daemon daedlock w/ containerd communication #35825
Description
CI is failing a lot with this:
15:31:46 ----------------------------------------------------------------------
15:31:46 FAIL: check_test.go:366: DockerSwarmSuite.TearDownTest
15:31:46
15:31:46 check_test.go:371:
15:31:46 d.Stop(c)
15:31:46 daemon/daemon.go:395:
15:31:46 t.Fatalf("Error while stopping the daemon %s : %v", d.id, err)
15:31:46 ... Error: Error while stopping the daemon d2305ce671729 : exit status 130
15:31:46
15:31:46
15:31:46 ----------------------------------------------------------------------
15:31:46 PANIC: docker_api_swarm_service_test.go:201: DockerSwarmSuite.TestAPISwarmServicesUpdateStartFirst
15:31:46
15:31:46 [d2305ce671729] waiting for daemon to start
15:31:46 [d2305ce671729] daemon started
15:31:46
15:31:46 [d2305ce671729] daemon started
15:31:46 Attempt #2: daemon is still running with pid 11196
15:31:46 Attempt #3: daemon is still running with pid 11196
15:31:46 Attempt #4: daemon is still running with pid 11196
15:31:46 [d2305ce671729] exiting daemon
15:31:46 ... Panic: Fixture has panicked (see related PANIC)
Here is the relevent daemon log: docker.log
The deadlock appears to be in communicating with containerd.
This could be due changes from #35809