We can't mount sysfs as rootless unless we manage the network
namespace. Problem: slirp4netns is now creating and managing a
network namespace separate from the OCI runtime, so we can't
mount sysfs in many circumstances. The `crun` OCI runtime will
automatically handle this by falling back to a bind mount, but
`runc` will not, so we didn't notice until RHEL gating tests ran
on the new branch.
Signed-off-by: Matthew Heon <mheon@redhat.com>
* Add a new `pkg/copy` to centralize all container-copy related code.
* The new code is based on Buildah's `copier` package.
* The compat `/archive` endpoints use the new `copy` package.
* Update docs and an several new tests.
* Includes many fixes, most notably, the look-up of volumes and mounts.
Breaking changes:
* Podman is now expecting that container-destination paths exist.
Before, Podman created the paths if needed. Docker does not do
that and I believe Podman should not either as it's a recipe for
masking errors. These errors may be user induced (e.g., a path
typo), or internal typos (e.g., when the destination may be a
mistakenly unmounted volume). Let's keep the magic low for such
a security sensitive feature.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
Docker defines an option of "default" which means to
use the default network. We should support this with
the same code path as --network="".
This is important for compatibility with the Docker API.
Fixes: https://github.com/containers/podman/issues/8544
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Add the mask and unmask option to the --security-opt flag
to allow users to specify paths to mask and unmask in the
container. If unmask=ALL, this will unmask all the paths we
mask by default.
Signed-off-by: Urvashi Mohnani <umohnani@redhat.com>
The network ID is not stored. It is just the sha256 hash from
the network name. There is a risk of a potential hash collision.
However it's very unlikely and even if we hit this it will
complain that more than network with this ID exists.
The main benefit is that the compat api can have proper
network ID support. Also this adds the support for
`podman network ls --format "{{.ID}}"` and `--filter id=<ID>`.
It also ensures that we can do network rm <ID> and network
inspect <ID>.
Since we use a hash this commit is backwards compatible even for
already existing networks.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
As described in issue #8507 this commit contains a breaking
change which is not wanted in v2.2.
We can discuss later if we want this in 3.0 or not.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Add label support for podman network create. Use the `args`
field in the cni config file to store the podman labels.
Use `podman_labels` as key name and store the labels as
map[string]string.
For reference: https://github.com/containernetworking/cni/blob/master/CONVENTIONS.md#args-in-network-confighttps://github.com/containernetworking/cni/blob/spec-v0.4.0/SPEC.md#network-configuration
Example snippet:
```
...
"args": {
"podman_labels": {
"key1":"value1",
"key2":"value2"
}
}
...
```
Make podman network list support several filters. Supported filters are name,
plugin, driver and label. Filters with different keys work exclusive. Several label
filters work exclusive and the other filter keys are working inclusive.
Also adjust the compat api to support labels in network create and list.
Breaking changes:
- podman network ls -f shortform is used for --filter instead --format
This matches docker and other podman commands (container ps, volume ps)
- libpod network list endpoint filter parameter is removed. Instead the
filters paramter should be used as json encoded map[string][]string.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
In k8s a persistent volume claim (PVC) allow pods to define a volume
by referencing the name of a PVC. The PVC basically contains criterias
that k8s then use to select which storage source it will use for the
volume.
Podman only provide one abtracted storage, the named volumes, and
create them if they don't exists yet. So this patch simply use a
volume with the name of the PVC.
Signed-off-by: Alban Bedel <albeu@free.fr>
Replace the simple map of names to paths with a map of names to a struct
to allow passing more parameters. Also move the code to parse the volumes
to its own file to avoid making the playKubePod() function overly complex.
Finally rework the kube volumes test to also be ready to support more
volume types.
Signed-off-by: Alban Bedel <albeu@free.fr>
First, make sure we are only trying to remove the network
interface if we are root.
Second, if we cannot get the interface name (e.g macvlan config)
then we should not fail. Just remove the config file.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
For docker compatibility we need to support --platform
flag.
podman create --platform
podman run --platform
podman pull --platform
Since we have --override-os and --override-arch already
this can be done just by modifying the client to split
the --platform call into os and arch and then pass those
options to the server side.
Fixes: https://github.com/containers/podman/issues/6244
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The cni plugin `tuning` is required to set a custom mac address.
This plugin is configured in the default cni config file which is
packaged with podman but was not included the generated config form
`podman network create`.
