summaryrefslogtreecommitdiff
path: root/libpod/oci.go
diff options
context:
space:
mode:
authorMatthew Heon <matthew.heon@pm.me>2019-12-12 16:19:36 -0500
committerMatthew Heon <matthew.heon@pm.me>2019-12-12 16:35:37 -0500
commitbd44fd5c815fc750fd6b60899328564bee74e6e5 (patch)
tree980a0e85eec4f3a646c1f0ac63c05634fd13510b /libpod/oci.go
parentc2dab75f0ec1502e075cf5174d1b308267648fda (diff)
downloadpodman-bd44fd5c815fc750fd6b60899328564bee74e6e5.tar.gz
podman-bd44fd5c815fc750fd6b60899328564bee74e6e5.tar.bz2
podman-bd44fd5c815fc750fd6b60899328564bee74e6e5.zip
Reap exec sessions on cleanup and removal
We currently rely on exec sessions being removed from the state by the Exec() API itself, on detecting the session stopping. This is not a reliable method, though. The Podman frontend for exec could be killed before the session ended, or another Podman process could be holding the lock and prevent update (most notable in `run --rm`, when a container with an active exec session is stopped). To resolve this, add a function to reap active exec sessions from the state, and use it on cleanup (to clear sessions after the container stops) and remove (to do the same when --rm is passed). This is a bit more complicated than it ought to be because Kata and company exist, and we can't guarantee the exec session has a PID on the host, so we have to plumb this through to the OCI runtime. Fixes #4666 Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Diffstat (limited to 'libpod/oci.go')
-rw-r--r--libpod/oci.go6
1 files changed, 3 insertions, 3 deletions
diff --git a/libpod/oci.go b/libpod/oci.go
index 9e761788e..05a2f37db 100644
--- a/libpod/oci.go
+++ b/libpod/oci.go
@@ -23,9 +23,6 @@ type OCIRuntime interface {
// CreateContainer creates the container in the OCI runtime.
CreateContainer(ctr *Container, restoreOptions *ContainerCheckpointOptions) error
// UpdateContainerStatus updates the status of the given container.
- // It includes a switch for whether to perform a hard query of the
- // runtime. If unset, the exit file (if supported by the implementation)
- // will be used.
UpdateContainerStatus(ctr *Container) error
// StartContainer starts the given container.
StartContainer(ctr *Container) error
@@ -59,6 +56,9 @@ type OCIRuntime interface {
// If timeout is 0, SIGKILL will be sent immediately, and SIGTERM will
// be omitted.
ExecStopContainer(ctr *Container, sessionID string, timeout uint) error
+ // ExecUpdateStatus checks the status of a given exec session.
+ // Returns true if the session is still running, or false if it exited.
+ ExecUpdateStatus(ctr *Container, sessionID string) (bool, error)
// ExecContainerCleanup cleans up after an exec session exits.
// It removes any files left by the exec session that are no longer
// needed, including the attach socket.