Fix a bug in the resolution of images in the Docker compat API.
When looking up an image by a short name, the name may match
an image that does not live on Docker Hub. The resolved name
should be used for normalization instead of the input name to
make sure that `busybox` can resolve to `registry.com/busybox`
if present in the local storage.
Fixes: #14291
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Create an auto-update event for each invocation, independent if images
and containers are updated or not. Those events will be indicated in
the events already but users will now know why.
Fixes: #14283
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
The init binary until now has been bind-mounted to /dev/init which
breaks when bind-mounting to /dev. Instead mount the init to
/run/podman-init. The reasoning for using /run is that it is already
used for other runtime data such as secrets.
Fixes: #14251
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
e2e tests tends to fail when running with multiple nodes because
the same device folder name is used accross all nodes
Signed-off-by: Cosmin Tupangiu <cosmin@redhat.com>
* Remove duplicate or unused types and constants
* Move all documetation-only models and responses into swagger package
* Remove all unecessary names, go-swagger will determine names from
struct declarations
* Use Libpod suffix to differentiate between compat and libpod models
and responses. Taken from swagger:operation declarations.
* Models and responses that start with lowercase are for swagger use
only while uppercase are used "as is" in the code and swagger comments
* Used gofumpt on new code
```release-note
```
Signed-off-by: Jhon Honce <jhonce@redhat.com>
- add test
- fix bug when a character device set in a volume as a block device
is seen as block device in _pkg/specgen/generate/kube/volume.go_.
At this stage the type does not matter much because the devices are
recreated at lower layer but the bug allowed a CharDevice volume to be
passed to lower layer as a BlockDevice.
Signed-off-by: Cosmin Tupangiu <cosmin@redhat.com>
Support running `podman play kube` in systemd by exploiting the
previously added "service containers". During `play kube`, a service
container is started before all the pods and containers, and is stopped
last. The service container communicates its conmon PID via sdnotify.
Add a new systemd template to dispatch such k8s workloads. The argument
of the template is the path to the k8s file. Note that the path must be
escaped for systemd not to bark:
Let's assume we have a `top.yaml` file in the home directory:
```
$ escaped=$(systemd-escape ~/top.yaml)
$ systemctl --user start podman-play-kube@$escaped.service
```
Closes: https://issues.redhat.com/browse/RUN-1287
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Use the host UID and host GID mapping when building the local pause
image for a Pod with a custom mapping. Otherwise, the mappings are off
and the build fails. Propagating the mapping to the build container is
not needed since the pause image ships merely a copied `catatonit` from
the host.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=2083997
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Send the main PID only once. Previously, `(*Container).start()` and
the conmon handler sent them ~simultaneously and went into a race.
I noticed the issue while debugging a WIP PR.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Add the notion of a "service container" to play kube. A service
container is started before the pods in play kube and is (reverse)
linked to them. The service container is stopped/removed *after*
all pods it is associated with are stopped/removed.
In other words, a service container tracks the entire life cycle
of a service started via `podman play kube`. This is required to
enable `play kube` in a systemd unit file.
The service container is only used when the `--service-container`
flag is set on the CLI. This flag has been marked as hidden as it
is not meant to be used outside the context of `play kube`. It is
further not supported on the remote client.
The wiring with systemd will be done in a later commit.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
If the RunAsUser, RunAsGroup, SupplementalGroups settings are not
overriden in the container security context, then take the value from
the pod security context.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Commit b58e7e7f11 was not fully rebased before merging and is now
breaking CI because commit 69c479b16e made the underlying error
visible. Using journald inside the container tests is not supported.
Fixes#14162
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
currently tags cause a panic due to an uninitialized map. Initialize the map
and add parsing to make sure we are only tagging with journald
resolves#13356
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
Pulling the K8s pause image seems unnecessarily expensive to me. Let's
use the testgitest_v2s2 one which is under our control and weighs only a
couple of KB.
This cut the execution time in less than half on my machine. Since it's
network bound and I am running on fibre, I expect more significant speed
ups in slower networks.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Squash the --quiet test into another one. The test was more complex
than necessary and can easily be squashed into another one to avoid one
more expensive pull.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Once upon a time, the tests actually pulled from Docker Hub. This has
changed with the rate limits, so we can safely remove the redundant
tests to speed up CI.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Rather than assuming a filesystem path, the API service URI is recorded
in the libpod runtime configuration and then reported as requested.
Note: All schemes other than "unix" are hard-coded to report URI exists.
Fixes#12023
Signed-off-by: Jhon Honce <jhonce@redhat.com>
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Instead of using the main module we should vendor the test tools in a
different directory. That way we do not add extra dependencies to the
main module which can be problemetic for packages or other users.
This is already done in buildah so this makes us more consitent.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
podman system reset did not clean up machines fully, leaving some config
files, and breaking machines. Now it removes all machines files fully.
Signed-off-by: Ashley Cui <acui@redhat.com>
The default log driver is not used when using play kube
without --log-driver. The LogDriver function needs to
be called in order to use the default log driver.
fixes#13781
Signed-off-by: Niall Crowe <nicrowe@redhat.com>
Polish the push and pull benchmarks. In particular, make sure to not be
network bound during these benchmarks by running a local registry and
pushing a local image that can later on be pulled.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
since the network config is a string map, json.unmarshal does not recognize
the config and spec as the same entity, need to map this option manually
resolves#13713
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
We can vendor the test dependencies such as go-md2man, git-validation
and goimports. This allows us to always install the same version as
specified in go.mod. Also we do not rely on a network connection for
this.
