feat(metric): add dragonfly-client-metric module (#1362)

* feat(metric): add dragonfly-client-metric module

Signed-off-by: Gaius <gaius.qi@gmail.com>

* chore(deps): update dragonfly-client versions to 1.0.22

Signed-off-by: Gaius <gaius.qi@gmail.com>

---------

Signed-off-by: Gaius <gaius.qi@gmail.com>

Cargo.lock

@@ -986,7 +986,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "anyhow",
  "bytes",
@@ -999,6 +999,7 @@ dependencies = [
  "dragonfly-client-backend",
  "dragonfly-client-config",
  "dragonfly-client-core",
+ "dragonfly-client-metric",
  "dragonfly-client-storage",
  "dragonfly-client-util",
  "fastrand",
@@ -1025,7 +1026,6 @@ dependencies = [
  "path-absolutize",
  "percent-encoding",
  "pprof",
- "prometheus",
  "prost-wkt-types",
  "rcgen",
  "reqwest",
@@ -1060,7 +1060,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client-backend"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "dragonfly-api",
  "dragonfly-client-core",
@@ -1091,7 +1091,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client-config"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "bytesize",
  "bytesize-serde",
@@ -1121,7 +1121,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client-core"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "headers 0.4.1",
  "hyper 1.6.0",
@@ -1140,7 +1140,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client-init"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "anyhow",
  "clap",
@@ -1155,9 +1155,24 @@ dependencies = [
  "url",
 ]
 
+[[package]]
+name = "dragonfly-client-metric"
+version = "1.0.22"
+dependencies = [
+ "dragonfly-api",
+ "dragonfly-client-config",
+ "dragonfly-client-util",
+ "fs2",
+ "lazy_static",
+ "prometheus",
+ "tokio",
+ "tracing",
+ "warp",
+]
+
 [[package]]
 name = "dragonfly-client-storage"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "bincode",
  "bytes",
@@ -1168,6 +1183,7 @@ dependencies = [
  "dragonfly-api",
  "dragonfly-client-config",
  "dragonfly-client-core",
+ "dragonfly-client-metric",
  "dragonfly-client-util",
  "fs2",
  "leaky-bucket",
@@ -1186,7 +1202,7 @@ dependencies = [
 
 [[package]]
 name = "dragonfly-client-util"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "base64 0.22.1",
  "bytesize",
@@ -1597,7 +1613,7 @@ dependencies = [
 
 [[package]]
 name = "hdfs"
-version = "1.0.21"
+version = "1.0.22"
 dependencies = [
  "dragonfly-client-backend",
  "dragonfly-client-core",

Cargo.toml

@@ -9,10 +9,11 @@ members = [
     "dragonfly-client-storage",
     "dragonfly-client-util",
     "dragonfly-client-backend/examples/plugin",
+    "dragonfly-client-metric",
 ]
 
 [workspace.package]
-version = "1.0.21"
+version = "1.0.22"
 authors = ["The Dragonfly Developers"]
 homepage = "https://d7y.io/"
 repository = "https://github.com/dragonflyoss/client.git"
@@ -22,13 +23,14 @@ readme = "README.md"
 edition = "2021"
 
 [workspace.dependencies]
-dragonfly-client = { path = "dragonfly-client", version = "1.0.21" }
+dragonfly-client = { path = "dragonfly-client", version = "1.0.22" }
-dragonfly-client-core = { path = "dragonfly-client-core", version = "1.0.21" }
+dragonfly-client-core = { path = "dragonfly-client-core", version = "1.0.22" }
-dragonfly-client-config = { path = "dragonfly-client-config", version = "1.0.21" }
+dragonfly-client-config = { path = "dragonfly-client-config", version = "1.0.22" }
-dragonfly-client-storage = { path = "dragonfly-client-storage", version = "1.0.21" }
+dragonfly-client-storage = { path = "dragonfly-client-storage", version = "1.0.22" }
-dragonfly-client-backend = { path = "dragonfly-client-backend", version = "1.0.21" }
+dragonfly-client-backend = { path = "dragonfly-client-backend", version = "1.0.22" }
-dragonfly-client-util = { path = "dragonfly-client-util", version = "1.0.21" }
+dragonfly-client-metric = { path = "dragonfly-client-metric", version = "1.0.22" }
-dragonfly-client-init = { path = "dragonfly-client-init", version = "1.0.21" }
+dragonfly-client-util = { path = "dragonfly-client-util", version = "1.0.22" }
+dragonfly-client-init = { path = "dragonfly-client-init", version = "1.0.22" }
 dragonfly-api = "=2.1.68"
 thiserror = "2.0"
 futures = "0.3.31"