Fixes#8385
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
This command exists in docker and is also in our documentation.
Also remove mentions of `podman ls` or `podman list`. These
commands do not exists in podman or docker.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
We allow a container to be connected to several cni networks
but only if they are listed comma sperated. This is not intuitive
for users especially since the flag parsing allows multiple string
flags but only would take the last value. see: spf13/pflag#72
Also get rid of the extra parsing logic for pods. The invalid options
are already handled by `pkg/specgen`.
A test is added to prevent a future regression.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Currently we don't document which end of the podman-remote client server
operations uses the containers.conf. This PR begins documenting this
and then testing to make sure the defaults follow the rules.
Fixes: https://github.com/containers/podman/issues/7657
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
this enables the ability to connect and disconnect a container from a
given network. it is only for the compatibility layer. some code had to
be refactored to avoid circular imports.
additionally, tests are being deferred temporarily due to some
incompatibility/bug in either docker-py or our stack.
Signed-off-by: baude <bbaude@redhat.com>
Filters with the same key work inclusive with the only exception being
`label` which is exclusive. Filters with different keys always work exclusive.
Also update the documentation with the new behavior.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Add some more tests, document cases where remote will not work
Add FIXMEs for tests that should work on podman-remote but currently
do not.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Add some more tests, document cases where remote will not work
Add FIXMEs for tests that should work on podman-remote but currently
do not.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
All of our filters worked exclusive resulting in `--filter status=created --filter status=exited` to return nothing.
In docker filters with the same key work inclusive with the only exception being `label` which is exclusive. Filters with different keys always work exclusive.
This PR aims to match the docker behavior with podman.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
The test were working fine. Just my IDE doesn't like the files
without the suffix and I don't like red errors.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Having play kube start the pod is not always appropriate, one might
for example like to have the pod running as a set of systemd services.
Add a `start` option to the command line and API to control if the pod
should be started or not; it defaults to true for backward
compatibility.
Signed-off-by: Alban Bedel <albeu@free.fr>
Remove references on contrib/gate. The gating image is currently not
maintained and was replaced in the CI rewrite. We can disable builds
in Quay once the change is merged.
Note that running `make validate` in a container is still desired.
A similar approach may be re-added in the future.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
When looking up a container's cgroup path, parse /proc/[PID]/cgroup.
This will work across all cgroup managers and configurations and is
supported on cgroups v1 and v2.
Fixes: #8265
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
The --hostname and containername should always be added to containers.
Added some tests to make sure you can always ping the hostname and container
name from within the container.
Fixes: https://github.com/containers/podman/issues/8095
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
in an effort to speed up the remote testing, we should be using
lookaside storage to avoid pull images as well as importing multiple
images into the RW store.
one test was removed and added into system test by Ed in #8325
Signed-off-by: baude <bbaude@redhat.com>
Supposed to be able to search for labels with a given value.
Previously it meant searching for label key and label value:
--filter label=key --filter label=value
Add some documentation and integration tests for it as well.
Signed-off-by: Anders F Björklund <anders.f.bjorklund@gmail.com>
Check to see if we are recording the version of buildah
used to build the image as a label in the image.
Also we should make sure the filter "since" works.
We are only testing "after", which we don't document.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
When making containers, we want to lock all named volumes we are
adding the container to, to ensure they aren't removed from under
us while we are working. Unfortunately, this code did not account
for a container having the same volume mounted in multiple places
so it could deadlock. Add a map to ensure that we don't lock the
same name more than once to resolve this.
Fixes#8221
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
we need to migrate play kube away from using the old container creation
method. the new approach is specgen and this aligns play kube with
container creation in the rest of podman.
Signed-off-by: baude <bbaude@redhat.com>
The IPv6 e2e tests on the CI for rootles mode fails because
it needs the ip6tables modules loaded.
Example error:
stdout="", stderr="failed to list chains: running [/sbin/ip6tables -t nat -S --wait]: exit status 3: modprobe: can't change directory to '/lib/modules': No such file or directory\nip6tables v1.8.4 (legacy): can't initialize ip6tables table `nat': Table does not exist (do you need to insmod?)\nPerhaps ip6tables or your kernel needs to be upgraded.\n\n"
Signed-off-by: Antonio Ojea <aojea@redhat.com>
add e2e test that checks that is not possible to create
different networks with the same subnet, in IPv6 neither
in IPv4
Signed-off-by: Antonio Ojea <aojea@redhat.com>
enable the ipv6 flag in podman network to be able to create
dual-stack networks for containers.