The advantage with this method is that dependabot will also update the
dependencies for us and we do not have to hardcode versions in the
Makefile.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Add the notion of an "exit policy" to a pod. This policy controls the
behaviour when the last container of pod exits. Initially, there are
two policies:
- "continue" : the pod continues running. This is the default policy
when creating a pod.
- "stop" : stop the pod when the last container exits. This is the
default behaviour for `play kube`.
In order to implement the deferred stop of a pod, add a worker queue to
the libpod runtime. The queue will pick up work items and in this case
helps resolve dead locks that would otherwise occur if we attempted to
stop a pod during container cleanup.
Note that the default restart policy of `play kube` is "Always". Hence,
in order to really solve #13464, the YAML files must set a custom
restart policy; the tests use "OnFailure".
Fixes: #13464
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
When a container is run in the host network namespace we have to keep
the same resolv.conf content and not use the systemd-resolve detection
logic.
But also make sure we still allow --dns options.
Fixes#14055
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
The files /etc/hosts, /etc/hostname and /etc/resolv.conf should always
be owned by the root user in the container. This worked correct for
/etc/hostname and /etc/hosts but not for /etc/resolv.conf.
A container run with --userns keep-id would have the reolv.conf file
owned by the current container user which is wrong.
Consolidate some common code in a new helper function to make the code more
cleaner.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
The test has been broken since it was added 4 years ago. Instead of
using hardcoded paths we should use tmp files.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
The errcheck linter makes sure that errors are always check and not
ignored by accident. It spotted a lot of unchecked errors, mostly in the
tests but also some real problem in the code.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Two for this error:
invalid indirect of pod.Spec.DNSConfig.Options[0]
...and one for a gofmt error (spaces).
Signed-off-by: Ed Santiago <santiago@redhat.com>
This is a very late followup to my ginkgo-improving work of 2021.
It has been stuck since December because it requires gomega 1.17,
which we've just enabled.
This commit is simply a copy-paste of a command I saved in
my TODO list many months ago:
sed -i -e 's/Expect(\([^ ]\+\)\.\([a-zA-Z0-9]\+\))\.To(Equal(/Expect(\1).To(HaveField(\"\2\", /' test/e2e/*_test.go
Signed-off-by: Ed Santiago <santiago@redhat.com>
add a new option to completely disable xfs quota usage for a volume.
xfs quota set on a volume, even just for tracking disk usage, can
cause weird errors if the volume is later re-used by a container with
a different quota projid. More specifically, link(2) and rename(2)
might fail with EXDEV if the source file has a projid that is
different from the parent directory.
To prevent such kind of issues, the volume should be created
beforehand with `podman volume create -o o=noquota $ID`
Closes: https://github.com/containers/podman/issues/14049
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
I think we forgot to bump the version in the main branch. It should be
v4.1.0-dev now.
Also set the min api version to 4.0.0 as on the podman 4.0 branch.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
We need to use the host network when it is set in the config and
--network was not used.
This regression was added in 3e9af2029f.
Fixes#14015
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
We should not exclude contianers by name. If a users has a container
with the name "inf" it is currently skipped. This is wrong. The k8s yaml
does not contain infra containers so we do not have to skip them.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
The linter ensures a common code style.
- use switch/case instead of else if
- use if instead of switch/case for single case statement
- add space between comment and text
- detect the use of defer with os.Exit()
- use short form var += "..." instead of var = var + "..."
- detect problems with append()
```
newSlice := append(orgSlice, val)
```
This could lead to nasty bugs because the orgSlice will be changed in
place if it has enough capacity too hold the new elements. Thus we
newSlice might not be a copy.
Of course most of the changes are just cosmetic and do not cause any
logic errors but I think it is a good idea to enforce a common style.
This should help maintainability.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Changes as of 2022-04-21:
- apply-podman-deltas: minor cleanup
- buildah-tests.diff: deal with:
. buildah #3894 (the registry one), which affected helpers.bash in
a way that resulted in conflicts here; and
. buildah #3917 (etchosts), which caused offset-only diffs
with no conflicts
- Reevaluate the bud skip list, and reenable some tests that
seems to be passing now under podman:
. bud with specified context ...
. two tests that require a local registry (which buildah now runs)
. bud with --cgroup-parent
Signed-off-by: Ed Santiago <santiago@redhat.com>
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Add more benchmarks for the most common and performance-critical image
commands. Benchmarks for `podman build` should go into a separate
section.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
[skip ci]
While chasing a flake, I discovered that our alpine_nginx
image is broken: it returns 404 on all requests. We never
caught this because--surprise!--curl exits 0 even when
server returns 4xx/5xx status.
Let's be strict: add -f (--fail) option to all invocations
of curl.
And, although I couldn't identify the root cause of the
flake (in "run two containers with the same IP" test),
I can at least fix the broken wait-for-nginx loop, bump
up the number of retries, and improve diagnostics on
failure. And add a strict error-message check.
Signed-off-by: Ed Santiago <santiago@redhat.com>
This reverts commit cc3790f332.
We can't change rootful to rootfull because `rootful` is written into the machine config. Changing this will break json unmarshalling, which will break existing machines.
[NO NEW TESTS NEEDED]
Signed-off-by: Ashley Cui <acui@redhat.com>
Do not use a list of statuses outside of libpod to validate container
statuses. Removing status was never added to the list.
Fixes: https://github.com/containers/podman/issues/13986
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Starting the podman service in debug-mode causes aardvark to run in
debug mode. This does unexpected things with file-descriptors leading
to a test-hang. Thanks to @Luap99 for the fix.
