{msg, args})
+}
+
+func (logger SpyLogger) Trace(msg string, args ...interface{}) {
+	logger.Log(hclog.Trace, msg, args...)
+}
+
+func (logger SpyLogger) Debug(msg string, args ...interface{}) {
+	logger.Log(hclog.Debug, msg, args...)
+}
+
+func (logger SpyLogger) Info(msg string, args ...interface{}) {
+	logger.Log(hclog.Info, msg, args...)
+}
+
+func (logger SpyLogger) Warn(msg string, args ...interface{}) {
+	logger.Log(hclog.Warn, msg, args...)
+}
+
+func (logger SpyLogger) Error(msg string, args ...interface{}) {
+	logger.Log(hclog.Error, msg, args...)
+}
+
+func (logger SpyLogger) IsTrace() bool {
+	panic("implement me")
+}
+
+func (logger SpyLogger) IsDebug() bool {
+	panic("implement me")
+}
+
+func (logger SpyLogger) IsInfo() bool {
+	panic("implement me")
+}
+
+func (logger SpyLogger) IsWarn() bool {
+	panic("implement me")
+}
+
+func (logger SpyLogger) IsError() bool {
+	panic("implement me")
+}
+
+func (logger SpyLogger) ImpliedArgs() []interface{} {
+	panic("implement me")
+}
+
+func (logger SpyLogger) With(args ...interface{}) hclog.Logger {
+	panic("implement me")
+}
+
+func (logger SpyLogger) Name() string {
+	return "spy logger"
+}
+
+func (logger SpyLogger) Named(name string) hclog.Logger {
+	panic("implement me")
+}
+
+func (logger SpyLogger) ResetNamed(name string) hclog.Logger {
+	panic("implement me")
+}
+
+func (logger SpyLogger) SetLevel(level hclog.Level) {
+	panic("implement me")
+}
+
+func (logger SpyLogger) StandardLogger(opts *hclog.StandardLoggerOptions) *log.Logger {
+	panic("implement me")
+}
+
+func (logger SpyLogger) StandardWriter(opts *hclog.StandardLoggerOptions) io.Writer {
+	panic("implement me")
+}
--- a/storage/clickhousespanstore/mocks/spylogger_test.go
+++ b/storage/clickhousespanstore/mocks/spylogger_test.go
@ -0,0 +1,170 @@
+package mocks
+
+import (
+	"math/rand"
+	"strconv"
+	"testing"
+
+	hclog "github.com/hashicorp/go-hclog"
+	"github.com/stretchr/testify/assert"
+)
+
+const (
+	maxLogCount = 80
+	maxArgCount = 10
+)
+
+func TestSpyLogger_AssertLogsEmpty(t *testing.T) {
+	logger := NewSpyLogger()
+	logger.AssertLogsEmpty(t)
+}
+
+func TestSpyLogger_AssertLogsOfLevelEqualNoArgs(t *testing.T) {
+	logger := NewSpyLogger()
+	var logs = make([][]LogMock, levelCount)
+	for level, levelLogs := range logs {
+		logsCount := rand.Intn(maxLogCount)
+		for i := 0; i < logsCount; i++ {
+			msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+			levelLogs = append(levelLogs, LogMock{Msg: msg})
+			logger.Log(hclog.Level(level+1), msg)
+		}
+		logs[level] = levelLogs
+	}
+
+	for level, levelLogs := range logs {
+		logger.AssertLogsOfLevelEqual(t, hclog.Level(level+1), levelLogs)
+	}
+}
+
+func TestSpyLogger_AssertLogsOfLevelEqualArgs(t *testing.T) {
+	logger := NewSpyLogger()
+	var logs = make([][]LogMock, levelCount)
+	for level, levelLogs := range logs {
+		logsCount := rand.Intn(maxLogCount)
+		for i := 0; i < logsCount; i++ {
+			msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+			args := generateArgs(rand.Intn(maxArgCount))
+			levelLogs = append(levelLogs, LogMock{Msg: msg, Args: args})
+			logger.Log(hclog.Level(level+1), msg, args...)
+		}
+		logs[level] = levelLogs
+	}
+
+	for level, levelLogs := range logs {
+		logger.AssertLogsOfLevelEqual(t, hclog.Level(level+1), levelLogs)
+	}
+}
+
+func TestSpyLogger_Trace(t *testing.T) {
+	logger := NewSpyLogger()
+	logsCount := rand.Intn(maxLogCount)
+	logs := make([]LogMock, 0, logsCount)
+	for i := 0; i < logsCount; i++ {
+		msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+		args := generateArgs(rand.Intn(maxArgCount))
+		logs = append(logs, LogMock{Msg: msg, Args: args})
+		logger.Trace(msg, args...)
+	}
+
+	logger.AssertLogsOfLevelEqual(t, hclog.Trace, logs)
+}
+
+func TestSpyLogger_Debug(t *testing.T) {
+	logger := NewSpyLogger()
+	logsCount := rand.Intn(maxLogCount)
+	logs := make([]LogMock, 0, logsCount)
+	for i := 0; i < logsCount; i++ {
+		msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+		args := generateArgs(rand.Intn(maxArgCount))
+		logs = append(logs, LogMock{Msg: msg, Args: args})
+		logger.Debug(msg, args...)
+	}
+
+	logger.AssertLogsOfLevelEqual(t, hclog.Debug, logs)
+}
+
+func TestSpyLogger_Info(t *testing.T) {
+	logger := NewSpyLogger()
+	logsCount := rand.Intn(maxLogCount)
+	logs := make([]LogMock, 0, logsCount)
+	for i := 0; i < logsCount; i++ {
+		msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+		args := generateArgs(rand.Intn(maxArgCount))
+		logs = append(logs, LogMock{Msg: msg, Args: args})
+		logger.Info(msg, args...)
+	}
+
+	logger.AssertLogsOfLevelEqual(t, hclog.Info, logs)
+}
+
+func TestSpyLogger_Warn(t *testing.T) {
+	logger := NewSpyLogger()
+	logsCount := rand.Intn(maxLogCount)
+	logs := make([]LogMock, 0, logsCount)
+	for i := 0; i < logsCount; i++ {
+		msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+		args := generateArgs(rand.Intn(maxArgCount))
+		logs = append(logs, LogMock{Msg: msg, Args: args})
+		logger.Warn(msg, args...)
+	}
+
+	logger.AssertLogsOfLevelEqual(t, hclog.Warn, logs)
+}
+
+func TestSpyLogger_Error(t *testing.T) {
+	logger := NewSpyLogger()
+	logsCount := rand.Intn(maxLogCount)
+	logs := make([]LogMock, 0, logsCount)
+	for i := 0; i < logsCount; i++ {
+		msg := "msg" + strconv.FormatUint(rand.Uint64(), 10)
+		args := generateArgs(rand.Intn(maxArgCount))
+		logs = append(logs, LogMock{Msg: msg, Args: args})
+		logger.Error(msg, args...)
+	}
+
+	logger.AssertLogsOfLevelEqual(t, hclog.Error, logs)
+}
+
+func TestSpyLogger_Name(t *testing.T) {
+	assert.Equal(t, "spy logger", NewSpyLogger().Name())
+}
+
+func TestNotImplemented(t *testing.T) {
+	logger := NewSpyLogger()
+
+	tests := map[string]struct {
+		function assert.PanicTestFunc
+	}{
+		"is_trace":        {function: func() { _ = logger.IsTrace() }},
+		"is_debug":        {function: func() { _ = logger.IsDebug() }},
+		"is_info":         {function: func() { _ = logger.IsInfo() }},
+		"is_warn":         {function: func() { _ = logger.IsWarn() }},
+		"is_error":        {function: func() { _ = logger.IsError() }},
+		"implied_args":    {function: func() { _ = logger.ImpliedArgs() }},
+		"with":            {function: func() { _ = logger.With() }},
+		"named":           {function: func() { _ = logger.Named("") }},
+		"reset_named":     {function: func() { _ = logger.ResetNamed("") }},
+		"set_level":       {function: func() { logger.SetLevel(hclog.NoLevel) }},
+		"standard_logger": {function: func() { _ = logger.StandardLogger(nil) }},
+		"standard_writer": {function: func() { _ = logger.StandardWriter(nil) }},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			assert.Panics(t, test.function, "implement me")
+		})
+	}
+}
+
+func generateArgs(count int) []interface{} {
+	args := make([]interface{}, 0, 2*count)
+	for j := 0; j < count; j++ {
+		args = append(
+			args,
+			"key"+strconv.FormatUint(rand.Uint64(), 10),
+			"value"+strconv.FormatUint(rand.Uint64(), 10),
+		)
+	}
+	return args
+}
--- a/storage/clickhousespanstore/params.go
+++ b/storage/clickhousespanstore/params.go
@ -0,0 +1,19 @@
+package clickhousespanstore
+
+import (
+	"database/sql"
+	"time"
+
+	hclog "github.com/hashicorp/go-hclog"
+)
+
+// WorkerParams contains parameters that are shared between WriteWorkers
+type WorkerParams struct {
+	logger     hclog.Logger
+	db         *sql.DB
+	indexTable TableName
+	spansTable TableName
+	tenant     string
+	encoding   Encoding
+	delay      time.Duration
+}
--- a/storage/clickhousespanstore/pool.go
+++ b/storage/clickhousespanstore/pool.go
@ -0,0 +1,131 @@
+package clickhousespanstore
+
+import (
+	"math"
+	"sync"
+
+	"github.com/jaegertracing/jaeger/model"
+	"github.com/prometheus/client_golang/prometheus"
+)
+
+var (
+	numDiscardedSpans = prometheus.NewCounter(prometheus.CounterOpts{
+		Name: "jaeger_clickhouse_discarded_spans",
+		Help: "Count of spans that have been discarded due to pending writes exceeding max_span_count",
+	})
+	numPendingSpans = prometheus.NewGauge(prometheus.GaugeOpts{
+		Name: "jaeger_clickhouse_pending_spans",
+		Help: "Number of spans that are currently pending, counts against max_span_count",
+	})
+)
+
+// WriteWorkerPool is a worker pool for writing batches of spans.
+// Given a new batch, WriteWorkerPool creates a new WriteWorker.
+// If the number of currently processed spans if more than maxSpanCount, then the oldest worker is removed.
+type WriteWorkerPool struct {
+	params *WorkerParams
+
+	finish  chan bool
+	done    sync.WaitGroup
+	batches chan []*model.Span
+
+	maxSpanCount int
+	mutex        sync.Mutex
+	workers      workerHeap
+	workerDone   chan *WriteWorker
+}
+
+var registerPoolMetrics sync.Once
+
+func NewWorkerPool(params *WorkerParams, maxSpanCount int) WriteWorkerPool {
+	registerPoolMetrics.Do(func() {
+		prometheus.MustRegister(numDiscardedSpans, numPendingSpans)
+	})
+
+	return WriteWorkerPool{
+		params:  params,
+		finish:  make(chan bool),
+		done:    sync.WaitGroup{},
+		batches: make(chan []*model.Span),
+
+		mutex:      sync.Mutex{},
+		workers:    newWorkerHeap(100),
+		workerDone: make(chan *WriteWorker),
+
+		maxSpanCount: maxSpanCount,
+	}
+}
+
+func (pool *WriteWorkerPool) Work() {
+	finish := false
+	nextWorkerID := int32(1)
+	pendingSpanCount := 0
+	for {
+		// Initialize to zero, or update value from previous loop
+		numPendingSpans.Set(float64(pendingSpanCount))
+
+		pool.done.Add(1)
+		select {
+		case batch := <-pool.batches:
+			batchSize := len(batch)
+			if pool.checkLimit(pendingSpanCount, batchSize) {
+				// Limit disabled or batch fits within limit, write the batch.
+				worker := WriteWorker{
+					workerID: nextWorkerID,
+
+					params: pool.params,
+					batch:  batch,
+
+					finish:     make(chan bool),
+					workerDone: pool.workerDone,
+					done:       sync.WaitGroup{},
+				}
+				if nextWorkerID == math.MaxInt32 {
+					nextWorkerID = 1
+				} else {
+					nextWorkerID++
+				}
+				pool.workers.AddWorker(&worker)
+				pendingSpanCount += batchSize
+				go worker.Work()
+			} else {
+				// Limit exceeded, complain
+				numDiscardedSpans.Add(float64(batchSize))
+				pool.params.logger.Error("Discarding batch of spans due to exceeding pending span count", "batch_size", batchSize, "pending_span_count", pendingSpanCount, "max_span_count", pool.maxSpanCount)
+			}
+		case worker := <-pool.workerDone:
+			// The worker has finished, subtract its work from the count and clean it from the heap.
+			pendingSpanCount -= len(worker.batch)
+			if err := pool.workers.RemoveWorker(worker); err != nil {
+				pool.params.logger.Error("could not remove worker", "worker", worker, "error", err)
+			}
+		case <-pool.finish:
+			pool.workers.CloseWorkers()
+			finish = true
+		}
+		pool.done.Done()
+
+		if finish {
+			break
+		}
+	}
+}
+
+func (pool *WriteWorkerPool) WriteBatch(batch []*model.Span) {
+	pool.batches <- batch
+}
+
+func (pool *WriteWorkerPool) Close() {
+	pool.finish <- true
+	pool.done.Wait()
+}
+
+// checkLimit returns whether batchSize fits within the maxSpanCount
+func (pool *WriteWorkerPool) checkLimit(pendingSpanCount int, batchSize int) bool {
+	if pool.maxSpanCount <= 0 {
+		return true
+	}
+
+	// Check limit, add batchSize if within limit
+	return pendingSpanCount+batchSize <= pool.maxSpanCount
+}
--- a/storage/clickhousespanstore/reader.go
+++ b/storage/clickhousespanstore/reader.go
@ -10,10 +10,9 @@ import (
 	"time"