This is required to be compatible with docker, where --ipv6
really means dual stack.
podman, unlike docker, support IPv6 only containers since
07e3f1bba9.
Signed-off-by: Antonio Ojea <aojea@redhat.com>
ed identified that the dnsname integration test does not use a unique
name and therefore cannot be cleaned up. this was made worse by a
improper defer statement to remove the network should the test fail.
Signed-off-by: baude <bbaude@redhat.com>
podman can now support adding network aliases when running containers
(--network-alias). It requires an updated dnsname plugin as well as an
updated ocicni to work properly.
Signed-off-by: baude <bbaude@redhat.com>
As of this commit, in Fedora 33, without without `CAP_NET_ADMIN` and
`CAP_NET_RAW`, require setting `net.ipv3.ping_group_range` in order for
the `ping` command to work inside a container. However, not all images
`ping` are created equal. For whatever reason, the busybox version in
the busybox container image, does not function. Switch to the Alpine
image's busybox ping, which seems to work fine.
Signed-off-by: Chris Evich <cevich@redhat.com>
When using multiple filters, return a volume that matches any one of the used filters, rather than matching both of the filters.
This is for compatibility with docker's cli, and more importantly, the apiv2 compat endpoint
Closes#6765
Signed-off-by: Ashley Cui <acui@redhat.com>
Using a function like ContainSubstring or Equal is better because if
the test fails it will log a descriptive error that includes the
actual string generated during the test. This is more helpful than a
function like BeTrue that will only indicate that an assertion failed
without giving further details of the failure.
Signed-off-by: Debarshi Ray <rishi@fedoraproject.org>
fedora does not have the the ability in rootless to set cpu limits.
this requires a simple fix for fedora 33 to pass ci tests.
Signed-off-by: baude <bbaude@redhat.com>
When --userns=keep-id is used, Podman is supposed to set up the home
directory of the user inside the container to match that on the host
as long as the home directory or any of its parents are marked as
volumes to be bind mounted into the container.
Currently, the test only considers the case where the home directory
itself is bind mounted into the container. It doesn't cover the Podman
code that walks through all the bind mounts looking for ancestors in
case the home directory itself wasn't specified as a bind mount.
Therefore, this improves the existing test added in commit
6ca8067956 ("Setup HOME environment when using --userns=keep-id")
Note that this test can't be run as root. The home directory of the
root user is /root, and it's parent is /. Bind mounting the entire /
from the host into the container prevents it from starting:
Error: openat2 ``: No such file or directory: OCI not found
Signed-off-by: Debarshi Ray <rishi@fedoraproject.org>
if --userns=keep-id is specified and not --user is specified, take the
unprivileged capabilities code path so that ambient capabilities are
honored in the container.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
if the kernel supports ambient capabilities (Linux 4.3+), also set
them when running with euid != 0.
This is different that what Moby does, as ambient capabilities are
never set.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
When creating a container in a pod the podname was always set as
the dns entry. This is incorrect when the container is not part
of the pods network namespace. This happend both rootful and
rootless. To fix this check if we are part of the pods network
namespace and if not use the container name as dns entry.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
If the image name not a manifest list type, enable manifest inspect to return manifest of single image manifest type vnd.docker.distribution.manifest.v2+json.
Signed-off-by: Qi Wang <qiwan@redhat.com>
Remove the search limit check since the c/image v5.6.0 supports pagination and can give result over 100 entries.
Signed-off-by: Qi Wang <qiwan@redhat.com>
Containers that share IPC Namespaces share each others
/dev/shm, which means a private /dev/shm needs to be setup
for the infra container.
Added a system test and an e2e test to make sure the
/dev/shm is shared.
Fixes: https://github.com/containers/podman/issues/8181
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Followon to #7965 (mirror registry). mirror.gcr.io doesn't
cache all the images we need, and I can't find a way to
add to its cache, so let's just use quay.io for those
images that it can't serve.
Tools used:
skopeo copy --all docker://docker.io/library/alpine:3.10.2 \
docker://quay.io/libpod/alpine:3.10.2
...and also:
docker.io/library/alpine:3.2
docker.io/library/busybox:latest
docker.io/library/busybox:glibc
docker.io/library/busybox:1.30.1
docker.io/library/redis:alpine
docker.io/libpod/alpine-with-bogus-seccomp:label
docker.io/libpod/alpine-with-seccomp:label
docker.io/libpod/alpine_healthcheck:latest
docker.io/libpod/badhealthcheck:latest
Since most of those were new quay.io/libpod images, they required
going in through the quay.io GUI, image, settings, Make Public.