Signed-off-by: Chris Evich <cevich@redhat.com>
golint, scopelint and interfacer are deprecated. golint is replaced by
revive. This linter is better because it will also check for our error
style: `error strings should not be capitalized or end with punctuation or a newline`
scopelint is replaced by exportloopref (already endabled)
interfacer has no replacement but I do not think this linter is
important.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Because /etc/hosts is shared for all containers with a shared network
namespace you should not be able to add hosts from a joined container.
Only the primary netns container can set the hosts.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
When we connect or disconnect from a network we also have to update
/etc/hosts to ensure we only have valid entries in there.
This also fixes problems with docker-compose since this makes use of
network connect/disconnect.
Fixes#12533
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
When we lookup the hostname for a given container we have to check if
the container is joined to another utsns and use this hostname then
instead.
This fixes a problem where the `hostname` command would use the correct
name but /etc/hostname would contain a different name.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Use the new logic from c/common to create the hosts file. This will help
to better allign the hosts files between buildah and podman.
Also this fixes several bugs:
- remove host entries when container is stopped and has a netNsCtr
- add entries for containers in a pod
- do not duplicate entries in the hosts file
- use the correct slirp ip when an userns is used
Features:
- configure host.containers.internal entry in containers.conf
- configure base hosts file in containers.conf
Fixes#12003Fixes#13224
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Following commit ensures that csv escaping is supported while using
inline `--mount=type=......` flag with `podman run` by using
`encoding/csv` to parse options instead of performing a `split.String(`
by `comma`.
Closes: https://github.com/containers/podman/issues/13922
Signed-off-by: Aditya R <arajan@redhat.com>
When run on an F36 host using netavark/aardvark-dns, for whatever
underlying reason most checkpoint/restore tests are emitting an error
similar to:
`criu: Symbol `__rseq_offset' has different size in shared object,
consider re-linking`
This extraneous output is causing the basic checkpoint system test to
fail. Since, all other testing of checkpoint/restore feature is
passing (also with the extraneous message) loosen the system test
sensitivity to match.
Signed-off-by: Chris Evich <cevich@redhat.com>
We are inconsistent on the name, we should stick with rootfull.
[NO NEW TESTS NEEDED] Existing tests should handle this and no tests for
machines exists yet.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
From a security point of view, it would be nice to be able to map a
rootless usernamespace that does not use your own UID within the
container.
This would add protection against a hostile process escapping the
container and reading content in your homedir.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
It seems this breaks older version of `podman-remote` users hence it
looks like this patch would be a better candidate for podman `5.0`
Problem
* Client with `4.0` cannot interact with a server of `4.1`
Plan this patch for podman `5.0`
This reverts commit 0cebd158b6.
Signed-off-by: Aditya R <arajan@redhat.com>
On F36 / podman 4, at the time of this commit there is no
`/etc/containers/storage.conf` installed by default. Since the
test volume-mounts this file into the container, it was failing. Fix
this by using a conditional volume-mount based on the file existing (or
not).
Signed-off-by: Chris Evich <cevich@redhat.com>
When this option was added to the e2e tests, there was no CI Automation
support for running remote tests w/ netavark. When added, many
e2e test errors/failures are generated due to this option not being
valid for the remote client. Fix this in the tests by conditionally
adding the option if the test is running the remote client.
Signed-off-by: Chris Evich <cevich@redhat.com>
Problem: the system test 'is()' checker was poorly thought out.
For example, there is no way to check for inequality or for
absence of a substring.
Solution, step 1: introduce new assert(), copied almost verbatim
from buildah, where it has been successful in addressing the
gaps in is().
The logical next step is to search the tests for 'die' and
for 'run', looking for negative assertions which we can
replace with assert(). There were a lot, and in the process
I found a number of ugly bugs in the tests themselves. I've
taken the liberty of fixing these.
Important note: at this time we have both assert() and is().
Replacing all instances of is() would be impossible to review.
Signed-off-by: Ed Santiago <santiago@redhat.com>
The patch introduces the following test cases:
1. An attempt to checkpoint a container that does not exist should fail.
2. Checkpoint of a running container with --create-image should create a
checkpoint image.
3. A single checkpoint image can be used to restore multiple containers,
each with a different name.
4. Restoring multiple containers from checkpoint images with a single
restore command.
Signed-off-by: Radostin Stoyanov <radostin@redhat.com>
Add a proof of concept for benchmarking Podman. The benchmarks are
implemented by means of the end-to-end test suite but hidden behind
a `benchmarks` build tag. Running `make localbenchmarks` will run
`test/e2e` with the specific build tag and set ginkgo's "focus" to
the specific "Podman Benchmark Suite" to only run this spec and skip
all others.
ginkgo will print a report before terminating listing the CPU and memory
stats for each benchmark. New benchmarks can easily be added via the
`newBenchmark` function that also supports adding an `init()` function
to each benchmark which allows for performing certain setups for the
specific benchmark. For instance, benchmarking `podman start` requires
creating a container beforehand.
Podman may be called more than once in the main function of a benchmark
but note that the displayed memory consumption is then a sum of all
Podman invocations. The memory consumption is collected via
`/usr/bin/time`.