ci/Dockerfile

@@ -29,6 +29,9 @@ COPY dragonfly-client-backend/src ./dragonfly-client-backend/src
 COPY dragonfly-client-backend/examples/plugin/Cargo.toml ./dragonfly-client-backend/examples/plugin/Cargo.toml
 COPY dragonfly-client-backend/examples/plugin/src ./dragonfly-client-backend/examples/plugin/src
 
+COPY dragonfly-client-metric/Cargo.toml ./dragonfly-client-metric/Cargo.toml
+COPY dragonfly-client-metric/src ./dragonfly-client-backend/src
+
 COPY dragonfly-client-util/Cargo.toml ./dragonfly-client-util/Cargo.toml
 COPY dragonfly-client-util/src ./dragonfly-client-util/src
 

ci/Dockerfile.debug

@@ -29,6 +29,9 @@ COPY dragonfly-client-backend/src ./dragonfly-client-backend/src
 COPY dragonfly-client-backend/examples/plugin/Cargo.toml ./dragonfly-client-backend/examples/plugin/Cargo.toml
 COPY dragonfly-client-backend/examples/plugin/src ./dragonfly-client-backend/examples/plugin/src
 
+COPY dragonfly-client-metric/Cargo.toml ./dragonfly-client-metric/Cargo.toml
+COPY dragonfly-client-metric/src ./dragonfly-client-backend/src
+
 COPY dragonfly-client-util/Cargo.toml ./dragonfly-client-util/Cargo.toml
 COPY dragonfly-client-util/src ./dragonfly-client-util/src
 

ci/Dockerfile.dfinit

@@ -29,6 +29,9 @@ COPY dragonfly-client-backend/src ./dragonfly-client-backend/src
 COPY dragonfly-client-backend/examples/plugin/Cargo.toml ./dragonfly-client-backend/examples/plugin/Cargo.toml
 COPY dragonfly-client-backend/examples/plugin/src ./dragonfly-client-backend/examples/plugin/src
 
+COPY dragonfly-client-metric/Cargo.toml ./dragonfly-client-metric/Cargo.toml
+COPY dragonfly-client-metric/src ./dragonfly-client-backend/src
+
 COPY dragonfly-client-util/Cargo.toml ./dragonfly-client-util/Cargo.toml
 COPY dragonfly-client-util/src ./dragonfly-client-util/src
 

dragonfly-client-metric/Cargo.toml

@@ -0,0 +1,22 @@
+[package]
+name = "dragonfly-client-metric"
+description = "Metric collection for the dragonfly client"
+version.workspace = true
+authors.workspace = true
+homepage.workspace = true
+repository.workspace = true
+keywords.workspace = true
+license.workspace = true
+readme.workspace = true
+edition.workspace = true
+
+[dependencies]
+dragonfly-api.workspace = true
+dragonfly-client-config.workspace = true
+dragonfly-client-util.workspace = true
+lazy_static.workspace = true
+tokio.workspace = true
+tracing.workspace = true
+warp.workspace = true
+fs2.workspace = true
+prometheus = { version = "0.13", features = ["process"] }

dragonfly-client-metric/src/lib.rs

@@ -1,5 +1,5 @@
 /*
- *     Copyright 2023 The Dragonfly Authors
+ *     Copyright 2025 The Dragonfly Authors
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.