 	"github.com/gogo/protobuf/proto"
-	"github.com/opentracing/opentracing-go"
-
 	"github.com/jaegertracing/jaeger/model"
 	"github.com/jaegertracing/jaeger/storage/spanstore"
+	opentracing "github.com/opentracing/opentracing-go"
 )

 const (
@ -31,20 +30,24 @@ var (
 // TraceReader for reading spans from ClickHouse
 type TraceReader struct {
 	db              *sql.DB
-	operationsTable string
-	indexTable      string
-	spansTable      string
+	operationsTable TableName
+	indexTable      TableName
+	spansTable      TableName
+	tenant          string
+	maxNumSpans     uint
 }

 var _ spanstore.Reader = (*TraceReader)(nil)

 // NewTraceReader returns a TraceReader for the database
-func NewTraceReader(db *sql.DB, operationsTable, indexTable, spansTable string) *TraceReader {
+func NewTraceReader(db *sql.DB, operationsTable, indexTable, spansTable TableName, tenant string, maxNumSpans uint) *TraceReader {
 	return &TraceReader{
 		db:              db,
 		operationsTable: operationsTable,
 		indexTable:      indexTable,
 		spansTable:      spansTable,
+		tenant:          tenant,
+		maxNumSpans:     maxNumSpans,
 	}
 }

@ -58,19 +61,29 @@ func (r *TraceReader) getTraces(ctx context.Context, traceIDs []model.TraceID) (
 	span, _ := opentracing.StartSpanFromContext(ctx, "getTraces")
 	defer span.Finish()

-	values := make([]interface{}, len(traceIDs))
+	args := make([]interface{}, len(traceIDs))
 	for i, traceID := range traceIDs {
-		values[i] = traceID.String()
+		args[i] = traceID.String()
 	}

-	// It's more efficient to do PREWHERE on traceID to then only read needed models:
+	// It's more efficient to do PREWHERE on traceID to the only read needed models:
 	// * https://clickhouse.tech/docs/en/sql-reference/statements/select/prewhere/
-	query := fmt.Sprintf("SELECT model FROM %s PREWHERE traceID IN (%s)", r.spansTable, "?"+strings.Repeat(",?", len(values)-1))
+	//nolint:gosec  , G201: SQL string formatting
+	query := fmt.Sprintf("SELECT model FROM %s PREWHERE traceID IN (%s)", r.spansTable, "?"+strings.Repeat(",?", len(traceIDs)-1))
+
+	if r.tenant != "" {
+		query += " AND tenant = ?"
+		args = append(args, r.tenant)
+	}
+
+	if r.maxNumSpans > 0 {
+		query += fmt.Sprintf(" ORDER BY timestamp LIMIT %d BY traceID", r.maxNumSpans)
+	}

 	span.SetTag("db.statement", query)
-	span.SetTag("db.args", values)
+	span.SetTag("db.args", args)

-	rows, err := r.db.QueryContext(ctx, query, values...)
+	rows, err := r.db.QueryContext(ctx, query, args...)
 	if err != nil {
 		return nil, err
 	}
@ -82,7 +95,8 @@ func (r *TraceReader) getTraces(ctx context.Context, traceIDs []model.TraceID) (
 	for rows.Next() {
 		var serialized string

-		if err := rows.Scan(&serialized); err != nil {
+		err = rows.Scan(&serialized)
+		if err != nil {
 			return nil, err
 		}

@ -143,7 +157,7 @@ func (r *TraceReader) getStrings(ctx context.Context, sql string, args ...interf

 	defer rows.Close()

-	values := []string{}
+	values := make([]string, 0)

 	for rows.Next() {
 		var value string
@ -169,11 +183,19 @@ func (r *TraceReader) GetServices(ctx context.Context) ([]string, error) {
 		return nil, errNoOperationsTable
 	}

-	query := fmt.Sprintf("SELECT service FROM %s GROUP BY service", r.operationsTable)
+	query := fmt.Sprintf("SELECT service FROM %s", r.operationsTable)
+	args := make([]interface{}, 0)

+	if r.tenant != "" {
+		query += " WHERE tenant = ?"
+		args = append(args, r.tenant)
+	}
+
+	query += " GROUP BY service"
 	span.SetTag("db.statement", query)
+	span.SetTag("db.args", args)

-	return r.getStrings(ctx, query)
+	return r.getStrings(ctx, query, args...)
 }

 // GetOperations fetches operations in the service and empty slice if service does not exists
@ -188,20 +210,44 @@ func (r *TraceReader) GetOperations(
 		return nil, errNoOperationsTable
 	}

-	query := fmt.Sprintf("SELECT operation FROM %s WHERE service = ? GROUP BY operation", r.operationsTable)
-	args := []interface{}{params.ServiceName}
+	//nolint:gosec  , G201: SQL string formatting
+	query := fmt.Sprintf("SELECT operation, spankind FROM %s WHERE", r.operationsTable)
+	args := make([]interface{}, 0)
+
+	if r.tenant != "" {
+		query += " tenant = ? AND"
+		args = append(args, r.tenant)
+	}
+
+	query += " service = ? GROUP BY operation, spankind ORDER BY operation"
+	args = append(args, params.ServiceName)

 	span.SetTag("db.statement", query)
 	span.SetTag("db.args", args)

-	names, err := r.getStrings(ctx, query, args...)
+	rows, err := r.db.QueryContext(ctx, query, args...)
 	if err != nil {
 		return nil, err
 	}

-	operations := make([]spanstore.Operation, len(names))
-	for i, name := range names {
-		operations[i].Name = name
+	defer rows.Close()
+
+	operations := make([]spanstore.Operation, 0)
+
+	for rows.Next() {
+		var name, spanKind string
+		if err := rows.Scan(&name, &spanKind); err != nil {
+			return nil, err
+		}
+		operation := spanstore.Operation{Name: name}
+		if spanKind != "" {
+			operation.SpanKind = spanKind
+		}
+		operations = append(operations, operation)
+	}
+
+	if err := rows.Err(); err != nil {
+		return nil, err
 	}

 	return operations, nil
@ -249,7 +295,7 @@ func (r *TraceReader) FindTraceIDs(ctx context.Context, params *spanstore.TraceQ
 		timeSpan = minTimespanForProgressiveSearch
 	}

-	found := []model.TraceID{}
+	found := make([]model.TraceID, 0)

 	for step := 0; step < maxProgressiveSteps; step++ {
 		if len(found) >= params.NumTraces {
@ -278,7 +324,7 @@ func (r *TraceReader) FindTraceIDs(ctx context.Context, params *spanstore.TraceQ
 		found = append(found, foundInRange...)

 		end = start
-		timeSpan = timeSpan * 2
+		timeSpan *= 2
 	}

 	return found, nil
@ -288,7 +334,7 @@ func (r *TraceReader) findTraceIDsInRange(ctx context.Context, params *spanstore
 	span, ctx := opentracing.StartSpanFromContext(ctx, "findTraceIDsInRange")
 	defer span.Finish()

-	if end.Before(start) || end.UTC() == start.UTC() {
+	if end.Before(start) || end == start {
 		return []model.TraceID{}, nil
 	}

@ -301,34 +347,36 @@ func (r *TraceReader) findTraceIDsInRange(ctx context.Context, params *spanstore
 	query := fmt.Sprintf("SELECT DISTINCT traceID FROM %s WHERE service = ?", r.indexTable)
 	args := []interface{}{params.ServiceName}

+	if r.tenant != "" {
+		query += " AND tenant = ?"
+		args = append(args, r.tenant)
+	}
+
 	if params.OperationName != "" {
-		query = query + " AND operation = ?"
+		query += " AND operation = ?"
 		args = append(args, params.OperationName)
 	}

-	query = query + " AND -toUnixTimestamp(timestamp) <= -toUnixTimestamp(?)"
-	args = append(args, start.UTC().Format("2006-01-02T15:04:05"))
-
-	query = query + " AND -toUnixTimestamp(timestamp) >= -toUnixTimestamp(?)"
-	args = append(args, end.UTC().Format("2006-01-02T15:04:05"))
+	query += " AND timestamp >= ? AND timestamp <= ?"
+	args = append(args, start, end)

 	if params.DurationMin != 0 {
-		query = query + " AND durationUs >= ?"
+		query += " AND durationUs >= ?"
 		args = append(args, params.DurationMin.Microseconds())
 	}

 	if params.DurationMax != 0 {
-		query = query + " AND durationUs <= ?"
+		query += " AND durationUs <= ?"
 		args = append(args, params.DurationMax.Microseconds())
 	}

 	for key, value := range params.Tags {
-		query = query + " AND has(tags, ?)"
-		args = append(args, fmt.Sprintf("%s=%s", key, value))
+		query += " AND has(tags.key, ?) AND has(splitByChar(',', tags.value[indexOf(tags.key, ?)]), ?)"
+		args = append(args, key, key, value)
 	}

 	if len(skip) > 0 {
-		query = query + fmt.Sprintf(" AND traceID NOT IN (%s)", "?"+strings.Repeat(",?", len(skip)-1))
+		query += fmt.Sprintf(" AND traceID NOT IN (%s)", "?"+strings.Repeat(",?", len(skip)-1))
 		for _, traceID := range skip {
 			args = append(args, traceID.String())
 		}
@ -336,7 +384,7 @@ func (r *TraceReader) findTraceIDsInRange(ctx context.Context, params *spanstore