A benchmark's report is split into CPU and memory as displayed below:
```
[CPU] podman images:
Fastest Time: 0.146s
Slowest Time: 0.187s
Average Time: 0.180s ± 0.015s
[MEM] podman images:
Smallest: 41892.0KB
Largest: 42792.0KB
Average: 42380.7KB ± 286.4KB
```
Note that the benchmarks are not wired into the CI yet. They are meant
as a proof of concept. More benchmarks and the plumbing into CI will
happen in a later change.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
in specgen, CLI path uses the given memory limit to define the swap value (if not already specified)
add a route to this piece of code from within the api handlers
resolves#13145
Signed-off-by: cdoern <cdoern@redhat.com>
`DefineCreateFlags` was excluding clone from using the memory-swappiness flag leading the value to be zero
when our deafult is -1. Rearrange the if/else to give clone these memory related options
resolves#13856
Signed-off-by: cdoern <cdoern@redhat.com>
It allows to customize the entry that is written to the `/etc/passwd`
file when --passwd is used.
Closes: https://github.com/containers/podman/issues/13185
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
This reverts commit e133a06d2f.
@nalind found a proper fix in c/storage [1] to address the performance
issue. So we really don't need the flag anymore. Note the flag has
never made it into any release.
[1] d76b3606fc
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Add new functions to logfile.go for rotating and truncating
the events log file once the log file and its contents
exceed the maximum size limit while keeping 50% of the
log file's content
Also add tests to verify log rotation and truncation
Signed-off-by: Niall Crowe <nicrowe@redhat.com>
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Current error messages are really awful, and cause great
suffering every time someone adds a new subcommand. Let's
see if these are better.
Signed-off-by: Ed Santiago <santiago@redhat.com>
--tail=1 is not working f you restart a container with journald logging.
We see the exit status and then call into the logging a second time
causing all of the logs to print.
Removing the tail log on exited seems to fix the problem.
Fixes: https://github.com/containers/podman/issues/13098
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
For better docker compatibility we should use the bridge network mode as
default for rootless. This was already done previously but commit
535818414c introduced this regression in v4.0.
Since the apiv2 test are only run rootful we cannot catch this problem
in CI.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Bad code got committed by accident: test description on run_podman
line, not test line.
Did not seem to affect tests, but fix it anyway.
Signed-off-by: Ed Santiago <santiago@redhat.com>
`--mount` should allow setting driver specific options using
`volume-opt` when `type=volume` is set.
This ensures parity with docker's `volume-opt`.
Signed-off-by: Aditya R <arajan@redhat.com>
The API endpoints should properly honour the `no_hosts=true` setting in
containers.conf.
Fixes#13719
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Add a new flag to set the ipam-driver. Also adds a new ipam driver none
mode which only creates interfaces but does not assign addresses.
Fixes#13521
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Convert container entrypoint from string to an array inorder to make
sure there is parity between `podman inspect` and `docker inspect`
Signed-off-by: Aditya R <arajan@redhat.com>
Add a --size option to podman images to allow for disabling computing
the size of listed images. If listing images is critical to
performance, user may chose to turn off size computation to speed things
up.
Context: #13755
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
This fixes#13756.
All the mechanics to create anonymous volumes is already present, but
there's still a validation preventing that path from being taken. We
remove the validation, which allows the volume to be created
successfully.
Signed-off-by: Andrew Aylett <andrew@aylett.co.uk>
--cap-add is useful when running a privileged container with UID != 0,
so that individual capabilities can be added to the container process.
Closes: https://github.com/containers/podman/issues/13449
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Ensure a directory added to .containerignore on client is not included
in tar sent to remote podman API service
* Clean up podman invocations to not include duplicate --remote and
--url flags
* Use pkill vs. pgrep when cleaning up podman API service in tests
* Add exit code when logging error when testing
Closes#13535
Signed-off-by: Jhon Honce <jhonce@redhat.com>
When you run podman-remote unsahre for example you currently get:
Error: unrecognized command `podman-remote unshare`
This is because we do not add the command to the cobra tree when we run
in remote mode. However this is a bad user experience since it is not
clear that the command is only supported for local podman. Users are
left wondering why this does not work and could think the documentation
is wrong.
To fix it we add a clear error message:
Error: cannot use command "podman-remote unshare" with the remote podman client
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
`podman play kube` creates a new volume for configmap, if same configmap
is applied again volume can be re-used, there is no need to remove and
re-create the volume again
Signed-off-by: Aditya R <arajan@redhat.com>
Since netavark does not support dhcp yet we have to create a macvlan
network with a static subnet. Since we do not use the network to run
containers the actual subnet does not matter.
Also stop hard coding the network id. Unlike the cni backend the
netavark backend creates random ids that are not predictable.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
- removed: v1.9.0, v2.0.6
+ added: v3.4.0
(Cannot add v4 because there's no such image on quay. As soon
as one appears, we should add it.)
Add a workaround for a UTS namespace conflict new in v3.4
Signed-off-by: Ed Santiago <santiago@redhat.com>
The slirp4netns port forwarder was not updated to make use of the new
port format. This results in a problem when port ranges are used since
it does not read the range field from the port.
Update the logic to iterate through all ports with the range and
protocols. Also added a system test for port ranges with slirp4netns,
rootlesskit and the bridge network mode.
Fixes#13643
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
the infra Inherit function was not properly passing pod volume information to new containers
alter the inherit function and struct to use the new `ConfigToSpec` function used in clone
pick and choose the proper entities from a temp spec and validate them on the spegen side rather
than passing directly to a config
resolves#13548
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
Signed-off-by: cdoern <cdoern@redhat.com>
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
With podman4 we support netavark, however old versions will still use
cni. Since netavark and cni can conflict we should not mix them.
Remove the network setup from the inital podman command and create the
directories manually to prevent such conflicts.