Signed-off-by: Ed Santiago <santiago@redhat.com>
podman inspect only had the capabilities to inspect containers and images. if a user wanted to inspect a pod, volume, or network, they would have to use `podman network inspect`, `podman pod inspect` etc. Docker's cli allowed users to inspect both volumes and networks using regular inspect, so this commit gives the user the functionality
If the inspect type is not specified using --type, the order of inspection is:
containers
images
volumes
networks
pods
meaning if container that has the same name as an image, podman inspect would return the container inspect.
To avoid duplicate code, podman network inspect and podman volume inspect now use the inspect package as well. Podman pod inspect does not because podman pod inspect returns a single json object while podman inspect can return multiple)
Signed-off-by: Ashley Cui <acui@redhat.com>
Detached containers and detach keys are only created with the podman run, i
exec, and start commands. We do not store the detach key sequence or the
detach flags in the database, nor does Docker. The current code was ignoreing
these fields but documenting that they can be used.
Fix podman create man page and --help output to no longer indicate that
--detach and --detach-keys works.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This does not match Docker, which does not add hostname in this
case, but it seems harmless enough.
Fixes#8095
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Make a distinction between pods that are completely running (all
containers running) and those that have some containers going,
but not all, by introducing an intermediate state between Stopped
and Running called Degraded. A Degraded pod has at least one, but
not all, containers running; a Running pod has all containers
running.
First step to a solution for #7213.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
In older versions of podman, we supported decimal numbers defaulting
to microseconds. This PR fixes to allow users to continue to specify
only digits.
Also cleaned up documentation to fully describe what input for --interval flag.
Finally improved testing on podman wait to actually make sure the command succeeded.
Fixed tests to work on podman-remote.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
- apiv2 - the 'ten /info requests' test is flaking often,
taking ~8 seconds (our limit is 7, up from 5 a few weeks
ago). Brent suggested that the first /info call might be
expensive, because it needs to access storage. So, let's
prime it by running one /info outside the timing loop.
And, because even that continues to fail, bump it up
to 10 seconds and file #8076 to track the slowdown.
- toolbox test - WaitForReady() has timed out, even on one
occasion causing a run failure because it failed 3 times.
Solution: bump up timeout from 2s to 5s. Not really great,
but CI systems are underpowered, and it's not unreasonable
that 2s might be too low.
- sdnotify test - add a 'podman wait' between stop & rm.
This may prevent a "cannot rm container as it is running"
race condition.
While working on this, Brent and I noticed a few ways that
test-apiv2 logging can be improved:
- test name: when request is POST, display the jsonified
parameters, not the original input ones. This should
make it much easier to reproduce failures.
- use curl's "--write-out" option to capture http code,
content type, and request time. We were getting the
first two via grep from logged headers; this is cleaner.
And there was no other way to get timing. We now include
the timing as X-Response-Time in the log file.
- abort on *any* curl error, not just 7 (cannot connect).
Any error at all from curl is bad news.
Signed-off-by: Ed Santiago <santiago@redhat.com>
When a container either joins a pod that shares the network
namespace or uses `--net=container:` to share the network
namespace of another container, it does not have its own copy of
the CNI results used to generate `podman inspect` output. As
such, to inspect these containers, we should be going to the
container we share the namespace with for network info.
Fixes#8073
Signed-off-by: Matthew Heon <mheon@redhat.com>
When defining multiple ports (same src/dst) like `-p 80:80 -p 443:443`
then podman will not show the complete output on `podman ps` (only
`0.0.0.0:80->80/tcp` in the example). This also applies to port ranges.
This patch refactors the port loop by pre-checking for ranges and
displaying them correctly to the end user.
Signed-off-by: Sascha Grunert <sgrunert@suse.com>
When a container uses --net=host the default hostname is set to
the host's hostname. However, we were not creating any entries
in `/etc/hosts` despite having a hostname, which is incorrect.
This hostname, for Docker compat, will always be the hostname of
the host system, not the container, and will be assigned to IP
127.0.1.1 (not the standard localhost address).