dragonfly-client-storage/Cargo.toml

@@ -12,6 +12,7 @@ edition.workspace = true
 [dependencies]
 dragonfly-client-core.workspace = true
 dragonfly-client-config.workspace = true
+dragonfly-client-metric.workspace = true
 dragonfly-client-util.workspace = true
 dragonfly-api.workspace = true
 chrono.workspace = true

dragonfly-client-storage/src/server/tcp.rs

@@ -18,6 +18,9 @@ use crate::Storage;
 use bytes::{Bytes, BytesMut};
 use dragonfly_api::common::v2::TrafficType;
 use dragonfly_client_core::{Error as ClientError, Result as ClientResult};
+use dragonfly_client_metric::{
+    collect_upload_piece_failure_metrics, collect_upload_piece_started_metrics,
+};
 use dragonfly_client_util::{id_generator::IDGenerator, shutdown};
 use leaky_bucket::RateLimiter;
 use std::net::SocketAddr;
@@ -157,7 +160,11 @@ impl TCPServerHandler {
                 Span::current().record("remote_address", remote_address.as_str());
                 Span::current().record("task_id", task_id);
                 Span::current().record("piece_id", piece_id.as_str());
+
+                // Collect upload piece started metrics.
+                collect_upload_piece_started_metrics();
                 info!("start upload piece content");
+
                 match self.handle_piece(piece_id.as_str(), task_id).await {
                     Ok((piece_content, mut content_reader)) => {
                         let piece_content_bytes: Bytes = piece_content.into();
@@ -174,6 +181,9 @@ impl TCPServerHandler {
                         self.write_stream(&mut content_reader, &mut writer).await?;
                     }
                     Err(err) => {
+                        // Collect upload piece failure metrics.
+                        collect_upload_piece_failure_metrics();
+
                         let error_response: Bytes =
                             Vortex::Error(Header::new_error(err.len() as u32), err).into();
                         self.write_response(error_response, &mut writer).await?;
@@ -203,7 +213,11 @@ impl TCPServerHandler {
                 Span::current().record("remote_address", remote_address.as_str());
                 Span::current().record("task_id", task_id);
                 Span::current().record("piece_id", piece_id.as_str());
+
+                // Collect upload piece started metrics.
+                collect_upload_piece_started_metrics();
                 info!("start upload persistent cache piece content");
+
                 match self
                     .handle_persistent_cache_piece(piece_id.as_str(), task_id)
                     .await
@@ -227,6 +241,9 @@ impl TCPServerHandler {
                         self.write_stream(&mut content_reader, &mut writer).await?;
                     }
                     Err(err) => {
+                        // Collect upload piece failure metrics.
+                        collect_upload_piece_failure_metrics();
+
                         let error_response: Bytes =
                             Vortex::Error(Header::new_error(err.len() as u32), err).into();
                         self.write_response(error_response, &mut writer).await?;

dragonfly-client/Cargo.toml

@@ -27,6 +27,7 @@ dragonfly-client-core.workspace = true
 dragonfly-client-config.workspace = true
 dragonfly-client-storage.workspace = true
 dragonfly-client-backend.workspace = true
+dragonfly-client-metric.workspace = true
 dragonfly-client-util.workspace = true
 dragonfly-api.workspace = true
 rcgen.workspace = true
@@ -75,7 +76,6 @@ opentelemetry_sdk = { version = "0.29.0", default-features = false, features = [
 opentelemetry-semantic-conventions = { version = "0.30.0", features = ["semconv_experimental"] }
 rolling-file = "0.2.0"
 pprof = { version = "0.15", features = ["flamegraph", "protobuf-codec"] }
-prometheus = { version = "0.13", features = ["process"] }
 tonic-health = "0.12.3"
 tower = { version = "0.4.13", features = ["limit", "load-shed", "buffer"] }
 indicatif = "0.18.0"

dragonfly-client/src/bin/dfdaemon/main.rs

@@ -23,13 +23,13 @@ use dragonfly_client::grpc::{
     manager::ManagerClient, scheduler::SchedulerClient,
 };
 use dragonfly_client::health::Health;
-use dragonfly_client::metrics::Metrics;
 use dragonfly_client::proxy::Proxy;
 use dragonfly_client::resource::{persistent_cache_task::PersistentCacheTask, task::Task};
 use dragonfly_client::stats::Stats;
 use dragonfly_client::tracing::init_tracing;
 use dragonfly_client_backend::BackendFactory;
 use dragonfly_client_config::{dfdaemon, VersionValueParser};
+use dragonfly_client_metric::Metrics;
 use dragonfly_client_storage::{server::tcp::TCPServer, Storage};
 use dragonfly_client_util::{id_generator::IDGenerator, net::Interface, shutdown};
 use leaky_bucket::RateLimiter;

