github/kata-containers
1
0
Fork 1
mirror of https://github.com/kata-containers/kata-containers.git synced 2025-04-29 12:14:48 +00:00
Code Issues Projects Releases Wiki Activity

runtime: stop getting OOM events when ttrpc: closed error

Browse Source 
getOOMEvents is a long-waiting call, it will retry when failed.
For cases of agent shutdown, the retry should stop.

When the agent hasn't detected agent has died, we can also check
whether the error is "ttrpc: closed".

Fixes: #3815

Signed-off-by: bin <bin@hyper.sh>
This commit is contained in:
bin 2022-03-29 16:39:01 +08:00
parent 9495316145
commit fb8be96194
1 changed files with 3 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/runtime/pkg/containerd-shim-v2/wait.go
View File

@ -15,7 +15,6 @@ import (
"github.com/containerd/containerd/api/types/task" "github.com/containerd/containerd/api/types/task"
"github.com/containerd/containerd/mount" "github.com/containerd/containerd/mount"
"github.com/sirupsen/logrus" "github.com/sirupsen/logrus"
"google.golang.org/grpc/codes"
"github.com/kata-containers/kata-containers/src/runtime/pkg/oci" "github.com/kata-containers/kata-containers/src/runtime/pkg/oci"
) )
@ -156,13 +155,11 @@ func watchOOMEvents(ctx context.Context, s *service) {
default: default:
containerID, err := s.sandbox.GetOOMEvent(ctx) containerID, err := s.sandbox.GetOOMEvent(ctx)
if err != nil { if err != nil {
shimLog.WithError(err).Warn("failed to get OOM event from sandbox") if err.Error() == "ttrpc: closed" || err.Error() == "Dead agent" {
// If the GetOOMEvent call is not implemented, then the agent is most likely an older version, shimLog.WithError(err).Warn("agent has shutdown, return from watching of OOM events")
// stop attempting to get OOM events.
// for rust agent, the response code is not found
if isGRPCErrorCode(codes.NotFound, err) || err.Error() == "Dead agent" {
return return
} }
shimLog.WithError(err).Warn("failed to get OOM event from sandbox")
time.Sleep(defaultCheckInterval) time.Sleep(defaultCheckInterval)
continue continue
} }