Also, when `--hostname` and `--net=host` are both passed, still
use the hostname from `--hostname`, not the host's hostname (we
still use the host's hostname by default in this case if the
`--hostname` flag is not passed).
Fixes#8054
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Previously, using an invalid image name would produce an error like
this:
Error: error encountered while bringing up pod test-pod-0: invalid reference format
This message didn't specify that there was an problem with an image
name, and it didn't specify which image name had a problem if there were
multiple. Now the error reads:
Error: error encountered while bringing up pod test-pod-0: Failed to parse image "./myimage": invalid reference format
Signed-off-by: Jordan Christiansen <xordspar0@gmail.com>
These options are now fully supported in the remote API and should no
longer be hidden and/or documented as non supported.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Currently the HOME environment is set to /root if
the user does not override it.
Also walk the parent directories of users homedir
to see if it is volume mounted into the container,
if yes, then set it correctly.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
* system df
* events
* fix error handling from go routine
* update tests to use gomega matchers for better error messages
* system info
* version
* volume inspect
Signed-off-by: Jhon Honce <jhonce@redhat.com>
We were only including the CNI Network fields in the output of
`podman inspect` when the container was not running. It's simple
enough to fix (populate with empty structs, since we can't fill
anything without a CNI response to get IP address assigned, etc).
This is necessary for Docker compatibility.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
In the past, Toolbox[0] has been affected by several of Podman's
bugs/changes of behaviour. This is one of the steps to assure that as
Podman progresses, Podman itself and subsequently Toolbox do not regress.
One of the other steps is including Toolbox's system tests in Podman's
gating systems (which and to what extent is yet to be decided on).
The tests are trying to stress parts of Podman that Toolbox needs for
its functionality: permission to handle some system files, correct
values/permissions/limits in certain parts, management of users and
groups, mounting of paths,.. The list is most likely longer and
therefore more commits will be needed to control every aspect of the
Toolbox/Podman relationship :).
Some test cases in test/e2e/toolbox_test.go rely on some tools being
present in the base image[1]. That is not the case with the common
ALPINE image or the basic Fedora image.
Some tests might be duplicates of already existing tests. I'm more in
favour of having those duplicates. Thanks to that it will be clear what
functionality/behaviour Toolbox requires.
[0] https://github.com/containers/toolbox
[1] https://github.com/containers/toolbox/#image-requirements
Signed-off-by: Ondřej Míchal <harrymichal@seznam.cz>
Currenly if a user specifies the name or ID of an external storage
container, we report an error to them.
buildah from scratch
working-container-2
podman rm working-container-2
Error: no container with name or ID working-container-2 found: no such container
Since the user specified the correct name and the container is in storage we
force them to specify --storage to remove it. This is a bad experience for the
user.
This change will just remove the container from storage. If the container
is known by libpod, it will remove the container from libpod as well.
The podman rm --storage option has been deprecated, and removed from docs.
Also cleaned documented options that are not available to podman-remote.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Following commands:
* systemd generate
* networks inspect
* pod stats
* Fixed test where format was quoted and then quoted again
* Fixed bug where output never printed '--' on missed reads
* pod ps
Signed-off-by: Jhon Honce <jhonce@redhat.com>
In Podman 1.9.3, `podman run -p 80` would assign port 80 in the
container to a random port on the host. In Podman 2.0 and up, it
assigned Port 80 in the container to Port 80 on the host. This is
an easy fix, fortunately - just need to remove the bit that
assumed host port, if not given, should be set to container port.
We also had a test for the bad behavior, so fix it to test for
the correct way of doing things.
Fixes#7947
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
due to a lack of "locking" on cni operations, we could get ourselves in trouble when doing rapid creation or removal of networks. added a simple file lock to deal with the collision and because it is not considered a performent path, use of the file lock should be ok. if proven otherwise in the future, some generic shared memory lock should be implemented for libpod and also used here.
moved pkog/network to libpod/network because libpod is now being pulled into the package and it has therefore lost its generic nature. this will make it easier to absorb into libpod as we try to make the network closer to core operations.
Fixes: #7807
Signed-off-by: baude <bbaude@redhat.com>
We do not populate the hostname field with the IP Address
when running within a user namespace.
Fixes https://github.com/containers/podman/issues/7490
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Docker supports log-opt max_size and so does conmon (ALthough poorly).
Adding support for this allows users to at least make sure their containers
logs do not become a DOS vector.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
All containers within a Pod need to run with the same SELinux
label, unless overwritten by the user.