Also the update to 4.0 changes the network db structure. While it is
compatible from 3.X to 4.0 it will fail the other way around. In this
test it will happen because the cleanup process still uses the old
podman while the network connect/disconnect test already changed the db
format. Therefore the cleanup process cannot see any networks and will
not tear it down. The following start will fail because the ip address
is already assigned.
Fixes#13679
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Following commit ensures that podman generates a valid event on `podman
container rename` where event specifies that it is a rename event and
container name swtichted to the latest name.
Signed-off-by: Aditya R <arajan@redhat.com>
WalkDir should be faster the Walk, since we often do
not need to stat files.
[NO NEW TESTS NEEDED] Existing tests should find errors.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
When podman gets an error it prints out "Error: " before
printing the error string. If the error message starts with
error, we end up with
Error: error ...
This PR Removes all of these stutters.
logrus.Error() also prints out that this is an error, so no need for the
error stutter.
[NO NEW TESTS NEEDED]
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
systemd expects the container_uuid environment variable be set
when it is running in a container.
Fixes: https://github.com/containers/podman/issues/13187
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The kernel never sets the inheritable capabilities for a process, they
are only set by userspace. Emulate the same behavior.
Closes: CVE-2022-27649
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Allows users to set `--os` , `--arch` and `--variant` of the image
created from the custom import.
Following is useful when user is already aware of the values which are
correct for their generated rootfs
Signed-off-by: Aditya R <arajan@redhat.com>
podman build fails on remote build when using a relative context directory.
This is because the context dir was not being added to the tar, so when remote
the compat build function would not be able to stat the contextDir.
resolves#13293
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
Still an unknown number remains but I am running out of patience.
Adding dots is not the best use of my time.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Make sure we're waiting for the ls container to finish to prevent
potential flakes or future regressions.
Spotted while enabling a linter.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Containers started with `--no-healthcheck` are configured to contain no
healthcheck and test configured as `NONE`. Podman shows wrong status as
such use cases.
Following commit fixes the faulty behavior of stauts field for
containers started with `--no-healthcheck`
Signed-off-by: Aditya R <arajan@redhat.com>
The calculate the percentage we need floating point numbers. The current
code however casted the result of reclaimable/size to an int first.
Casting to an int in go will just discard the decimal points, thus the
result was either 0 or 1 so if multiplied by 100 it would show up as 0%
or 100%.
To fix this we have to multiply by 100 first before casting the result
to an int. Also add a check for div by zero which results in NaN and use
math.Round() to correctly round a number.
Ref #13516
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
One of the pod tests was leaving a stray image behind,
causing scary red warnings in CI logs. Clean that up.
Also, now that #13541 has merged, use 'rmi --ignore' instead of
ignoring exit status from rmi
Signed-off-by: Ed Santiago <santiago@redhat.com>
Add an `--ignore` flag to `podman image rm` to instruct ignoring image
if a specified image does not exist and to not throw an error. Other
commands (e.g., `podman container rm`) already support this flag.
Such an `--ignore` flag can come in handy in clean-up scripcts such as
the teardown phases in the Podman tests.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Update the login tests to reflect the latest changes to allow http{s}
prefixes (again) to address bugzilla.redhat.com/show_bug.cgi?id=2062072.
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
Compose v2 uses dashes as separators instead of hyphens. This
broke some tests that relied upon container names. Set the name
conditionally to make it safe for both.
Signed-off-by: Matthew Heon <mheon@redhat.com>
Make sure to ignore local {container,docker}ignore files when building a
local pause image. Otherwise, we may mistakenly not be able to copy
catatonit into the build container.
Fixes: #13529
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
When podman generate systemd is invoked, it previously did not check if
container-prefix or pod-prefix are empty. When these are empty, the file name
starts with the separator, which is hyphen by default. This results in files
like '-containername.service'.
The code now checks if these prefixes are empty. If they are, the filename no
longer adds a separator. Instead, it uses name or ID of the container or pod.
Closes#13272
Signed-off-by: Nirmal Patel <npate012@gmail.com>
* systemctl stop podman.service will now return exit code 0
* Update test framework to support JSON boolean and numeric values
Signed-off-by: Jhon Honce <jhonce@redhat.com>
There is a lot of unneeded code, k8s is the by far the biggest
dependency in podman. We should remove as much as possible so that we
only have the stuff left that we use.
This is just a quick skim over the code which removes a lot of the
generated code and many packages that are now unused.
I know that this will be impossible to properly review. I will try to
make smaller changes in follow up work.
Right now this reduces about 8 MB in binary size!!!
[NO NEW TESTS NEEDED] Hopefully existing tests will catch any problems.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
We only need a small part of the k8s dependencies but they are the
biggest dependencies in podman by far. Moving them into podman allows us
to remove the unnecessary parts.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
All podman tests in CI expects exit code 125, which might not be true since exit code from runtime
is relayed as it is without any modification both in `buildah` and `podman`. Following behviour is seen when PR https://github.com/containers/buildah/pull/3809
added a test here https://github.com/containers/buildah/blob/main/tests/bud.bats#L3183 which relays exit code from runtime as it
is, in case of both `podman` and `buildah`. However apart from this test case no other test case was able to trigger this behviour hence
marking this test as an anomaly. Since its debatable if we should override this returned error number or not hence adding a note here.