 	// Sorting by service is required for early termination of primary key scan:
 	// * https://github.com/ClickHouse/ClickHouse/issues/7102
-	query = query + " ORDER BY service, -toUnixTimestamp(timestamp) LIMIT ?"
+	query += " ORDER BY service, timestamp DESC LIMIT ?"
 	args = append(args, params.NumTraces-len(skip))

 	span.SetTag("db.statement", query)
--- a/storage/clickhousespanstore/reader_test.go
+++ b/storage/clickhousespanstore/reader_test.go
--- a/storage/clickhousespanstore/tablename.go
+++ b/storage/clickhousespanstore/tablename.go
@ -0,0 +1,7 @@
+package clickhousespanstore
+
+type TableName string
+
+func (tableName TableName) ToLocal() TableName {
+	return tableName + "_local"
+}
--- a/storage/clickhousespanstore/tablename_test.go
+++ b/storage/clickhousespanstore/tablename_test.go
@ -0,0 +1,13 @@
+package clickhousespanstore
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestTableName_ToLocal(t *testing.T) {
+	tableName := TableName("some_table")
+	assert.Equal(t, tableName+"_local", tableName.ToLocal())
+
+}
--- a/storage/clickhousespanstore/worker.go
+++ b/storage/clickhousespanstore/worker.go
@ -0,0 +1,274 @@
+package clickhousespanstore
+
+import (
+	"encoding/json"
+	"fmt"
+	"sort"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/gogo/protobuf/proto"
+	"github.com/jaegertracing/jaeger/model"
+)
+
+var delays = []int{2, 3, 5, 8}
+
+// WriteWorker writes spans to CLickHouse.
+// Given a batch of spans, WriteWorker attempts to write them to database.
+// Interval in seconds between attempts changes due to delays slice, then it remains the same as the last value in delays.
+type WriteWorker struct {
+	// workerID is an arbitrary identifier for keeping track of this worker in logs
+	workerID   int32
+	params     *WorkerParams
+	batch      []*model.Span
+	finish     chan bool
+	workerDone chan *WriteWorker
+	done       sync.WaitGroup
+}
+
+func (worker *WriteWorker) Work() {
+	worker.done.Add(1)
+
+	defer worker.done.Done()
+
+	// TODO: look for specific error(connection refused | database error)
+	if err := worker.writeBatch(worker.batch); err != nil {
+		worker.params.logger.Error("Could not write a batch of spans", "error", err, "worker_id", worker.workerID)
+	} else {
+		worker.close()
+		return
+	}
+	attempt := 0
+	for {
+		currentDelay := worker.getCurrentDelay(&attempt, worker.params.delay)
+		timer := time.After(currentDelay)
+		select {
+		case <-worker.finish:
+			worker.close()
+			return
+		case <-timer:
+			if err := worker.writeBatch(worker.batch); err != nil {
+				worker.params.logger.Error("Could not write a batch of spans", "error", err, "worker_id", worker.workerID)
+			} else {
+				worker.close()
+				return
+			}
+		}
+	}
+}
+
+func (worker *WriteWorker) Close() {
+	worker.finish <- true
+	worker.done.Wait()
+}
+
+func (worker *WriteWorker) getCurrentDelay(attempt *int, delay time.Duration) time.Duration {
+	if *attempt < len(delays) {
+		*attempt++
+	}
+	return time.Duration(int64(delays[*attempt-1]) * delay.Nanoseconds())
+}
+
+func (worker *WriteWorker) close() {
+	worker.workerDone <- worker
+}
+
+func (worker *WriteWorker) writeBatch(batch []*model.Span) error {
+	worker.params.logger.Debug("Writing spans", "size", len(batch))
+	if err := worker.writeModelBatch(batch); err != nil {
+		return err
+	}
+
+	if worker.params.indexTable != "" {
+		if err := worker.writeIndexBatch(batch); err != nil {
+			return err
+		}
+	}
+
+	return nil
+}
+
+func (worker *WriteWorker) writeModelBatch(batch []*model.Span) error {
+	tx, err := worker.params.db.Begin()
+	if err != nil {
+		return err
+	}
+
+	committed := false
+
+	defer func() {
+		if !committed {
+			// Clickhouse does not support real rollback
+			_ = tx.Rollback()
+		}
+	}()
+
+	var query string
+	if worker.params.tenant == "" {
+		query = fmt.Sprintf("INSERT INTO %s (timestamp, traceID, model) VALUES (?, ?, ?)", worker.params.spansTable)
+	} else {
+		query = fmt.Sprintf("INSERT INTO %s (tenant, timestamp, traceID, model) VALUES (?, ?, ?, ?)", worker.params.spansTable)
+	}
+
+	statement, err := tx.Prepare(query)
+	if err != nil {
+		return err
+	}
+
+	defer statement.Close()
+
+	for _, span := range batch {
+		var serialized []byte
+
+		if worker.params.encoding == EncodingJSON {
+			serialized, err = json.Marshal(span)
+		} else {
+			serialized, err = proto.Marshal(span)
+		}
+
+		if err != nil {
+			return err
+		}
+
+		if worker.params.tenant == "" {
+			_, err = statement.Exec(span.StartTime, span.TraceID.String(), serialized)
+		} else {
+			_, err = statement.Exec(worker.params.tenant, span.StartTime, span.TraceID.String(), serialized)
+		}
+		if err != nil {
+			return err
+		}
+	}
+
+	committed = true
+
+	return tx.Commit()
+}
+
+func (worker *WriteWorker) writeIndexBatch(batch []*model.Span) error {
+	tx, err := worker.params.db.Begin()
+	if err != nil {
+		return err
+	}
+
+	committed := false
+
+	defer func() {
+		if !committed {
+			// Clickhouse does not support real rollback
+			_ = tx.Rollback()
+		}
+	}()
+
+	var query string
+	if worker.params.tenant == "" {
+		query = fmt.Sprintf(
+			"INSERT INTO %s (timestamp, traceID, service, operation, durationUs, tags.key, tags.value) VALUES (?, ?, ?, ?, ?, ?, ?)",
+			worker.params.indexTable,
+		)
+	} else {
+		query = fmt.Sprintf(
+			"INSERT INTO %s (tenant, timestamp, traceID, service, operation, durationUs, tags.key, tags.value) VALUES (?, ?, ?, ?, ?, ?, ?, ?)",
+			worker.params.indexTable,
+		)
+	}
+
+	statement, err := tx.Prepare(query)
+	if err != nil {
+		return err
+	}
+
+	defer statement.Close()
+
+	for _, span := range batch {
+		keys, values := uniqueTagsForSpan(span)
+		if worker.params.tenant == "" {
+			_, err = statement.Exec(
+				span.StartTime,
+				span.TraceID.String(),
+				span.Process.ServiceName,
+				span.OperationName,
+				uint64(span.Duration.Microseconds()),
+				keys,
+				values,
+			)
+		} else {
+			_, err = statement.Exec(
+				worker.params.tenant,
+				span.StartTime,
+				span.TraceID.String(),
+				span.Process.ServiceName,
+				span.OperationName,
+				uint64(span.Duration.Microseconds()),
+				keys,
+				values,
+			)
+		}
+		if err != nil {
+			return err
+		}
+	}
+
+	committed = true
+
+	return tx.Commit()
+}
+
+func uniqueTagsForSpan(span *model.Span) (keys, values []string) {
+	uniqueTags := make(map[string][]string, len(span.Tags)+len(span.Process.Tags))
+
+	for i := range span.Tags {
+		key := tagKey(&span.GetTags()[i])
+		uniqueTags[key] = append(uniqueTags[key], tagValue(&span.GetTags()[i]))
+	}
+
+	for i := range span.Process.Tags {
+		key := tagKey(&span.GetProcess().GetTags()[i])
+		uniqueTags[key] = append(uniqueTags[key], tagValue(&span.GetProcess().GetTags()[i]))
+	}
+
+	for _, event := range span.Logs {
+		for i := range event.Fields {
+			key := tagKey(&event.GetFields()[i])
+			uniqueTags[key] = append(uniqueTags[key], tagValue(&event.GetFields()[i]))
+		}
+	}
+
+	keys = make([]string, 0, len(uniqueTags))
+	for k := range uniqueTags {
+		keys = append(keys, k)
+	}
+	sort.Strings(keys)
+
+	values = make([]string, 0, len(uniqueTags))
+	for _, key := range keys {
+		values = append(values, strings.Join(unique(uniqueTags[key]), ","))
+	}
+
+	return keys, values
+}
+
+func tagKey(kv *model.KeyValue) string {
+	return kv.Key
+}
+
+func tagValue(kv *model.KeyValue) string {
+	return kv.AsString()
+}
+
+func unique(slice []string) []string {
+	if len(slice) == 1 {
+		return slice
+	}
+
+	keys := make(map[string]bool)
+	list := []string{}
+	for _, entry := range slice {
+		if _, value := keys[entry]; !value {
+			keys[entry] = true
+			list = append(list, entry)
+		}
+	}
+	return list
+}
--- a/storage/clickhousespanstore/worker_test.go
+++ b/storage/clickhousespanstore/worker_test.go
@ -0,0 +1,552 @@
+package clickhousespanstore
+
+import (
+	"database/sql"
+	"database/sql/driver"
+	"encoding/json"
+	"fmt"
+	"math/rand"
+	"strconv"
+	"testing"
+	"time"
+
+	sqlmock "github.com/DATA-DOG/go-sqlmock"
+	"github.com/gogo/protobuf/proto"
+	hclog "github.com/hashicorp/go-hclog"
+	"github.com/jaegertracing/jaeger/model"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore/mocks"
+)
+
+const (
+	testTagCount      = 10
+	testLogCount      = 5
+	testLogFieldCount = 5
+	testIndexTable    = "test_index_table"
+	testSpansTable    = "test_spans_table"
+	testTenant        = "test_tenant"
+)
+
+type expectation struct {
+	preparation string
+	execArgs    [][]driver.Value
+}
+
+var (
+	errorMock = fmt.Errorf("error mock")
+	process   = model.NewProcess("test_service", []model.KeyValue{model.String("test_process_key", "test_process_value")})
+	testSpan  = model.Span{
+		TraceID:       model.NewTraceID(1, 2),
+		SpanID:        model.NewSpanID(3),
+		OperationName: "GET /unit_test",
+		StartTime:     testStartTime,
+		Process:       process,
+		Tags:          []model.KeyValue{model.String("test_string_key", "test_string_value"), model.Int64("test_int64_key", 4)},
+		Logs:          []model.Log{{Timestamp: testStartTime, Fields: []model.KeyValue{model.String("test_log_key", "test_log_value")}}},
+		Duration:      time.Minute,
+	}
+	testSpans             = []*model.Span{&testSpan}
+	keys, values          = uniqueTagsForSpan(&testSpan)
+	indexWriteExpectation = expectation{
+		preparation: fmt.Sprintf("INSERT INTO %s (timestamp, traceID, service, operation, durationUs, tags.key, tags.value) VALUES (?, ?, ?, ?, ?, ?, ?)", testIndexTable),
+		execArgs: [][]driver.Value{{
+			testSpan.StartTime,
+			testSpan.TraceID.String(),
+			testSpan.Process.GetServiceName(),
+			testSpan.OperationName,
+			uint64(testSpan.Duration.Microseconds()),
+			keys,
+			values,
+		}}}
+	indexWriteExpectationTenant = expectation{
+		preparation: fmt.Sprintf("INSERT INTO %s (tenant, timestamp, traceID, service, operation, durationUs, tags.key, tags.value) VALUES (?, ?, ?, ?, ?, ?, ?, ?)", testIndexTable),
+		execArgs: [][]driver.Value{{
+			testTenant,
+			testSpan.StartTime,
+			testSpan.TraceID.String(),
+			testSpan.Process.GetServiceName(),
+			testSpan.OperationName,
+			uint64(testSpan.Duration.Microseconds()),
+			keys,
+			values,
+		}}}
+	writeBatchLogs = []mocks.LogMock{{Msg: "Writing spans", Args: []interface{}{"size", len(testSpans)}}}
+)
+
+func TestSpanWriter_TagKeyValue(t *testing.T) {
+	tests := map[string]struct {
+		kv       model.KeyValue
+		expected string
+	}{
+		"string value":       {kv: model.String("tag_key", "tag_string_value"), expected: "tag_string_value"},
+		"true value":         {kv: model.Bool("tag_key", true), expected: "true"},
+		"false value":        {kv: model.Bool("tag_key", false), expected: "false"},
+		"positive int value": {kv: model.Int64("tag_key", 1203912), expected: "1203912"},
+		"negative int value": {kv: model.Int64("tag_key", -1203912), expected: "-1203912"},
+		"float value":        {kv: model.Float64("tag_key", 0.005009), expected: "0.005009"},
+	}
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			assert.Equal(t, test.expected, tagValue(&test.kv), "Incorrect tag value string")
+		})
+	}
+}
+
+func TestSpanWriter_UniqueTagsForSpan(t *testing.T) {
+	tests := map[string]struct {
+		tags           []model.KeyValue
+		processTags    []model.KeyValue
+		logs           []model.Log
+		expectedKeys   []string
+		expectedValues []string
+	}{
+		"default": {
+			tags:           []model.KeyValue{model.String("key2", "value")},
+			processTags:    []model.KeyValue{model.Int64("key3", 412)},
+			logs:           []model.Log{{Fields: []model.KeyValue{model.Float64("key1", .5)}}},
+			expectedKeys:   []string{"key1", "key2", "key3"},
+			expectedValues: []string{"0.5", "value", "412"},
+		},
+		"repeating tags": {
+			tags:           []model.KeyValue{model.String("key2", "value"), model.String("key2", "value")},
+			processTags:    []model.KeyValue{model.Int64("key3", 412)},
+			logs:           []model.Log{{Fields: []model.KeyValue{model.Float64("key1", .5)}}},
+			expectedKeys:   []string{"key1", "key2", "key3"},
+			expectedValues: []string{"0.5", "value", "412"},
+		},
+		"repeating keys": {
+			tags:           []model.KeyValue{model.String("key2", "value_a"), model.String("key2", "value_b")},
+			processTags:    []model.KeyValue{model.Int64("key3", 412)},
+			logs:           []model.Log{{Fields: []model.KeyValue{model.Float64("key1", .5)}}},
+			expectedKeys:   []string{"key1", "key2", "key3"},
+			expectedValues: []string{"0.5", "value_a,value_b", "412"},
+		},
+		"repeating values": {
+			tags:           []model.KeyValue{model.String("key2", "value"), model.Int64("key4", 412)},
+			processTags:    []model.KeyValue{model.Int64("key3", 412)},
+			logs:           []model.Log{{Fields: []model.KeyValue{model.Float64("key1", .5)}}},
+			expectedKeys:   []string{"key1", "key2", "key3", "key4"},
+			expectedValues: []string{"0.5", "value", "412", "412"},
+		},
+	}
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			process := model.Process{Tags: test.processTags}
+			span := model.Span{Tags: test.tags, Process: &process, Logs: test.logs}
+			actualKeys, actualValues := uniqueTagsForSpan(&span)
+			assert.Equal(t, test.expectedKeys, actualKeys)
+			assert.Equal(t, test.expectedValues, actualValues)
+		})
+	}
+}
+
+func TestSpanWriter_General(t *testing.T) {
+	spanJSON, err := json.Marshal(&testSpan)
+	require.NoError(t, err)
+	modelWriteExpectationJSON := getModelWriteExpectation(spanJSON, "")
+	modelWriteExpectationJSONTenant := getModelWriteExpectation(spanJSON, testTenant)
+	spanProto, err := proto.Marshal(&testSpan)
+	require.NoError(t, err)
+	modelWriteExpectationProto := getModelWriteExpectation(spanProto, "")