dragonfly-client/src/grpc/dfdaemon_download.rs

@@ -14,15 +14,6 @@
  * limitations under the License.
  */
 
-use crate::metrics::{
-    collect_delete_host_failure_metrics, collect_delete_host_started_metrics,
-    collect_delete_task_failure_metrics, collect_delete_task_started_metrics,
-    collect_download_task_failure_metrics, collect_download_task_finished_metrics,
-    collect_download_task_started_metrics, collect_list_task_entries_failure_metrics,
-    collect_list_task_entries_started_metrics, collect_stat_task_failure_metrics,
-    collect_stat_task_started_metrics, collect_upload_task_failure_metrics,
-    collect_upload_task_finished_metrics, collect_upload_task_started_metrics,
-};
 use crate::resource::{persistent_cache_task, task};
 use dragonfly_api::common::v2::{CacheTask, PersistentCacheTask, Priority, Task, TaskType};
 use dragonfly_api::dfdaemon::v2::{
@@ -44,6 +35,15 @@ use dragonfly_client_core::{
     error::{ErrorType, OrErr},
     Error as ClientError, Result as ClientResult,
 };
+use dragonfly_client_metric::{
+    collect_delete_host_failure_metrics, collect_delete_host_started_metrics,
+    collect_delete_task_failure_metrics, collect_delete_task_started_metrics,
+    collect_download_task_failure_metrics, collect_download_task_finished_metrics,
+    collect_download_task_started_metrics, collect_list_task_entries_failure_metrics,
+    collect_list_task_entries_started_metrics, collect_stat_task_failure_metrics,
+    collect_stat_task_started_metrics, collect_upload_task_failure_metrics,
+    collect_upload_task_finished_metrics, collect_upload_task_started_metrics,
+};
 use dragonfly_client_util::{
     digest::{verify_file_digest, Digest},
     http::{get_range, hashmap_to_headermap, headermap_to_hashmap},

dragonfly-client/src/grpc/dfdaemon_upload.rs

@@ -14,14 +14,6 @@
  * limitations under the License.
  */
 
-use crate::metrics::{
-    collect_delete_task_failure_metrics, collect_delete_task_started_metrics,
-    collect_download_task_failure_metrics, collect_download_task_finished_metrics,
-    collect_download_task_started_metrics, collect_stat_task_failure_metrics,
-    collect_stat_task_started_metrics, collect_update_task_failure_metrics,
-    collect_update_task_started_metrics, collect_upload_piece_failure_metrics,
-    collect_upload_piece_finished_metrics, collect_upload_piece_started_metrics,
-};
 use crate::resource::{persistent_cache_task, task};
 use dragonfly_api::common::v2::{
     CacheTask, Host, Network, PersistentCacheTask, Piece, Priority, Task, TaskType,
@@ -46,6 +38,14 @@ use dragonfly_client_core::{
     error::{ErrorType, OrErr},
     Error as ClientError, Result as ClientResult,
 };
+use dragonfly_client_metric::{
+    collect_delete_task_failure_metrics, collect_delete_task_started_metrics,
+    collect_download_task_failure_metrics, collect_download_task_finished_metrics,
+    collect_download_task_started_metrics, collect_stat_task_failure_metrics,
+    collect_stat_task_started_metrics, collect_update_task_failure_metrics,
+    collect_update_task_started_metrics, collect_upload_piece_failure_metrics,
+    collect_upload_piece_finished_metrics, collect_upload_piece_started_metrics,
+};
 use dragonfly_client_util::{
     http::{get_range, hashmap_to_headermap, headermap_to_hashmap},
     id_generator::TaskIDParameter,