Signed-off-by: Aditya R <arajan@redhat.com>
Bumps
c/buildah to -> `v1.24.3-0.20220310160415-5ec70bf01ea5`
c/storage to -> `v1.38.3-0.20220308085612-93ce26691863`
c/image to -> `v5.20.1-0.20220310094651-0d8056ee346f`
Signed-off-by: Aditya R <arajan@redhat.com>
Allow users to add annotions in the podman play kube command.
This PR Also fixes the fact that annotations in the pod spec were
not being passed down to containers.
Fixes: https://github.com/containers/podman/issues/12968
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The permissions on disk were wrong since we were not converting to
octal.
Fixes: https://github.com/containers/podman/issues/13108
[NO NEW TESTS NEEDED] Since we don't currently test using the docker
client
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
pprof tests are way too flaky, and are causing problems for
community contributors who don't have privs to press Re-run.
There has been no activity or interest in fixing the bug,
and it's not something I can fix. So, just disable the test.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Secret-verify-leak is causing flakes, when running in parallel tests.
This is because remote secrets are copied into the context directory to
send to the API server, and secret-verify-leak is doing a COPY * and
then checking if the temporary secret file ends up in the container or
not. Since all the temporary files are prefixed with
"podman-build-secret", this test checks if podman-build-secret is in the
image. However, when run in parallel with other tests, other temporary
podman-build-secrets might be in the context dir. Moving
secret-verify-leak into its own directory makes sure that the context
dir is used only by this one test.
Also renamed Dockerfile -> Containerfile and cleaned up unused
Containerfiles.
Signed-off-by: Ashley Cui <acui@redhat.com>
The default rule should be set at the head of device configuration.
Otherwise, rules for user devices are overridden by the default rule so
that any access to the user devices are denied.
Signed-off-by: Hironori Shiina <shiina.hironori@jp.fujitsu.com>
Error out if the kube yaml passed to play kube has more
than one container or init container with the same name.
Signed-off-by: Urvashi Mohnani <umohnani@redhat.com>
* Add configuration to add report header for python client used in tests
* Move report headers into the individual test runners vs runner.sh
Signed-off-by: Jhon Honce <jhonce@redhat.com>
do not attempt to use cgroups with pods if the cgroups are disabled.
A similar check is already in place for containers.
Closes: https://github.com/containers/podman/issues/13411
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
While resolving `workdir` we mostly create a `workdir` when `stat`
fails with `ENOENT` or `ErrNotExist` however following cases are not
true when user explicitly specifies a `workdir` while `running` using
`--workdir` which tells `podman` to only use workdir if its exists on
the container. Following configuration is implicity set with other
`run` mechanism like `podman play kube`
Problem with explicit `--workdir` or similar implicit config in `podman play
kube` is that currently podman ignores the fact that workdir can also be
a `symlink` and actual `link` could be valid.
Hence following commit ensures that in such scenarios when a `workdir`
is not found and we cannot create a `workdir` podman must perform a
check to ensure that if `workdir` is a `symlink` and `link` is resolved
successfully and resolved link is present on the container then we
return as it is.
Docker performs a similar behviour.
Signed-off-by: Aditya R <arajan@redhat.com>
* Add which python client is being used to run tests, see "python
client" below.
* Remove redundate code from test classes
* Update/Add comments to modules and classes
======================================================= test session starts ========================================================
platform linux -- Python 3.10.0, pytest-6.2.4, py-1.10.0, pluggy-0.13.1
python client -- DockerClient
rootdir: /home/jhonce/Projects/go/src/github.com/containers/podman
plugins: requests-mock-1.8.0
collected 33 items
test/python/docker/compat/test_containers.py ...s.............. [ 54%]
test/python/docker/compat/test_images.py ............ [ 90%]
test/python/docker/compat/test_system.py ... [100%]
Note: Follow-up PRs will verify the test results and expand the tests.
Signed-off-by: Jhon Honce <jhonce@redhat.com>
This option was requested so that users could specify alternate
locations to find context directories for each image build. It
requites the --build option to be set.
Partion Fix: https://github.com/containers/podman/issues/12485
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Allow users to commit containers into a single layer.
Usage
```bash
podman container commit --squash <name>
```
Signed-off-by: Aditya R <arajan@redhat.com>
When removing a container created with a --volumes-from a container
created with a built in volume, we complain if the original container
still exists. Since this is an expected state, we should not complain
about it.
Fixes: https://github.com/containers/podman/issues/12808
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
When a test which creates a network fail it will not remove the network.
The teardown logic should remove the networks. Since there is no --all
option for network rm we use network prune --force.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
`podman play kube` tries to build images even if `--build` is set to
false so lets honor that and make `--build` , `true` by default so it
matches the original behviour.
Signed-off-by: Aditya R <arajan@redhat.com>
Add a extra `See 'podman command --help'` to the error output.
With this patch you now get:
```
$ podman run -h
Error: flag needs an argument: 'h' in -h
See 'podman run --help'
```
Fixes#13082Fixes#13002
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
We're running into problems that are impossible to diagnose
because we have no idea if the SUT is using netavark or CNI.
We've previously run into similar problems with runc/crun,
or cgroups 1/2.
This adds a one-line 'echo' with important system info. Now,
when viewing a full test log, it will be possible to view
system settings in one glance.
