mirror of https://github.com/containers/podman.git
19aabf440e
The remote API will wait 300s by default before conmon will call the cleanup. In the meantime when you inspect an exec session started with ExecStart() (so not attached) and it did exit we do not know that. If a caller inspects it they think it is still running. To prevent this we should sync the session based on the exec pid and update the state accordingly. For a reproducer see the test in this commit or the issue. Fixes #18424 Signed-off-by: Paul Holzinger <pholzing@redhat.com> |
||
|---|---|---|
| .. | ||
| fixture | ||
| attach_test.go | ||
| auth_test.go | ||
| common_test.go | ||
| connection_test.go | ||
| containers_test.go | ||
| create_test.go | ||
| exec_test.go | ||
| generator_test.go | ||
| images_test.go | ||
| info_test.go | ||
| manifests_test.go | ||
| networks_test.go | ||
| pods_test.go | ||
| resource_test.go | ||
| secrets_test.go | ||
| system_test.go | ||
| test_suite_test.go | ||
| types_test.go | ||
| volumes_test.go | ||