dragonfly-client/src/grpc/mod.rs

@@ -14,11 +14,11 @@
  * limitations under the License.
  */
 
-use crate::metrics::{
-    collect_prefetch_task_failure_metrics, collect_prefetch_task_started_metrics,
-};
 use dragonfly_api::dfdaemon::v2::DownloadTaskRequest;
 use dragonfly_client_core::{Error as ClientError, Result as ClientResult};
+use dragonfly_client_metric::{
+    collect_prefetch_task_failure_metrics, collect_prefetch_task_started_metrics,
+};
 use std::path::PathBuf;
 use std::time::Duration;
 use tonic::Request;

dragonfly-client/src/lib.rs

@@ -19,7 +19,6 @@ pub mod dynconfig;
 pub mod gc;
 pub mod grpc;
 pub mod health;
-pub mod metrics;
 pub mod proxy;
 pub mod resource;
 pub mod stats;

dragonfly-client/src/proxy/mod.rs

@@ -15,10 +15,6 @@
  */
 
 use crate::grpc::{dfdaemon_download::DfdaemonDownloadClient, REQUEST_TIMEOUT};
-use crate::metrics::{
-    collect_proxy_request_failure_metrics, collect_proxy_request_started_metrics,
-    collect_proxy_request_via_dfdaemon_metrics,
-};
 use crate::resource::{piece::MIN_PIECE_LENGTH, task::Task};
 use bytes::Bytes;
 use dragonfly_api::common::v2::{Download, TaskType};
@@ -29,6 +25,10 @@ use dragonfly_api::errordetails::v2::Backend;
 use dragonfly_client_config::dfdaemon::{Config, Rule};
 use dragonfly_client_core::error::{ErrorType, OrErr};
 use dragonfly_client_core::{Error as ClientError, Result as ClientResult};
+use dragonfly_client_metric::{
+    collect_proxy_request_failure_metrics, collect_proxy_request_started_metrics,
+    collect_proxy_request_via_dfdaemon_metrics,
+};
 use dragonfly_client_util::{
     http::{hashmap_to_headermap, headermap_to_hashmap},
     shutdown,

dragonfly-client/src/resource/piece.rs

@@ -15,16 +15,16 @@
  */
 
 use super::*;
-use crate::metrics::{
-    collect_backend_request_failure_metrics, collect_backend_request_finished_metrics,
-    collect_backend_request_started_metrics, collect_download_piece_traffic_metrics,
-    collect_upload_piece_traffic_metrics,
-};
 use chrono::Utc;
 use dragonfly_api::common::v2::{Hdfs, ObjectStorage, Range, TrafficType};
 use dragonfly_client_backend::{BackendFactory, GetRequest};
 use dragonfly_client_config::dfdaemon::Config;
 use dragonfly_client_core::{error::BackendError, Error, Result};
+use dragonfly_client_metric::{
+    collect_backend_request_failure_metrics, collect_backend_request_finished_metrics,
+    collect_backend_request_started_metrics, collect_download_piece_traffic_metrics,
+    collect_upload_piece_traffic_metrics,
+};
 use dragonfly_client_storage::{metadata, Storage};
 use dragonfly_client_util::id_generator::IDGenerator;
 use leaky_bucket::RateLimiter;

dragonfly-client/src/resource/task.rs

@@ -15,10 +15,6 @@
  */
 
 use crate::grpc::{scheduler::SchedulerClient, REQUEST_TIMEOUT};
-use crate::metrics::{
-    collect_backend_request_failure_metrics, collect_backend_request_finished_metrics,
-    collect_backend_request_started_metrics,
-};
 use dragonfly_api::common::v2::{
     Download, Hdfs, ObjectStorage, Peer, Piece, SizeScope, Task as CommonTask, TaskType,
     TrafficType,
@@ -43,6 +39,10 @@ use dragonfly_client_core::{
     error::{BackendError, DownloadFromParentFailed, ErrorType, OrErr},
     Error, Result as ClientResult,
 };
+use dragonfly_client_metric::{
+    collect_backend_request_failure_metrics, collect_backend_request_finished_metrics,
+    collect_backend_request_started_metrics,
+};
 use dragonfly_client_storage::{metadata, Storage};
 use dragonfly_client_util::{
     http::{hashmap_to_headermap, headermap_to_hashmap},