+	modelWriteExpectationProtoTenant := getModelWriteExpectation(spanProto, testTenant)
+	tests := map[string]struct {
+		encoding     Encoding
+		indexTable   TableName
+		tenant       string
+		spans        []*model.Span
+		expectations []expectation
+		action       func(writeWorker *WriteWorker, spans []*model.Span) error
+		expectedLogs []mocks.LogMock
+	}{
+		"write index batch": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			spans:        testSpans,
+			expectations: []expectation{indexWriteExpectation},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeIndexBatch(spans) },
+		},
+		"write index tenant batch": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			spans:        testSpans,
+			expectations: []expectation{indexWriteExpectationTenant},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeIndexBatch(spans) },
+		},
+		"write model batch JSON": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationJSON},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeModelBatch(spans) },
+		},
+		"write model tenant batch JSON": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationJSONTenant},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeModelBatch(spans) },
+		},
+		"write model batch Proto": {
+			encoding:     EncodingProto,
+			indexTable:   testIndexTable,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationProto},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeModelBatch(spans) },
+		},
+		"write model tenant batch Proto": {
+			encoding:     EncodingProto,
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationProtoTenant},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeModelBatch(spans) },
+		},
+		"write batch no index JSON": {
+			encoding:     EncodingJSON,
+			indexTable:   "",
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationJSON},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch no index Proto": {
+			encoding:     EncodingProto,
+			indexTable:   "",
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationProto},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch JSON": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationJSON, indexWriteExpectation},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch tenant JSON": {
+			encoding:     EncodingJSON,
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationJSONTenant, indexWriteExpectationTenant},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch Proto": {
+			encoding:     EncodingProto,
+			indexTable:   testIndexTable,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationProto, indexWriteExpectation},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch tenant Proto": {
+			encoding:     EncodingProto,
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			spans:        testSpans,
+			expectations: []expectation{modelWriteExpectationProtoTenant, indexWriteExpectationTenant},
+			action:       func(writeWorker *WriteWorker, spans []*model.Span) error { return writeWorker.writeBatch(spans) },
+			expectedLogs: writeBatchLogs,
+		},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			db, mock, err := mocks.GetDbMock()
+			require.NoError(t, err, "an error was not expected when opening a stub database connection")
+			defer db.Close()
+
+			spyLogger := mocks.NewSpyLogger()
+			worker := getWriteWorker(spyLogger, db, test.encoding, test.indexTable, test.tenant)
+
+			for _, expectation := range test.expectations {
+				mock.ExpectBegin()
+				prep := mock.ExpectPrepare(expectation.preparation)
+				for _, args := range expectation.execArgs {
+					prep.ExpectExec().WithArgs(args...).WillReturnResult(sqlmock.NewResult(1, 1))
+				}
+				mock.ExpectCommit()
+			}
+
+			assert.NoError(t, test.action(&worker, test.spans))
+			assert.NoError(t, mock.ExpectationsWereMet())
+			spyLogger.AssertLogsOfLevelEqual(t, hclog.Debug, test.expectedLogs)
+		})
+	}
+}
+
+func TestSpanWriter_BeginError(t *testing.T) {
+	tests := map[string]struct {
+		action       func(writeWorker *WriteWorker) error
+		expectedLogs []mocks.LogMock
+	}{
+		"write model batch": {action: func(writeWorker *WriteWorker) error { return writeWorker.writeModelBatch(testSpans) }},
+		"write index batch": {action: func(writeWorker *WriteWorker) error { return writeWorker.writeIndexBatch(testSpans) }},
+		"write batch": {
+			action:       func(writeWorker *WriteWorker) error { return writeWorker.writeBatch(testSpans) },
+			expectedLogs: writeBatchLogs,
+		},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			db, mock, err := mocks.GetDbMock()
+			require.NoError(t, err, "an error was not expected when opening a stub database connection")
+			defer db.Close()
+
+			spyLogger := mocks.NewSpyLogger()
+			writeWorker := getWriteWorker(spyLogger, db, EncodingJSON, testIndexTable, "")
+
+			mock.ExpectBegin().WillReturnError(errorMock)
+
+			assert.ErrorIs(t, test.action(&writeWorker), errorMock)
+			assert.NoError(t, mock.ExpectationsWereMet())
+			spyLogger.AssertLogsOfLevelEqual(t, hclog.Debug, test.expectedLogs)
+		})
+	}
+}
+
+func TestSpanWriter_PrepareError(t *testing.T) {
+	spanJSON, err := json.Marshal(&testSpan)
+	require.NoError(t, err)
+	modelWriteExpectation := getModelWriteExpectation(spanJSON, "")
+	modelWriteExpectationTenant := getModelWriteExpectation(spanJSON, testTenant)
+
+	tests := map[string]struct {
+		action       func(writeWorker *WriteWorker) error
+		tenant       string
+		expectation  expectation
+		expectedLogs []mocks.LogMock
+	}{
+		"write model batch": {
+			action:      func(writeWorker *WriteWorker) error { return writeWorker.writeModelBatch(testSpans) },
+			expectation: modelWriteExpectation,
+		},
+		"write model tenant batch": {
+			action:      func(writeWorker *WriteWorker) error { return writeWorker.writeModelBatch(testSpans) },
+			tenant:      testTenant,
+			expectation: modelWriteExpectationTenant,
+		},
+		"write index batch": {
+			action:      func(writeWorker *WriteWorker) error { return writeWorker.writeIndexBatch(testSpans) },
+			expectation: indexWriteExpectation,
+		},
+		"write index tenant batch": {
+			action:      func(writeWorker *WriteWorker) error { return writeWorker.writeIndexBatch(testSpans) },
+			tenant:      testTenant,
+			expectation: indexWriteExpectationTenant,
+		},
+		"write batch": {
+			action:       func(writeWorker *WriteWorker) error { return writeWorker.writeBatch(testSpans) },
+			expectation:  modelWriteExpectation,
+			expectedLogs: writeBatchLogs,
+		},
+		"write tenant batch": {
+			action:       func(writeWorker *WriteWorker) error { return writeWorker.writeBatch(testSpans) },
+			tenant:       testTenant,
+			expectation:  modelWriteExpectationTenant,
+			expectedLogs: writeBatchLogs,
+		},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			db, mock, err := mocks.GetDbMock()
+			require.NoError(t, err, "an error was not expected when opening a stub database connection")
+			defer db.Close()
+
+			spyLogger := mocks.NewSpyLogger()
+			spanWriter := getWriteWorker(spyLogger, db, EncodingJSON, testIndexTable, test.tenant)
+
+			mock.ExpectBegin()
+			mock.ExpectPrepare(test.expectation.preparation).WillReturnError(errorMock)
+			mock.ExpectRollback()
+
+			assert.ErrorIs(t, test.action(&spanWriter), errorMock)
+			assert.NoError(t, mock.ExpectationsWereMet())
+			spyLogger.AssertLogsOfLevelEqual(t, hclog.Debug, test.expectedLogs)
+		})
+	}
+}
+
+func TestSpanWriter_ExecError(t *testing.T) {
+	spanJSON, err := json.Marshal(&testSpan)
+	require.NoError(t, err)
+	modelWriteExpectation := getModelWriteExpectation(spanJSON, "")
+	modelWriteExpectationTenant := getModelWriteExpectation(spanJSON, testTenant)
+	tests := map[string]struct {
+		indexTable   TableName
+		tenant       string
+		expectations []expectation
+		action       func(writer *WriteWorker) error
+		expectedLogs []mocks.LogMock
+	}{
+		"write model batch": {
+			indexTable:   testIndexTable,
+			expectations: []expectation{modelWriteExpectation},
+			action:       func(writer *WriteWorker) error { return writer.writeModelBatch(testSpans) },
+		},
+		"write model tenant batch": {
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			expectations: []expectation{modelWriteExpectationTenant},
+			action:       func(writer *WriteWorker) error { return writer.writeModelBatch(testSpans) },
+		},
+		"write index batch": {
+			indexTable:   testIndexTable,
+			expectations: []expectation{indexWriteExpectation},
+			action:       func(writer *WriteWorker) error { return writer.writeIndexBatch(testSpans) },
+		},
+		"write index tenant batch": {
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			expectations: []expectation{indexWriteExpectationTenant},
+			action:       func(writer *WriteWorker) error { return writer.writeIndexBatch(testSpans) },
+		},
+		"write batch no index": {
+			indexTable:   "",
+			expectations: []expectation{modelWriteExpectation},
+			action:       func(writer *WriteWorker) error { return writer.writeBatch(testSpans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write batch": {
+			indexTable:   testIndexTable,
+			expectations: []expectation{modelWriteExpectation, indexWriteExpectation},
+			action:       func(writer *WriteWorker) error { return writer.writeBatch(testSpans) },
+			expectedLogs: writeBatchLogs,
+		},
+		"write tenant batch": {
+			indexTable:   testIndexTable,
+			tenant:       testTenant,
+			expectations: []expectation{modelWriteExpectationTenant, indexWriteExpectationTenant},
+			action:       func(writer *WriteWorker) error { return writer.writeBatch(testSpans) },
+			expectedLogs: writeBatchLogs,
+		},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			db, mock, err := mocks.GetDbMock()
+			require.NoError(t, err, "an error was not expected when opening a stub database connection")
+			defer db.Close()
+
+			spyLogger := mocks.NewSpyLogger()
+			writeWorker := getWriteWorker(spyLogger, db, EncodingJSON, testIndexTable, test.tenant)
+
+			for i, expectation := range test.expectations {
+				mock.ExpectBegin()
+				prep := mock.ExpectPrepare(expectation.preparation)
+				if i < len(test.expectations)-1 {
+					for _, args := range expectation.execArgs {
+						prep.ExpectExec().WithArgs(args...).WillReturnResult(sqlmock.NewResult(1, 1))
+					}
+					mock.ExpectCommit()
+				} else {
+					prep.ExpectExec().WithArgs(expectation.execArgs[0]...).WillReturnError(errorMock)
+					mock.ExpectRollback()
+				}
+			}
+
+			assert.ErrorIs(t, test.action(&writeWorker), errorMock)
+			assert.NoError(t, mock.ExpectationsWereMet())
+			spyLogger.AssertLogsOfLevelEqual(t, hclog.Debug, test.expectedLogs)
+		})
+	}
+}
+
+func getWriteWorker(spyLogger mocks.SpyLogger, db *sql.DB, encoding Encoding, indexTable TableName, tenant string) WriteWorker {
+	return WriteWorker{
+		params: &WorkerParams{
+			logger:     spyLogger,
+			db:         db,
+			spansTable: testSpansTable,
+			indexTable: indexTable,
+			tenant:     tenant,
+			encoding:   encoding,
+		},
+		workerDone: make(chan *WriteWorker),
+	}
+}
+
+func generateRandomSpans(count int) []*model.Span {
+	spans := make([]*model.Span, count)
+	for i := 0; i < count; i++ {
+		span := generateRandomSpan()
+		spans[i] = &span
+	}
+	return spans
+}
+
+func generateRandomSpan() model.Span {
+	processTags := generateRandomKeyValues(testTagCount)
+	process := model.Process{
+		ServiceName: "service" + strconv.FormatUint(rand.Uint64(), 10),
+		Tags:        processTags,
+	}
+	span := model.Span{
+		TraceID:       model.NewTraceID(rand.Uint64(), rand.Uint64()),
+		SpanID:        model.NewSpanID(rand.Uint64()),
+		OperationName: "operation" + strconv.FormatUint(rand.Uint64(), 10),
+		StartTime:     getRandomTime(),
+		Process:       &process,
+		Tags:          generateRandomKeyValues(testTagCount),
+		Logs:          generateRandomLogs(),
+		Duration:      time.Unix(rand.Int63n(1<<32), 0).Sub(time.Unix(0, 0)),
+	}
+	return span
+}
+
+func generateRandomLogs() []model.Log {
+	logs := make([]model.Log, 0, testLogCount)
+	for i := 0; i < testLogCount; i++ {
+		timestamp := getRandomTime()
+		logs = append(logs, model.Log{Timestamp: timestamp, Fields: generateRandomKeyValues(testLogFieldCount)})
+	}
+	return logs
+}
+
+func getRandomTime() time.Time {
+	return time.Unix(rand.Int63n(time.Now().Unix()), 0)
+}
+
+func generateRandomKeyValues(count int) []model.KeyValue {
+	tags := make([]model.KeyValue, 0, count)
+	for i := 0; i < count; i++ {
+		key := "key" + strconv.FormatUint(rand.Uint64(), 16)
+		value := "key" + strconv.FormatUint(rand.Uint64(), 16)
+		kv := model.KeyValue{Key: key, VType: model.ValueType_STRING, VStr: value}
+		tags = append(tags, kv)
+	}
+
+	return tags
+}
+
+func getModelWriteExpectation(spanJSON []byte, tenant string) expectation {
+	if tenant == "" {
+		return expectation{
+			preparation: fmt.Sprintf("INSERT INTO %s (timestamp, traceID, model) VALUES (?, ?, ?)", testSpansTable),
+			execArgs: [][]driver.Value{{
+				testSpan.StartTime,
+				testSpan.TraceID.String(),
+				spanJSON,
+			}},
+		}
+	} else {
+		return expectation{
+			preparation: fmt.Sprintf("INSERT INTO %s (tenant, timestamp, traceID, model) VALUES (?, ?, ?, ?)", testSpansTable),
+			execArgs: [][]driver.Value{{
+				tenant,
+				testSpan.StartTime,
+				testSpan.TraceID.String(),
+				spanJSON,
+			}},
+		}
+	}
+}
--- a/storage/clickhousespanstore/writer.go
+++ b/storage/clickhousespanstore/writer.go
@ -3,20 +3,13 @@ package clickhousespanstore
 import (
 	"context"
 	"database/sql"
-	"encoding/json"
-	"fmt"
-	"sort"
-	"strings"
 	"sync"
 	"time"

