opensource/podman
1
0
Fork 0
mirror of https://github.com/containers/podman.git synced 2025-10-20 12:43:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

[v4.4.1-rhel] fix --health-on-failure=restart in transient unit

Browse Source 
As described in #17777, the `restart` on-failure action did not behave
correctly when the health check is being run by a transient systemd
unit.  It ran just fine when being executed outside such a unit, for
instance, manually or, as done in the system tests, in a scripted
fashion.

There were two issue causing the `restart` on-failure action to
misbehave:

1) The transient systemd units used the default `KillMode=cgroup` which
   will nuke all processes in the specific cgroup including the recently
   restarted container/conmon once the main `podman healthcheck run`
   process exits.

2) Podman attempted to remove the transient systemd unit and timer
   during restart.  That is perfectly fine when manually restarting the
   container but not when the restart itself is being executed inside
   such a transient unit.  Ultimately, Podman tried to shoot itself in
   the foot.

Fix both issues by moving the restart logic in the cleanup process.
Instead of restarting the container, the `healthcheck run` will just
stop the container and the cleanup process will restart the container
once it has turned unhealthy.

Backport of commit 95634154303f5b8c3d5c92820e2a3545c54f0bc8.

Fixes: #17777
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=2180125
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=2180126
Signed-off-by: Valentin Rothberg <vrothberg@redhat.com>
This commit is contained in:
Valentin Rothberg
2023-03-17 12:46:33 +01:00
parent ffc2614b8e
commit a2e80c53d4
3 changed files with 81 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
libpod/container_internal.go
View File

@ -228,6 +228,15 @@ func (c *Container) handleExitFile(exitFile string, fi os.FileInfo) error {
}
func (c *Container) shouldRestart() bool {
if c.config.HealthCheckOnFailureAction == define.HealthCheckOnFailureActionRestart {
isUnhealthy, err := c.isUnhealthy()
if err != nil {
logrus.Errorf("Checking if container is unhealthy: %v", err)
} else if isUnhealthy {
return true
}
}
// If we did not get a restart policy match, return false
// Do the same if we're not a policy that restarts.
if !c.state.RestartPolicyMatch ||
@ -267,6 +276,12 @@ func (c *Container) handleRestartPolicy(ctx context.Context) (_ bool, retErr err
return false, err
}
if c.config.HealthCheckConfig != nil {
if err := c.removeTransientFiles(ctx, c.config.StartupHealthCheckConfig != nil && !c.state.StartupHCPassed); err != nil {
return false, err
}
}
// Is the container running again?
// If so, we don't have to do anything
if c.ensureState(define.ContainerStateRunning, define.ContainerStatePaused) {
@ -1431,6 +1446,7 @@ func (c *Container) restartWithTimeout(ctx context.Context, timeout uint) (retEr
if err := c.stop(timeout); err != nil {
return err
}
if c.config.HealthCheckConfig != nil {
if err := c.removeTransientFiles(context.Background(), c.config.StartupHealthCheckConfig != nil && !c.state.StartupHCPassed); err != nil {
logrus.Error(err.Error())