Also added a bunch of SELinux tests to make sure selinux labels
are correct on namespaces.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
A podman could not read logs written to journald properly, due to a tail config bug.
Added a system test to check this - since e2e tests don't like journald
Signed-off-by: Ashley Cui <acui@redhat.com>
* --format "table {{.field..." will print fields out in a table with
headings. Table keyword is removed, spaces between fields are
converted to tabs
* Update parse.MatchesJSONFormat()'s regex to be more inclusive
* Add report.Headers(), obtain all the field names to be used as
column headers, a map of field name to column headers may be provided
to override the field names
* Update several commands to use new functions
Signed-off-by: Jhon Honce <jhonce@redhat.com>
The network test created config files with random filenames
but the network name was static. Since the tests can run in
parallel podman was not able to distinguish the networks.
We need to make sure that each test has its own config file
and network name. This helps to prevent unnecessary flakes.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
The systemd test was inaccurately being skipped when a userns env. var.
flag was set. At best this is confusing to new developers, and at worse
it actively blocks an otherwise legitimate class of tests. Improve the
accuracy of skip-logic by adding/using a purpose built set of
functions.
Signed-off-by: Chris Evich <cevich@redhat.com>
If run&create image returns error: image contains manifest list, not a runnable image, find the local image that has digest matching the digest from the list and use the image from local storage for the command.
Signed-off-by: Qi Wang <qiwan@redhat.com>
When looking up local images, take the unqualified-serach registries of
the registries.conf into account (on top of "localhost/").
Also extend the integration tests to prevent future regressions.
Fixes: #6381
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
We need to handle removal of non existing network.
This allows the `removeCNINetwork` function
always to be called. This is needed by tests which are trying
to remove the network manually in order to prevent flakes.
Fixes#7809
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Test passes on Fedora because the registry server is one of the
defaults. However it is not typically configured on Ubuntu hosts, and
therefor this test can fail. While specifying the FQIN in the
dockerfile text is not an ideal solution, it cannot negatively
affect other tests which utilize `podmanTest.BuildImage`.
Signed-off-by: Chris Evich <cevich@redhat.com>
change capabilities handling to reflect what docker does.
Bounding: set to caplist
Inheritable: set to caplist
Effective: if uid != 0 then clear; else set to caplist
Permitted: if uid != 0 then clear; else set to caplist
Ambient: clear
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Fix the remote client to untag all tags of the specified image.
Instead of querying the image on the client side, support the
case where both, repo and tag, are empty and remove all tags.
Reuse the ABI implementation where possible. In retrospective,
the libpod untag endpoint should support a slice of strings to
batch remove tags rather than reaching out for each tag individually.
Enable the skipped test.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
In podman containers rm and podman images rm, the commands
exit with error code 1 if the object does not exists.
This PR implements similar functionality to volumes, networks, and Pods.
Similarly if volumes or Networks are in use by other containers, and return
exit code 2.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
If user sets namespace to host, then default sysctls need to be ignored
that are specific to that namespace.
--net=host ignore sysctls that begin with net.
--ipc=host ignore fs.mqueue
--uts=host ignore kernel.domainname and kernel.hostname
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The podman remote client ignored the force option due a typo.
If an error occured the remote client would panic with an
index out of range error.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Currently the --pull missing|always|never is ignored
This PR implements this for local API. For remote we
need to default to pullpolicy specified in the containers.conf
file.
Also fixed an issue when images were matching other images names
based on prefix, causing images to always be pulled.
I had named an image myfedora and when ever I pulled fedora, the system
thought that it there were two images named fedora since it was checking
for the name fedora as well as the prefix fedora. I changed it to check
for fedora and the prefix /fedora, to prefent failures like I had.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
We need to get more tests running in rootless mode. Since cgroupsV2 allows
management of cgroups in rootless environments a lot of more tests can be run.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The remote client does not support loading directories yet. To prevent
confusing error messages and to make the behaviour more explicit, check
if the input points to a directory and throw an error if needed.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
Add a new endpoint for container stats allowing for batch operations on
more than one container. The new endpoint deprecates the
single-container endpoint which will eventually be removed with the next
major release.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
Before this change, filters of the form `podman pod ps --filter
label=app=myapp` were not working. The results would include all pods
that contained the app label with any value. Looking at the code, this
makes sense. It appears that the second = and everything after it were
getting truncated.