Signed-off-by: Ed Santiago <santiago@redhat.com>
podman container clone takes the id of an existing continer and creates a specgen from the given container's config
recreating all proper namespaces and overriding spec options like resource limits and the container name if given in the cli options
this command utilizes the common function DefineCreateFlags meaning that we can funnel as many create options as we want
into clone over time allowing the user to clone with as much or as little of the original config as they want.
container clone takes a second argument which is a new name and a third argument which is an image name to use instead of the original container's
the current supported flags are:
--destroy (remove the original container)
--name (new ctr name)
--cpus (sets cpu period and quota)
--cpuset-cpus
--cpu-period
--cpu-rt-period
--cpu-rt-runtime
--cpu-shares
--cpuset-mems
--memory
--run
resolves#10875
Signed-off-by: cdoern <cdoern@redhat.com>
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
Signed-off-by: cdoern <cdoern@redhat.com>
The CONTAINERS_CONF environment variable can be used to override the
configuration file, which is useful for testing. However, at the moment
this variable is not propagated to conmon. That means in particular, that
conmon can't propagate it back to podman when invoking its --exit-command.
The mismatch in configuration between the starting and cleaning up podman
instances can cause a variety of errors.
This patch also adds two related test cases. One checks explicitly that
the correct CONTAINERS_CONF value appears in conmon's environment. The
other checks for a possible specific impact of this bug: if we use a
nonstandard name for the runtime (even if its path is just a regular crun),
then the podman container cleanup invoked at container exit will fail.
That has the effect of meaning that a container started with -d --rm won't
be correctly removed once complete.
Fixes#12917
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This comment refers to overiding $PODMAN although the code below does
nothing of the sort. Presumbly the comment has been outdated by altering
the containers.conf / $CONTAINERS_CONF instead.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We could remove the container running the volume plugins, before
the containers using the volume plugins; this could cause
unmounting the volumes to fail because the plugin could not be
contacted.
Signed-off-by: Matthew Heon <mheon@redhat.com>
It looks like some descriptions have changed on the docker registry
where we had been searching for images that include 'alpine'. We are
now seeing an image in the initial list that has 'alpine' in its
description.
Signed-off-by: Brent Baude <bbaude@redhat.com>
For the since and after imagve filter tests, instead of using the
read-only cache of images, we just use the empty r/w store. We then
build three images that are strictly predictable.
Signed-off-by: Brent Baude <bbaude@redhat.com>
Checkpoint/restore pod tests are not running with an older runc and now
that runc 1.1.0 appears in the repositories it was detected that the
tests were failing. This was not detected in CI as CI was not using runc
1.1.0 yet.
Signed-off-by: Adrian Reber <areber@redhat.com>
The `podman network connect` and `podman network disconnect`
commands give containers access to different networks than the
ones they were created with; these networks can also have DNS
servers associated with them. Until now, however, we did not
modify resolv.conf as network membership changed.
With this PR, `podman network connect` will add any new
nameservers supported by the new network to the container's
/etc/resolv.conf, and `podman network disconnect` command will do
the opposite, removing the network's nameservers from
`/etc/resolv.conf`.
Fixes#9603
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
* Ensure meaningful behaviour when called with /v3.x.x semantics
* Change return code to 409 from 500 when client attempts to use an
existing network name
* Update API bats test runner to support /v4.0.0 endpoints by default
Signed-off-by: Jhon Honce <jhonce@redhat.com>
When running podman inside systemd user units, it is possible that
systemd kills the rootless netns slirp4netns process because it was
started in the default unit cgroup. When the unit is stopped all
processes in that cgroup are killed. Since the slirp4netns process is
run once for all containers it should not be killed. To make sure
systemd will not kill the process we move it to the user.slice.
Fixes#13153
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Lot of clients are expecting proper `Content-type: application/json`
configured in response headers of `/build` compat api. Following commit
fixes that.
Fixes issues where code is setting header field after writing header
which is wrong. We must set `content-type` before we write and flush
http header.
Signed-off-by: Aditya R <arajan@redhat.com>
All the healthcheck return output now but systems tests is written to
expect empty output which seems wrong.
Modify jq output to contain newline character rather than actual newline
Signed-off-by: Aditya R <arajan@redhat.com>
It seems we are ignoring output from healthcheck session.
Open a valid pipe to healthcheck session in order read its output.
Use common pipe for both `stdout/stderr` since that was the previous
behviour as well.
Signed-off-by: Aditya R <arajan@redhat.com>
Append the podman dns seach domain to the host search domains when we
use the dnsname/aardvark server. Previously it would only use podman
seach domains and discard the host domains.
Fixes#13103
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
separated cgroupNS sharing from setting the pod as the cgroup parent,
made a new flag --share-parent which sets the pod as the cgroup parent for all
containers entering the pod
remove cgroup from the default kernel namespaces since we want the same default behavior as before which is just the cgroup parent.
resolves#12765
Signed-off-by: cdoern <cdoern@redhat.com>
Signed-off-by: cdoern <cbdoer23@g.holycross.edu>
Signed-off-by: cdoern <cdoern@redhat.com>
podman system prune should also remove all networks. When we want to
users to migrate to the new network stack we recommend to run podman
system reset. However this did not remove networks and if there were
still networks around we would continue to use cni since this was
considered an old system.
There is one exception for the default network. It should not be removed
since this could cause other issues when it no longer exists. The
network backend detection logic ignores the default network so this is
fine.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This reverts commit a1bc8cb52c.
Please see resolv.conf(5) search domains must be on the same line. If
you use multiple seach key words only the last one is used. I tested this
with alpine and it works correctly when they are on the same line so I
am not sure what issues Dan had with it but this is not correct.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Alpine does not seem to use search correctly when there are multiple
search domains on the same line. It only uses the first with the advent.