-	"github.com/hashicorp/go-hclog"
-	"github.com/jaegertracing/jaeger/storage/spanstore"
-
-	"github.com/gogo/protobuf/proto"
-	"go.uber.org/zap"
-
+	hclog "github.com/hashicorp/go-hclog"
 	"github.com/jaegertracing/jaeger/model"
+	"github.com/jaegertracing/jaeger/storage/spanstore"
+	"github.com/prometheus/client_golang/prometheus"
 )

 type Encoding string
@ -28,45 +21,76 @@ const (
 	EncodingProto Encoding = "protobuf"
 )

+var (
+	numWritesWithBatchSize = prometheus.NewCounter(prometheus.CounterOpts{
+		Name: "jaeger_clickhouse_writes_with_batch_size_total",
+		Help: "Number of clickhouse writes due to batch size criteria",
+	})
+	numWritesWithFlushInterval = prometheus.NewCounter(prometheus.CounterOpts{
+		Name: "jaeger_clickhouse_writes_with_flush_interval_total",
+		Help: "Number of clickhouse writes due to flush interval criteria",
+	})
+)
+
 // SpanWriter for writing spans to ClickHouse
 type SpanWriter struct {
-	logger     hclog.Logger
-	db         *sql.DB
-	indexTable string
-	spansTable string
-	encoding   Encoding
-	delay      time.Duration
-	size       int
-	spans      chan *model.Span
-	finish     chan bool
-	done       sync.WaitGroup
+	workerParams WorkerParams
+
+	size   int64
+	spans  chan *model.Span
+	finish chan bool
+	done   sync.WaitGroup
 }

+var registerWriterMetrics sync.Once
 var _ spanstore.Writer = (*SpanWriter)(nil)

 // NewSpanWriter returns a SpanWriter for the database
-func NewSpanWriter(logger hclog.Logger, db *sql.DB, indexTable string, spansTable string, encoding Encoding, delay time.Duration, size int) *SpanWriter {
+func NewSpanWriter(
+	logger hclog.Logger,
+	db *sql.DB,
+	indexTable,
+	spansTable TableName,
+	tenant string,
+	encoding Encoding,
+	delay time.Duration,
+	size int64,
+	maxSpanCount int,
+) *SpanWriter {
 	writer := &SpanWriter{
-		logger:     logger,
-		db:         db,
-		indexTable: indexTable,
-		spansTable: spansTable,
-		encoding:   encoding,
-		delay:      delay,
-		size:       size,
-		spans:      make(chan *model.Span, size),
-		finish:     make(chan bool),
+		workerParams: WorkerParams{
+			logger:     logger,
+			db:         db,
+			indexTable: indexTable,
+			spansTable: spansTable,
+			tenant:     tenant,
+			encoding:   encoding,
+			delay:      delay,
+		},
+		size:   size,
+		spans:  make(chan *model.Span, size),
+		finish: make(chan bool),
 	}

-	go writer.backgroundWriter()
+	writer.registerMetrics()
+	go writer.backgroundWriter(maxSpanCount)

 	return writer
 }

-func (w *SpanWriter) backgroundWriter() {
+func (w *SpanWriter) registerMetrics() {
+	registerWriterMetrics.Do(func() {
+		prometheus.MustRegister(numWritesWithBatchSize)
+		prometheus.MustRegister(numWritesWithFlushInterval)
+	})
+}
+
+func (w *SpanWriter) backgroundWriter(maxSpanCount int) {
+	pool := NewWorkerPool(&w.workerParams, maxSpanCount)
+	go pool.Work()
 	batch := make([]*model.Span, 0, w.size)

-	timer := time.After(w.delay)
+	timer := time.After(w.workerParams.delay)
 	last := time.Now()

 	for {
@ -79,23 +103,33 @@ func (w *SpanWriter) backgroundWriter() {
 		case span := <-w.spans:
 			batch = append(batch, span)
 			flush = len(batch) == cap(batch)
+			if flush {
+				w.workerParams.logger.Debug("Flush due to batch size", "size", len(batch))
+				numWritesWithBatchSize.Inc()
+			}
 		case <-timer:
-			timer = time.After(w.delay)
-			flush = time.Since(last) > w.delay && len(batch) > 0
+			timer = time.After(w.workerParams.delay)
+			flush = time.Since(last) > w.workerParams.delay && len(batch) > 0
+			if flush {
+				w.workerParams.logger.Debug("Flush due to timer")
+				numWritesWithFlushInterval.Inc()
+			}
 		case <-w.finish:
 			finish = true
 			flush = len(batch) > 0
+			w.workerParams.logger.Debug("Finish channel")
 		}

 		if flush {
-			if err := w.writeBatch(batch); err != nil {
-				w.logger.Error("Could not write a batch of spans", zap.Error(err))
-			}
+			pool.WriteBatch(batch)

 			batch = make([]*model.Span, 0, w.size)
 			last = time.Now()
 		}

+		if finish {
+			pool.Close()
+		}
 		w.done.Done()

 		if finish {
@ -104,107 +138,6 @@ func (w *SpanWriter) backgroundWriter() {
 	}
 }