Even though there was already a passing test that tested `podman pod ps
--filter label=io.podman.test.label=value1`, the test failed with the
above example with a label `app=myapp`. The new code works in both
cases.
Signed-off-by: Jordan Christiansen <xordspar0@gmail.com>
Most have been fixed, others I replaced with SkipIfRemote
Fix ContainerStart on tunnel, it needs to wait for the exit status
before returning.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Now that Dan has added helpful comments to each SkipIfRemote,
let's take the next step and include those messages in the
Skip() output so someone viewing test results can easily
see if a remote test is skipped for a real reason or for
a FIXME.
This commit is the result of a simple:
perl -pi -e 's;(SkipIfRemote)\(\)(\s+//\s+(.*))?;$1("$3");' *.go
in the test/e2e directory, with a few minor (manual) changes
in wording.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Remove ones that are not needed.
Document those that should be there.
Document those that should be fixed.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
When we execute podman-remote attach, we were not checking if the
container was in the correct state, this is leading to timeouts and
we had turned off remote testing.
Also added an IfRemote() function so we can turn on more tests when
using the "-l" flag for local, but use container name for remote.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Currently `podman image unmount` report every image that is mounted
when it unmounts them. We should only report unmounted actually mounted images.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Add a bunch of tests to ensure that --volumes-from
works as expected.
Also align the podman create and run man page.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Podman wants to guarantee that exec sessions retain the groups of
the container they are started in, unless explicitly overridden
by the user. This guarantee was broken for containers where the
`--user` flag was specified; this patch resolves that.
Somewhere in the Exec rewrite for APIv2, I changed the location
where the container's User is passed into the exec session
(similar to groups, we also want to preserve user unless
overridden). The lower-level Exec APIs already handled setting
user and group appropriately if not specified when the exec
session was created, but I added duplicate code to handle this
higher in the stack - and that code only handled setting user,
not supplemental groups, breaking support in that specific case.
Two things conspired to make this one hard to track down: first,
things were only broken if the container explicitly set a user;
otherwise, the container user would still appear to be unset to
the lower-level code, which would properly set supplemental
groups (this tricked our existing test into passing). Also, the
`crun` OCI runtime will add the groups without prompting, which
further masked the problem there. I debated making `runc` do the
same, but in the end it's better to fix this in Podman - it's
better to be explicit about what we want done so we will work
with all OCI runtimes.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
for podman-remote build operations, the iidfile, when used, needs to write the file to the client's local filesystem.
Signed-off-by: baude <bbaude@redhat.com>
Always return all associated names / repo tags of an image and fix a bug
with malformed repo tags.
Previously, Podman returned all names only with `--all` but this flag
only instructs to list intermediate images and should not alter
associated names. With `--all` Podman queried the repo tags of an image
which splits all *tagged* names into repository and tag which is then
reassembled to eventually be parsed again in the frontend. Lot's of
redundant CPU heat and buggy as the reassembly didn't consider digests
which ultimately broke parsing in the frontend.
Fixes: #7651
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
When using `podman play kube` with a YAML file that has pod labels,
apply those labels to the pods that podman makes.
For example, this Deployment spec has labels on a pod:
apiVersion: apps/v1
kind: Deployment
metadata:
name: myapp
labels:
app: myapp
spec:
selector:
matchLabels:
app: myapp
template:
metadata:
labels:
app: myapp
spec:
containers:
- name: web
image: nginx
ports:
- containerPort: 80
The pods that podman creates will have the label "app" set to "myapp" so
that these pods can be found with `podman pods ps --filter label=app`.
Signed-off-by: Jordan Christiansen <xordspar0@gmail.com>
Currently infr-command and --infra-image commands are ignored
from the user. This PR instruments them and adds tests for
each combination.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Enables podman create, pull, run, import to use --signature-policy option. Set it as hidden flag to be consistent with other commands.
Signed-off-by: Qi Wang <qiwan@redhat.com>
* Refactor/Rename channel.WriteCloser() to encapsulate the channel
* Refactor build endpoint to "live" stream buildah output channels
over API rather then buffering output
* Refactor bindings/tunnel build because endpoint changes
* building tar file now in bindings rather then depending on
caller
* Cleanup initiating extra image engine
* Remove setting fields to zero values (less noise in code)
* Update tests to support remote builds
Fixes#7136Fixes#7137
Signed-off-by: Jhon Honce <jhonce@redhat.com>