When podman runs within a separate network we are appending on
dns.podman as a search, if you add a search domain, then this causes the
local search on network to fail.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
these mount flags are already used for the /dev/shm mount on the host,
but they are not set for the bind mount itself.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
by default slirp4netns uses the tap0 device. When slirp4netns is
used, use that device by default instead of eth0.
Closes: https://github.com/containers/podman/issues/11695
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Often users want their overlayed volumes to be `non-volatile` in nature
that means that same `upper` dir can be re-used by one or more
containers but overall of nature of volumes still have to be `overlay`
so work done is still on a overlay not on the actual volume.
Following PR adds support for more advanced options i.e custom `workdir`
and `upperdir` for overlayed volumes. So that users can re-use `workdir`
and `upperdir` across new containers as well.
Usage
```console
$ podman run -it -v myvol:/data:O,upperdir=/path/persistant/upper,workdir=/path/persistant/work alpine sh
```
Signed-off-by: Aditya R <arajan@redhat.com>
podman network create --subnet, --gateway and --ip-range can now be
specified multiple times to join the network to more than one subnet.
This is very useful if you want to use a dual stack network and assign a
fixed ipv4 and ipv6 subnet. The order of the options is important here,
the first --gateway/--ip-range will be assigned to the first subnet and
so on.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Bumps Buildah to v1.24.0 and adopts the new values for pull:
true, false, never, and always. The pull-never and pull-always options
for the build command are still usable, but they have been removed from
the man page documentation with this change.
Signed-off-by: tomsweeneyredhat <tsweeney@redhat.com>
machinectl does not propogate error messages and adds extra lines in the output, exec.Cmd is able to clear the env besides PATH and TERM,
and use the given UID and GID to execute the command properly.
machinectl is still used to create a user session. Ubuntu support is limited by this.
Signed-off-by: cdoern <cdoern@redhat.com>
Minimal: only test Fedora.Latest.
Reason: podman 4.0.0-0.2.rc2 broke bodhi gating tests:
https://github.com/containers/podman/issues/12989
Requires skipping two recently-added tests that use 'podman unshare',
which doesn't work on remote.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Before this PR, the podman --help command shows the defaults
as runc and overlay even if the storage.conf and containers.conf
files do not match. This PR changes them to show the actual defaults
and in the case of storage driver, does not show the default at all.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
We should not check if the network supports dns when we create a
container with network aliases. This could be the case for containers
created by docker-compose for example if the dnsname plugin is not
installed or the user uses a macvlan config where we do not support dns.
Fixes#12972
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This PR fixes nil pointer dereference for configmap optional parameter.
When optional parameter is not passed, the code tried to acces the
parameter which caused nil pointer dereference.
Signed-off-by: Ondra Machacek <omachace@redhat.com>
Amongst other things 030-run.bats checks for sensible error messages when
attempting a "podman run" with a non-existent or inaccessible path. It
checks for these messages, which come from the low-level runtime, in a lot
of detail, including separate versions for runc and crun. This is fragile
in several ways:
* It's likely to fail if using a runtime other than crun or runc
* It relies on detecting whether the runtime is crun vs. runc using the
path, which could fail if the binary has been named something unusual
* It will break if crun or runc ever alter their error message (even if
it's just changing case)
This replaces the checked versions with a much more accepting regex which
will work for both the runc and crun messages, while still looking for the
essential pieces. This isn't guaranteed to work with other runtimes, but
it's much more likely to.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Amongst other things 005-info.bats tests that the host.ociRuntime.package
field in the output from "podman info" reports something containing "crun"
or "runc".
While those are obviously the dominant runtimes in practice, it seems
undesirable to explicitly exclude them in testing. So, remove this
specific test.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
`http:/host:port/images/load` fails to accept tar with more than one
images however manual load works as expected. Remove explicit check for
`1` image and only fail if result set has value less than `1`.
Signed-off-by: Aditya R <arajan@redhat.com>
systemd 250 on f36 has changed the output format of a message
our tests rely on. This breaks bodhi gating tests. Fix test
so it accepts a continually-growing set of systemd outputs.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Heuristic to initialize TunnelMode/remote podman:
- Podman built with remote tag
- Podman running on darwin or windows GOOS
- CONTAINER_HOST or CONTAINER_CONNECTION set in environment
- --remote flag given on command line
- From containers.conf, Engine.Remote == true and GOOS == linux
Otherwise, podman will run in ABIMode/linked against libpod library.
Fixes#12866
Signed-off-by: Jhon Honce <jhonce@redhat.com>
This commit includes:
* Handlers for generate systemd unit
with manually defined dependencies such as:
Wants=, After= and Requires=
* The new unit and e2e tests for checking generated systemd units
for container and pod with custom dependencies
* Documented descriptions for custom dependencies options
Signed-off-by: Eugene (Evgenii) Shubin <esendjer@gmail.com>
Valentin Rothberg
Cosmin Tupangiu
Jhon Honce
OpenShift Merge Robot
Aditya R
Daniel J Walsh
Giuseppe Scrivano
Paul Holzinger
cdoern
Ashley Cui
Niall Crowe
Ed Santiago
Chris Evich
aonoa
Radostin Stoyanov
Andrew Aylett
John Matthews
gcalin
jason
Giuseppe Scrivano
Matthew Heon
Nirmal Patel
LStandman
Lokesh Mandvekar
😎 Mostafa Emami
Hironori Shiina
Urvashi Mohnani
Xueyuan Chen
cdoern
Romain Geissler
David Gibson
Brent Baude
Adrian Reber
Matthew Heon
Yaron Dayagi
tomsweeneyredhat
Ondra Machacek
esendjer