-func (w *SpanWriter) writeBatch(batch []*model.Span) error {
-	if err := w.writeModelBatch(batch); err != nil {
-		return err
-	}
-
-	if w.indexTable != "" {
-		if err := w.writeIndexBatch(batch); err != nil {
-			return err
-		}
-	}
-
-	return nil
-}
-
-func (w *SpanWriter) writeModelBatch(batch []*model.Span) error {
-	tx, err := w.db.Begin()
-	if err != nil {
-		return err
-	}
-
-	commited := false
-
-	defer func() {
-		if !commited {
-			// Clickhouse does not support real rollback
-			_ = tx.Rollback()
-		}
-	}()
-
-	statement, err := tx.Prepare(fmt.Sprintf("INSERT INTO %s (timestamp, traceID, model) VALUES (?, ?, ?)", w.spansTable))
-	if err != nil {
-		return nil
-	}
-
-	defer statement.Close()
-
-	for _, span := range batch {
-		var serialized []byte
-
-		if w.encoding == EncodingJSON {
-			serialized, err = json.Marshal(span)
-		} else {
-			serialized, err = proto.Marshal(span)
-		}
-
-		if err != nil {
-			return err
-		}
-
-		_, err = statement.Exec(span.StartTime, span.TraceID.String(), serialized)
-		if err != nil {
-			return err
-		}
-	}
-
-	commited = true
-
-	return tx.Commit()
-}
-
-func (w *SpanWriter) writeIndexBatch(batch []*model.Span) error {
-	tx, err := w.db.Begin()
-	if err != nil {
-		return err
-	}
-
-	commited := false
-
-	defer func() {
-		if !commited {
-			// Clickhouse does not support real rollback
-			_ = tx.Rollback()
-		}
-	}()
-
-	statement, err := tx.Prepare(fmt.Sprintf("INSERT INTO %s (timestamp, traceID, service, operation, durationUs, tags) VALUES (?, ?, ?, ?, ?, ?)", w.indexTable))
-	if err != nil {
-		return err
-	}
-
-	defer statement.Close()
-
-	for _, span := range batch {
-		_, err = statement.Exec(
-			span.StartTime,
-			span.TraceID.String(),
-			span.Process.ServiceName,
-			span.OperationName,
-			span.Duration.Microseconds(),
-			uniqueTagsForSpan(span),
-		)
-		if err != nil {
-			return err
-		}
-	}
-
-	commited = true
-
-	return tx.Commit()
-}
-
 // WriteSpan writes the encoded span
 func (w *SpanWriter) WriteSpan(_ context.Context, span *model.Span) error {
 	w.spans <- span
@ -217,43 +150,3 @@ func (w *SpanWriter) Close() error {
 	w.done.Wait()
 	return nil
 }
-
-func uniqueTagsForSpan(span *model.Span) []string {
-	uniqueTags := make(map[string]struct{}, len(span.Tags)+len(span.Process.Tags))
-
-	buf := &strings.Builder{}
-
-	for _, kv := range span.Tags {
-		uniqueTags[tagString(buf, &kv)] = struct{}{}
-	}
-
-	for _, kv := range span.Process.Tags {
-		uniqueTags[tagString(buf, &kv)] = struct{}{}
-	}
-
-	for _, event := range span.Logs {
-		for _, kv := range event.Fields {
-			uniqueTags[tagString(buf, &kv)] = struct{}{}
-		}
-	}
-
-	tags := make([]string, 0, len(uniqueTags))
-
-	for kv := range uniqueTags {
-		tags = append(tags, kv)
-	}
-
-	sort.Strings(tags)
-
-	return tags
-}
-
-func tagString(buf *strings.Builder, kv *model.KeyValue) string {
-	buf.Reset()
-
-	buf.WriteString(kv.Key)
-	buf.WriteByte('=')
-	buf.WriteString(kv.AsString())
-
-	return buf.String()
-}
--- a/storage/config.go
+++ b/storage/config.go
@ -1,16 +1,151 @@
 package storage

-import "time"
+import (
+	"time"
+
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore"
+)
+
+type EncodingType string
+
+const (
+	defaultEncoding                     = JSONEncoding
+	JSONEncoding           EncodingType = "json"
+	ProtobufEncoding       EncodingType = "protobuf"
+	defaultMaxSpanCount                 = int(1e7)
+	defaultBatchSize                    = 10_000
+	defaultBatchDelay                   = time.Second * 5
+	defaultUsername                     = "default"
+	defaultDatabaseName                 = "default"
+	defaultMetricsEndpoint              = "localhost:9090"
+	defaultMaxNumSpans                  = 0
+
+	defaultSpansTable      clickhousespanstore.TableName = "jaeger_spans"
+	defaultSpansIndexTable clickhousespanstore.TableName = "jaeger_index"
+	defaultOperationsTable clickhousespanstore.TableName = "jaeger_operations"
+)

 type Configuration struct {
-	// ClickHouse address e.g. tcp://localhost:9000.
-	Address string `yaml:"address"`
-	// Directory with .sql files that are run at plugin startup.
-	InitSQLScriptsDir string `yaml:"init_sql_scripts_dir"`
 	// Batch write size. Default is 10_000.
-	BatchWriteSize int `yaml:"batch_write_size"`
+	BatchWriteSize int64 `yaml:"batch_write_size"`
 	// Batch flush interval. Default is 5s.
 	BatchFlushInterval time.Duration `yaml:"batch_flush_interval"`
+	// Maximal amount of spans that can be pending writes at a time.
+	// New spans exceeding this limit will be discarded,
+	// keeping memory in check if there are issues writing to ClickHouse.
+	// Check the "jaeger_clickhouse_discarded_spans" metric to keep track of discards.
+	// Default 10_000_000, or disable the limit entirely by setting to 0.
+	MaxSpanCount int `yaml:"max_span_count"`
 	// Encoding either json or protobuf. Default is json.
-	Encoding string `yaml:"encoding"`
+	Encoding EncodingType `yaml:"encoding"`
+	// ClickHouse address e.g. localhost:9000.
+	Address string `yaml:"address"`
+	// Directory with .sql files to run at plugin startup, mainly for integration tests.
+	// Depending on the value of init_tables, this can be run as a
+	// replacement or supplement to creating default tables for span storage.
+	// If init_tables is also enabled, the scripts in this directory will be run first.
+	InitSQLScriptsDir string `yaml:"init_sql_scripts_dir"`
+	// Whether to automatically attempt to create tables in ClickHouse.
+	// By default, this is enabled if init_sql_scripts_dir is empty,
+	// or disabled if init_sql_scripts_dir is provided.
+	InitTables *bool `yaml:"init_tables"`
+	// Indicates location of TLS certificate used to connect to database.
+	CaFile string `yaml:"ca_file"`
+	// Username for connection to database. Default is "default".
+	Username string `yaml:"username"`
+	// Password for connection to database.
+	Password string `yaml:"password"`
+	// Database name. Default is "default"
+	Database string `yaml:"database"`
+	// Endpoint for scraping prometheus metrics e.g. localhost:9090.
+	MetricsEndpoint string `yaml:"metrics_endpoint"`
+	// Whether to use SQL scripts supporting replication and sharding. Default false.
+	Replication bool `yaml:"replication"`
+	// If non-empty, enables multitenancy in SQL scripts, and assigns the tenant name for this instance.
+	Tenant string `yaml:"tenant"`
+	// Table with spans. Default "jaeger_spans_local" or "jaeger_spans" when replication is enabled.
+	SpansTable clickhousespanstore.TableName `yaml:"spans_table"`
+	// Span index table. Default "jaeger_index_local" or "jaeger_index" when replication is enabled.
+	SpansIndexTable clickhousespanstore.TableName `yaml:"spans_index_table"`
+	// Operations table. Default "jaeger_operations_local" or "jaeger_operations" when replication is enabled.
+	OperationsTable   clickhousespanstore.TableName `yaml:"operations_table"`
+	spansArchiveTable clickhousespanstore.TableName
+	// TTL for data in tables in days. If 0, no TTL is set. Default 0.
+	TTLDays uint `yaml:"ttl"`
+	// The maximum number of spans to fetch per trace. If 0, no limits is set. Default 0.
+	MaxNumSpans uint `yaml:"max_num_spans"`
+	// The maximum number of open connections to the database. Default is unlimited (see: https://pkg.go.dev/database/sql#DB.SetMaxOpenConns)
+	MaxOpenConns *uint `yaml:"max_open_conns"`
+	// The maximum number of database connections in the idle connection pool. Default 2. (see: https://pkg.go.dev/database/sql#DB.SetMaxIdleConns)
+	MaxIdleConns *uint `yaml:"max_idle_conns"`
+	// The maximum amount of milliseconds a database connection may be reused. Default = connections are never closed due to age (see: https://pkg.go.dev/database/sql#DB.SetConnMaxLifetime)
+	ConnMaxLifetimeMillis *uint `yaml:"conn_max_lifetime_millis"`
+	// The maximum amount of milliseconds a database connection may be idle. Default = connections are never closed due to idle time (see: https://pkg.go.dev/database/sql#DB.SetConnMaxIdleTime)
+	ConnMaxIdleTimeMillis *uint `yaml:"conn_max_idle_time_millis"`
+}
+
+func (cfg *Configuration) setDefaults() {
+	if cfg.BatchWriteSize == 0 {
+		cfg.BatchWriteSize = defaultBatchSize
+	}
+	if cfg.BatchFlushInterval == 0 {
+		cfg.BatchFlushInterval = defaultBatchDelay
+	}
+	if cfg.MaxSpanCount == 0 {
+		cfg.MaxSpanCount = defaultMaxSpanCount
+	}
+	if cfg.Encoding == "" {
+		cfg.Encoding = defaultEncoding
+	}
+	if cfg.InitTables == nil {
+		// Decide whether to init tables based on whether a custom script path was provided
+		var defaultInitTables bool
+		if cfg.InitSQLScriptsDir == "" {
+			defaultInitTables = true
+		} else {
+			defaultInitTables = false
+		}
+		cfg.InitTables = &defaultInitTables
+	}
+	if cfg.Username == "" {
+		cfg.Username = defaultUsername
+	}
+	if cfg.Database == "" {
+		cfg.Database = defaultDatabaseName
+	}
+	if cfg.MetricsEndpoint == "" {
+		cfg.MetricsEndpoint = defaultMetricsEndpoint
+	}
+	if cfg.MaxNumSpans == 0 {
+		cfg.MaxNumSpans = defaultMaxNumSpans
+	}
+	if cfg.SpansTable == "" {
+		if cfg.Replication {
+			cfg.SpansTable = defaultSpansTable
+			cfg.spansArchiveTable = defaultSpansTable + "_archive"
+		} else {
+			cfg.SpansTable = defaultSpansTable.ToLocal()
+			cfg.spansArchiveTable = (defaultSpansTable + "_archive").ToLocal()
+		}
+	} else {
+		cfg.spansArchiveTable = cfg.SpansTable + "_archive"
+	}
+	if cfg.SpansIndexTable == "" {
+		if cfg.Replication {
+			cfg.SpansIndexTable = defaultSpansIndexTable
+		} else {
+			cfg.SpansIndexTable = defaultSpansIndexTable.ToLocal()
+		}
+	}
+	if cfg.OperationsTable == "" {
+		if cfg.Replication {
+			cfg.OperationsTable = defaultOperationsTable
+		} else {
+			cfg.OperationsTable = defaultOperationsTable.ToLocal()
+		}
+	}
+}
+
+func (cfg *Configuration) GetSpansArchiveTable() clickhousespanstore.TableName {
+	return cfg.spansArchiveTable
 }
--- a/storage/config_test.go
+++ b/storage/config_test.go
@ -0,0 +1,128 @@
+package storage
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore"
+)
+
+func TestSetDefaults(t *testing.T) {
+	tests := map[string]struct {
+		replication bool
+		getField    func(Configuration) interface{}
+		expected    interface{}
+	}{
+		"username": {
+			getField: func(config Configuration) interface{} { return config.Username },
+			expected: defaultUsername,
+		},
+		"database name": {
+			getField: func(config Configuration) interface{} { return config.Database },
+			expected: defaultDatabaseName,
+		},
+		"encoding": {
+			getField: func(config Configuration) interface{} { return config.Encoding },
+			expected: defaultEncoding,
+		},
+		"batch write size": {
+			getField: func(config Configuration) interface{} { return config.BatchWriteSize },
+			expected: defaultBatchSize,
+		},
+		"batch flush interval": {
+			getField: func(config Configuration) interface{} { return config.BatchFlushInterval },
+			expected: defaultBatchDelay,
+		},
+		"max span count": {
+			getField: func(config Configuration) interface{} { return config.MaxSpanCount },
+			expected: defaultMaxSpanCount,
+		},
+		"metrics endpoint": {
+			getField: func(config Configuration) interface{} { return config.MetricsEndpoint },
+			expected: defaultMetricsEndpoint,
+		},
+		"spans table name local": {
+			getField: func(config Configuration) interface{} { return config.SpansTable },
+			expected: defaultSpansTable.ToLocal(),
+		},
+		"spans table name replication": {
+			replication: true,
+			getField:    func(config Configuration) interface{} { return config.SpansTable },
+			expected:    defaultSpansTable,
+		},
+		"index table name local": {
+			getField: func(config Configuration) interface{} { return config.SpansIndexTable },
+			expected: defaultSpansIndexTable.ToLocal(),
+		},
+		"index table name replication": {
+			replication: true,
+			getField:    func(config Configuration) interface{} { return config.SpansIndexTable },
+			expected:    defaultSpansIndexTable,
+		},
+		"operations table name local": {
+			getField: func(config Configuration) interface{} { return config.OperationsTable },
+			expected: defaultOperationsTable.ToLocal(),
+		},
+		"operations table name replication": {
+			replication: true,
+			getField:    func(config Configuration) interface{} { return config.OperationsTable },
+			expected:    defaultOperationsTable,
+		},
+		"max number spans": {
+			getField: func(config Configuration) interface{} { return config.MaxNumSpans },
+			expected: defaultMaxNumSpans,
+		},
+	}
+
+	for name, test := range tests {
+		t.Run(fmt.Sprintf("default %s", name), func(t *testing.T) {
+			config := Configuration{Replication: test.replication}
+			config.setDefaults()
+			assert.EqualValues(t, test.expected, test.getField(config))
+		})
+	}
+}
+
+func TestConfiguration_GetSpansArchiveTable(t *testing.T) {
+	tests := map[string]struct {
+		config                        Configuration
+		expectedSpansArchiveTableName clickhousespanstore.TableName
+	}{
+		"default_config_local":       {config: Configuration{}, expectedSpansArchiveTableName: (defaultSpansTable + "_archive").ToLocal()},
+		"default_config_replication": {config: Configuration{Replication: true}, expectedSpansArchiveTableName: defaultSpansTable + "_archive"},
+		"custom_spans_table":         {config: Configuration{SpansTable: "custom_table_name"}, expectedSpansArchiveTableName: "custom_table_name_archive"},
+	}
+
+	for name, test := range tests {
+		t.Run(name, func(t *testing.T) {
+			test.config.setDefaults()
+			assert.Equal(t, test.expectedSpansArchiveTableName, test.config.GetSpansArchiveTable())
+		})
+	}
+}
+
+func TestConfiguration_InitTables(test *testing.T) {
+	// for pointers below
+	t := true
+	f := false
+	tests := map[string]struct {
+		config             Configuration
+		expectedInitTables bool
+	}{
+		"scriptsempty_initnil":      {config: Configuration{}, expectedInitTables: true},
+		"scriptsprovided_initnil":   {config: Configuration{InitSQLScriptsDir: "hello"}, expectedInitTables: false},
+		"scriptsempty_inittrue":     {config: Configuration{InitTables: &t}, expectedInitTables: true},
+		"scriptsprovided_inittrue":  {config: Configuration{InitSQLScriptsDir: "hello", InitTables: &t}, expectedInitTables: true},
+		"scriptsempty_initfalse":    {config: Configuration{InitTables: &f}, expectedInitTables: false},
+		"scriptsprovided_initfalse": {config: Configuration{InitSQLScriptsDir: "hello", InitTables: &f}, expectedInitTables: false},
+	}
+
+	for name, testcase := range tests {
+		test.Run(name, func(t *testing.T) {
+			testcase.config.setDefaults()
+			assert.Equal(t, testcase.expectedInitTables, *(testcase.config.InitTables))
+		})
+	}
+}
--- a/storage/store.go
+++ b/storage/store.go
@ -1,146 +1,341 @@
 package storage

