When enable_ipv6=true is set for slirp4netns (default since podman v4),
we will try to set the accept sysctl. This sysctl will not exist on
systems that have ipv6 disabled. In this case we should not error and
just ignore the extra ipv6 setup.
Also the current logic to wait for the slirp4 setup was kinda broken, it
did not actually wait until the sysctl was set before starting slirp.
This should now be fixed by using two `sync.WaitGroup`s.
[NO NEW TESTS NEEDED]
Fixes#13388
Signed-off-by: Paul Holzinger <pholzing@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>
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>
(cherry picked from commit aafa80918a)
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 primarily served to protect us against shutting down the
Libpod runtime while operations (like creating a container) were
happening. However, it was very inconsistently implemented (a lot
of our longer-lived functions, like pulling images, just didn't
implement it at all...) and I'm not sure how much we really care
about this very-specific error case?
Removing it also removes a lot of potential deadlocks, which is
nice.
[NO NEW TESTS NEEDED]
Signed-off-by: Matthew Heon <mheon@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>
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>
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>
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>
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>
when running on NFS, a RemoveAll could cause EBUSY because of some
unlinked files that are still kept open and "silly renamed" to
.nfs$ID.
This is only half of the fix, as conmon needs to be fixed too.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2040379
Related: https://github.com/containers/conmon/pull/319
[NO NEW TESTS NEEDED] as it requires NFS as the underlying storage.
Signed-off-by: Giuseppe Scrivano <gscrivan@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>
the config.json file for the OCI runtime is never closed, this is a
problem when running on NFS, since it leaves around stale files that
cannot be unlinked.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
Waiting on an initialized sync.WaitGroup returns immediately.
Hence, move the goroutine to wait and close *after* reading
the logs.
Fixes: #12904
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
Improve our compatibility with Docker by better handling the
state strings that we print in `podman ps`. Docker capitalizes
all states in `ps` (we do not) - fix this in our PS code. Also,
stop normalizing ContainerStateConfigured to the "Created" state,
and instead make it always be Created, with the existing Created
state becoming Initialized.
I didn't rename the actual states because I'm somewhat reticent
to make such a large change a day before we leave for break. It's
somewhat confusing that ContainerStateConfigured now returns
Created, but internally and externally we're still consistent.
[NO NEW TESTS NEEDED] existing tests should catch anything that
broke.
I also consider this a breaking change. I will flag appropriately
on Github.
Fixes RHBZ#2010432 and RHBZ#2032561
Signed-off-by: Matthew Heon <mheon@redhat.com>
This change updates the CDI API to commit 46367ec063fda9da931d050b308ccd768e824364
which addresses some inconistencies in the previous implementation.
Signed-off-by: Evan Lezar <elezar@nvidia.com>
Support removing the entire pod when --depend is used on an infra
container. --all now implies --depend to properly support removing all
containers and not error out when hitting infra containers.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
move the check after the cgroup manager is set, so to correctly detect
--cgroup-manager=cgroupfs and do not raise a warning about dbus not
being present.
Closes: https://github.com/containers/podman/issues/12802
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
The libpod/network packages were moved to c/common so that buildah can
use it as well. To prevent duplication use it in podman as well and
remove it from here.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This directory needs to be world searchable so users can access it from
different user namespaces.
Fixes: https://github.com/containers/podman/issues/12779
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This option causes Podman to not only remove the specified containers
but all of the containers that depend on the specified
containers.
Fixes: https://github.com/containers/podman/issues/10360
Also ran codespell on the code
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
For ip/macvlan networks we cannot use the gateway as address for this
hostname. In this case the gateway is normally not on the host so we
just try to use a local ip instead.
[NO NEW TESTS NEEDED] We cannot run macvlan networks in CI.
Fixes#11351
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Remove hard code use of the DefaultInfraImage and rely on
getting this from containers.conf.
Fixes: https://github.com/containers/podman/issues/12771
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
I don't see where these With Functions are used, so removing them to
clean up code.
WithDefaultInfra* functions screwed me up and confused me.
[NO NEW TESTS NEEDED]
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