 import (
+	"crypto/tls"
+	"crypto/x509"
 	"database/sql"
-	"embed"
 	"fmt"
 	"io"
-	"io/ioutil"
 	"os"
 	"path/filepath"
 	"sort"
+	"strings"
+	"text/template"
 	"time"

-	_ "github.com/ClickHouse/clickhouse-go" // force SQL driver registration
-	"github.com/hashicorp/go-hclog"
+	clickhouse "github.com/ClickHouse/clickhouse-go/v2"
+	hclog "github.com/hashicorp/go-hclog"
 	"github.com/jaegertracing/jaeger/plugin/storage/grpc/shared"
 	"github.com/jaegertracing/jaeger/storage/dependencystore"
 	"github.com/jaegertracing/jaeger/storage/spanstore"

-	"github.com/pavolloffay/jaeger-clickhouse/storage/clickhousedependencystore"
-	"github.com/pavolloffay/jaeger-clickhouse/storage/clickhousespanstore"
-)
-
-const (
-	defaultBatchSize  = 10_000
-	defaultBatchDelay = time.Second * 5
+	jaegerclickhouse "github.com/jaegertracing/jaeger-clickhouse"
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousedependencystore"
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore"
 )

 type Store struct {
-	db     *sql.DB
-	logger hclog.Logger
-	cfg    Configuration
+	db            *sql.DB
+	writer        spanstore.Writer
+	reader        spanstore.Reader
+	archiveWriter spanstore.Writer
+	archiveReader spanstore.Reader
 }

-var _ shared.StoragePlugin = (*Store)(nil)
-var _ io.Closer = (*Store)(nil)
+var (
+	_ shared.StoragePlugin             = (*Store)(nil)
+	_ shared.ArchiveStoragePlugin      = (*Store)(nil)
+	_ shared.StreamingSpanWriterPlugin = (*Store)(nil)
+	_ io.Closer                        = (*Store)(nil)
+)

-func NewStore(logger hclog.Logger, cfg Configuration, embeddedSQLScripts embed.FS) (*Store, error) {
-	db, err := defaultConnector(cfg.Address)
+func NewStore(logger hclog.Logger, cfg Configuration) (*Store, error) {
+	cfg.setDefaults()
+	db, err := connector(cfg)
 	if err != nil {
 		return nil, fmt.Errorf("could not connect to database: %q", err)
 	}

-	if err := initializeDB(db, cfg.InitSQLScriptsDir, embeddedSQLScripts); err != nil {
-		db.Close()
+	if err := runInitScripts(logger, db, cfg); err != nil {
+		_ = db.Close()
 		return nil, err
 	}
-	if cfg.BatchWriteSize == 0 {
-		cfg.BatchWriteSize = defaultBatchSize
-	}
-	if cfg.BatchFlushInterval == 0 {
-		cfg.BatchFlushInterval = defaultBatchDelay
-	}
-	if cfg.Encoding == "" {
-		cfg.Encoding = string(clickhousespanstore.EncodingJSON)
+	if cfg.Replication {
+		return &Store{
+			db: db,
+			writer: clickhousespanstore.NewSpanWriter(
+				logger,
+				db,
+				cfg.SpansIndexTable,
+				cfg.SpansTable,
+				cfg.Tenant,
+				clickhousespanstore.Encoding(cfg.Encoding),
+				cfg.BatchFlushInterval,
+				cfg.BatchWriteSize,
+				cfg.MaxSpanCount,
+			),
+			reader: clickhousespanstore.NewTraceReader(
+				db,
+				cfg.OperationsTable,
+				cfg.SpansIndexTable,
+				cfg.SpansTable,
+				cfg.Tenant,
+				cfg.MaxNumSpans,
+			),
+			archiveWriter: clickhousespanstore.NewSpanWriter(
+				logger,
+				db,
+				"",
+				cfg.GetSpansArchiveTable(),
+				cfg.Tenant,
+				clickhousespanstore.Encoding(cfg.Encoding),
+				cfg.BatchFlushInterval,
+				cfg.BatchWriteSize,
+				cfg.MaxSpanCount,
+			),
+			archiveReader: clickhousespanstore.NewTraceReader(
+				db,
+				"",
+				"",
+				cfg.GetSpansArchiveTable(),
+				cfg.Tenant,
+				cfg.MaxNumSpans,
+			),
+		}, nil
 	}
 	return &Store{
-		db:     db,
-		logger: logger,
-		cfg:    cfg,
+		db: db,
+		writer: clickhousespanstore.NewSpanWriter(
+			logger,
+			db,
+			cfg.SpansIndexTable,
+			cfg.SpansTable,
+			cfg.Tenant,
+			clickhousespanstore.Encoding(cfg.Encoding),
+			cfg.BatchFlushInterval,
+			cfg.BatchWriteSize,
+			cfg.MaxSpanCount,
+		),
+		reader: clickhousespanstore.NewTraceReader(
+			db,
+			cfg.OperationsTable,
+			cfg.SpansIndexTable,
+			cfg.SpansTable,
+			cfg.Tenant,
+			cfg.MaxNumSpans,
+		),
+		archiveWriter: clickhousespanstore.NewSpanWriter(
+			logger,
+			db,
+			"",
+			cfg.GetSpansArchiveTable(),
+			cfg.Tenant,
+			clickhousespanstore.Encoding(cfg.Encoding),
+			cfg.BatchFlushInterval,
+			cfg.BatchWriteSize,
+			cfg.MaxSpanCount,
+		),
+		archiveReader: clickhousespanstore.NewTraceReader(
+			db,
+			"",
+			"",
+			cfg.GetSpansArchiveTable(),
+			cfg.Tenant,
+			cfg.MaxNumSpans,
+		),
 	}, nil
 }

-func initializeDB(db *sql.DB, initSQLScriptsDir string, embeddedScripts embed.FS) error {
-	var sqlStatements []string
-	if initSQLScriptsDir != "" {
-		filePaths, err := walkMatch(initSQLScriptsDir, "*.sql")
+func connector(cfg Configuration) (*sql.DB, error) {
+	var conn *sql.DB
+
+	options := clickhouse.Options{
+		Addr: []string{sanitize(cfg.Address)},
+		Auth: clickhouse.Auth{
+			Database: cfg.Database,
+			Username: cfg.Username,
+			Password: cfg.Password,
+		},
+		Compression: &clickhouse.Compression{
+			Method: clickhouse.CompressionLZ4,
+		},
+	}
+
+	if cfg.CaFile != "" {
+		caCert, err := os.ReadFile(cfg.CaFile)
+		if err != nil {
+			return nil, err
+		}
+		caCertPool := x509.NewCertPool()
+		caCertPool.AppendCertsFromPEM(caCert)
+		options.TLS = &tls.Config{
+			RootCAs: caCertPool,
+		}
+	}
+	conn = clickhouse.OpenDB(&options)
+
+	if cfg.MaxOpenConns != nil {
+		conn.SetMaxIdleConns(int(*cfg.MaxOpenConns))
+	}
+	if cfg.MaxIdleConns != nil {
+		conn.SetMaxIdleConns(int(*cfg.MaxIdleConns))
+	}
+	if cfg.ConnMaxLifetimeMillis != nil {
+		conn.SetConnMaxLifetime(time.Millisecond * time.Duration(*cfg.ConnMaxLifetimeMillis))
+	}
+	if cfg.ConnMaxIdleTimeMillis != nil {
+		conn.SetConnMaxIdleTime(time.Millisecond * time.Duration(*cfg.ConnMaxIdleTimeMillis))
+	}
+
+	if err := conn.Ping(); err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+type tableArgs struct {
+	Database string
+
+	SpansIndexTable   clickhousespanstore.TableName
+	SpansTable        clickhousespanstore.TableName
+	OperationsTable   clickhousespanstore.TableName
+	SpansArchiveTable clickhousespanstore.TableName
+
+	TTLTimestamp string
+	TTLDate      string
+
+	Multitenant bool
+	Replication bool
+}
+
+type distributedTableArgs struct {
+	Database string
+	Table    clickhousespanstore.TableName
+	Hash     string
+}
+
+func render(templates *template.Template, filename string, args interface{}) string {
+	var statement strings.Builder
+	err := templates.ExecuteTemplate(&statement, filename, args)
+	if err != nil {
+		panic(err)
+	}
+	return statement.String()
+}
+
+func runInitScripts(logger hclog.Logger, db *sql.DB, cfg Configuration) error {
+	var (
+		sqlStatements []string
+		ttlTimestamp  string
+		ttlDate       string
+	)
+	if cfg.TTLDays > 0 {
+		ttlTimestamp = fmt.Sprintf("TTL timestamp + INTERVAL %d DAY DELETE", cfg.TTLDays)
+		ttlDate = fmt.Sprintf("TTL date + INTERVAL %d DAY DELETE", cfg.TTLDays)
+	}
+	if cfg.InitSQLScriptsDir != "" {
+		filePaths, err := walkMatch(cfg.InitSQLScriptsDir, "*.sql")
 		if err != nil {
 			return fmt.Errorf("could not list sql files: %q", err)
 		}
 		sort.Strings(filePaths)
 		for _, f := range filePaths {
-			sqlStatement, err := ioutil.ReadFile(f)
+			sqlStatement, err := os.ReadFile(filepath.Clean(f))
 			if err != nil {
 				return err
 			}
 			sqlStatements = append(sqlStatements, string(sqlStatement))
 		}
-	} else {
-		f, err := embeddedScripts.ReadFile("sqlscripts/0001-jaeger-index.sql")
-		if err != nil {
-			return err
-		}
-		sqlStatements = append(sqlStatements, string(f))
-		f, err = embeddedScripts.ReadFile("sqlscripts/0002-jaeger-spans.sql")
-		if err != nil {
-			return err
-		}
-		sqlStatements = append(sqlStatements, string(f))
-		f, err = embeddedScripts.ReadFile("sqlscripts/0003-jaeger-operations.sql")
-		if err != nil {
-			return err
-		}
-		sqlStatements = append(sqlStatements, string(f))
 	}
-	return executeScripts(sqlStatements, db)
+	if *cfg.InitTables {
+		templates := template.Must(template.ParseFS(jaegerclickhouse.SQLScripts, "sqlscripts/*.tmpl.sql"))
+
+		args := tableArgs{
+			Database: cfg.Database,
+
+			SpansIndexTable:   cfg.SpansIndexTable,
+			SpansTable:        cfg.SpansTable,
+			OperationsTable:   cfg.OperationsTable,
+			SpansArchiveTable: cfg.GetSpansArchiveTable(),
+
+			TTLTimestamp: ttlTimestamp,
+			TTLDate:      ttlDate,
+
+			Multitenant: cfg.Tenant != "",
+			Replication: cfg.Replication,
+		}
+
+		if cfg.Replication {
+			// Add "_local" to the local table names, and omit it from the distributed tables below
+			args.SpansIndexTable = args.SpansIndexTable.ToLocal()
+			args.SpansTable = args.SpansTable.ToLocal()
+			args.OperationsTable = args.OperationsTable.ToLocal()
+			args.SpansArchiveTable = args.SpansArchiveTable.ToLocal()
+		}
+
+		sqlStatements = append(sqlStatements, render(templates, "jaeger-index.tmpl.sql", args))
+		sqlStatements = append(sqlStatements, render(templates, "jaeger-operations.tmpl.sql", args))
+		sqlStatements = append(sqlStatements, render(templates, "jaeger-spans.tmpl.sql", args))
+		sqlStatements = append(sqlStatements, render(templates, "jaeger-spans-archive.tmpl.sql", args))
+
+		if cfg.Replication {
+			// Now these tables omit the "_local" suffix
+			distargs := distributedTableArgs{
+				Table:    cfg.SpansTable,
+				Database: cfg.Database,
+				Hash:     "cityHash64(traceID)",
+			}
+			sqlStatements = append(sqlStatements, render(templates, "distributed-table.tmpl.sql", distargs))
+
+			distargs.Table = cfg.SpansIndexTable
+			sqlStatements = append(sqlStatements, render(templates, "distributed-table.tmpl.sql", distargs))
+
+			distargs.Table = cfg.GetSpansArchiveTable()
+			sqlStatements = append(sqlStatements, render(templates, "distributed-table.tmpl.sql", distargs))
+
+			distargs.Table = cfg.OperationsTable
+			distargs.Hash = "rand()"
+			sqlStatements = append(sqlStatements, render(templates, "distributed-table.tmpl.sql", distargs))
+		}
+	}
+	return executeScripts(logger, sqlStatements, db)
 }

 func (s *Store) SpanReader() spanstore.Reader {
-	return clickhousespanstore.NewTraceReader(s.db, "jaeger_operations_v2", "jaeger_index_v2", "jaeger_spans_v2")
+	return s.reader
 }

 func (s *Store) SpanWriter() spanstore.Writer {
-	return clickhousespanstore.NewSpanWriter(s.logger, s.db, "jaeger_index_v2", "jaeger_spans_v2", clickhousespanstore.Encoding(s.cfg.Encoding), s.cfg.BatchFlushInterval, s.cfg.BatchWriteSize)
+	return s.writer
 }

 func (s *Store) DependencyReader() dependencystore.Reader {
 	return clickhousedependencystore.NewDependencyStore()
 }

+func (s *Store) ArchiveSpanReader() spanstore.Reader {
+	return s.archiveReader
+}
+
+func (s *Store) ArchiveSpanWriter() spanstore.Writer {
+	return s.archiveWriter
+}
+
+func (s *Store) StreamingSpanWriter() spanstore.Writer {
+	return s.writer
+}
+
 func (s *Store) Close() error {
 	return s.db.Close()
 }

-func defaultConnector(datasource string) (*sql.DB, error) {
-	db, err := sql.Open("clickhouse", datasource)
-	if err != nil {
-		return nil, err
-	}
-
-	if err := db.Ping(); err != nil {
-		return nil, err
-	}
-
-	return db, nil
-}
-
-func executeScripts(sqlStatements []string, db *sql.DB) error {
+func executeScripts(logger hclog.Logger, sqlStatements []string, db *sql.DB) error {
 	tx, err := db.Begin()
 	if err != nil {
-		return nil
+		return err
 	}
 	committed := false
 	defer func() {
 		if !committed {
-			tx.Rollback()
+			_ = tx.Rollback()
 		}
 	}()

-	for _, file := range sqlStatements {
-		_, err = tx.Exec(file)
+	for _, statement := range sqlStatements {
+		logger.Debug("Running SQL statement", "statement", statement)
+		_, err = tx.Exec(statement)
 		if err != nil {
-			return fmt.Errorf("could not run sql %q: %q", file, err)
+			return fmt.Errorf("could not run sql %q: %q", statement, err)
 		}
 	}
 	committed = true
@ -168,3 +363,10 @@ func walkMatch(root, pattern string) ([]string, error) {
 	}
 	return matches, nil
 }
+
+// Earlier version of clickhouse-go used to expect address as tcp://host:port
+// while newer version of clickhouse-go expect address as host:port (without scheme)
+// so to maintain backward compatibility we clean it up
+func sanitize(addr string) string {
+	return strings.TrimPrefix(addr, "tcp://")
+}
--- a/storage/store_test.go
+++ b/storage/store_test.go
@ -0,0 +1,187 @@
+package storage
+
+import (
+	"database/sql"
+	"fmt"
+	"testing"
+
+	sqlmock "github.com/DATA-DOG/go-sqlmock"
+	hclog "github.com/hashicorp/go-hclog"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousedependencystore"
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore"
+	"github.com/jaegertracing/jaeger-clickhouse/storage/clickhousespanstore/mocks"
+)
+
+const (
+	testIndexTable        = "test_index_table"
+	testSpansTable        = "test_spans_table"
+	testOperationsTable   = "test_operation_table"
+	testSpansArchiveTable = "test_spans_archive_table"
+)
+
+var errorMock = fmt.Errorf("error mock")
+
+func TestStore_SpanWriter(t *testing.T) {
+	writer := clickhousespanstore.SpanWriter{}
+	store := Store{
+		writer: &writer,
+	}
+	assert.Equal(t, &writer, store.SpanWriter())
+}
+
+func TestStore_ArchiveSpanWriter(t *testing.T) {
+	writer := clickhousespanstore.SpanWriter{}
+	store := Store{
+		archiveWriter: &writer,
+	}
+	assert.Equal(t, &writer, store.ArchiveSpanWriter())
+}
+
+func TestStore_SpanReader(t *testing.T) {
+	reader := clickhousespanstore.TraceReader{}
+	store := Store{
+		reader: &reader,
+	}
+	assert.Equal(t, &reader, store.SpanReader())
+}
+
+func TestStore_ArchiveSpanReader(t *testing.T) {
+	reader := clickhousespanstore.TraceReader{}
+	store := Store{
+		archiveReader: &reader,
+	}
+	assert.Equal(t, &reader, store.ArchiveSpanReader())
+}
+
+func TestStore_DependencyReader(t *testing.T) {
+	store := Store{}
+	assert.Equal(t, &clickhousedependencystore.DependencyStore{}, store.DependencyReader())
+}
+
+func TestStore_Close(t *testing.T) {
+	db, mock, err := mocks.GetDbMock()
+	require.NoError(t, err)
+	defer db.Close()
+
+	logger := mocks.NewSpyLogger()
+	store := newStore(db, logger)
+
+	mock.ExpectClose()
+	require.NoError(t, store.Close())
+	assert.NoError(t, mock.ExpectationsWereMet())
+	logger.AssertLogsEmpty(t)
+}
+
+func newStore(db *sql.DB, logger mocks.SpyLogger) Store {
+	return Store{
+		db: db,
+		writer: clickhousespanstore.NewSpanWriter(
+			logger,
+			db,
+			testIndexTable,
+			testSpansTable,
+			"",
+			clickhousespanstore.EncodingJSON,
+			0,
+			0,
+			0,
+		),
+		reader: clickhousespanstore.NewTraceReader(
+			db,
+			testOperationsTable,
+			testIndexTable,
+			testSpansTable,
+			"",
+			0,
+		),
+		archiveWriter: clickhousespanstore.NewSpanWriter(
+			logger,
+			db,
+			testIndexTable,
+			testSpansArchiveTable,
+			"",
+			clickhousespanstore.EncodingJSON,
+			0,
+			0,
+			0,
+		),
+		archiveReader: clickhousespanstore.NewTraceReader(
+			db,
+			testOperationsTable,
+			testIndexTable,
+			testSpansArchiveTable,
+			"",
+			0,
+		),
+	}
+}
+
+func TestStore_executeScripts(t *testing.T) {
+	db, mock, err := mocks.GetDbMock()
+	require.NoError(t, err)
+	defer db.Close()
+
+	spyLogger := mocks.NewSpyLogger()
+	scripts := []string{
+		"first SQL script",
+		"second_SQL_script",
+	}
+
+	mock.ExpectBegin()
+	for _, script := range scripts {
+		mock.ExpectExec(script).WillReturnResult(sqlmock.NewResult(1, 1))
+	}
+	mock.ExpectCommit()
+	err = executeScripts(spyLogger, scripts, db)
+	require.NoError(t, err)
+	assert.NoError(t, mock.ExpectationsWereMet())
+	spyLogger.AssertLogsOfLevelEqual(t, hclog.Debug, func() []mocks.LogMock {
+		res := make([]mocks.LogMock, len(scripts))
+		for i, script := range scripts {
+			res[i] = mocks.LogMock{Msg: "Running SQL statement", Args: []interface{}{"statement", script}}
+		}
+		return res
+	}())
+}
+
+func TestStore_executeScriptsExecuteError(t *testing.T) {
+	db, mock, err := mocks.GetDbMock()
+	require.NoError(t, err)
+	defer db.Close()
+
+	spyLogger := mocks.NewSpyLogger()
+	scripts := []string{
+		"first SQL script",
+		"second_SQL_script",
+	}
+
+	mock.ExpectBegin()
+	mock.ExpectExec(scripts[0]).WillReturnError(errorMock)
+	mock.ExpectRollback()
+	err = executeScripts(spyLogger, scripts, db)
+	assert.EqualError(t, err, fmt.Sprintf("could not run sql %q: %q", scripts[0], errorMock))
+	spyLogger.AssertLogsOfLevelEqual(
+		t,
+		hclog.Debug,
+		[]mocks.LogMock{{Msg: "Running SQL statement", Args: []interface{}{"statement", scripts[0]}}},
+	)
+}
+
+func TestStore_executeScriptBeginError(t *testing.T) {
+	db, mock, err := mocks.GetDbMock()
+	require.NoError(t, err)
+	defer db.Close()
+
+	spyLogger := mocks.NewSpyLogger()
+	scripts := []string{
+		"first SQL script",
+		"second_SQL_script",
+	}
+
+	mock.ExpectBegin().WillReturnError(errorMock)
+	err = executeScripts(spyLogger, scripts, db)
+	assert.EqualError(t, err, errorMock.Error())
+}