Skip to content

fix issue75: seg fault and listen volcano.sock fail #76

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
linuxfhy wants to merge 1 commit into from
Closed

fix issue75: seg fault and listen volcano.sock fail #76

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 13 additions & 2 deletions pkg/plugin/nvidia/server.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -186,9 +186,20 @@ func (m *NvidiaDevicePlugin) DevicesNum() int {

// Serve starts the gRPC server of the device plugin.
func (m *NvidiaDevicePlugin) Serve() error {

sock, err := net.Listen("unix", m.socket)
if err != nil {
return err
log.Printf("Listen sock fail and retry for '%s': %s", m.resourceName, err)
err = os.Remove(m.socket)
if err != nil {
log.Printf("Error deleting file: %s, %v\n", m.socket, err)
return err
}
sock, err = net.Listen("unix", m.socket)
if err != nil {
log.Printf("Retry Listen sock fail '%s': %s", m.resourceName, err)
return err
}
}

pluginapi.RegisterDevicePluginServer(m.server, m)
Expand Down Expand Up @@ -343,6 +354,7 @@ func (m *NvidiaDevicePlugin) Allocate(ctx context.Context, reqs *pluginapi.Alloc
}

sort.Sort(availablePods)
util.UseClient(m.kubeInteractor.clientset)

Comment on lines +357 to 358
Copy link

Copilot AI Mar 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ensure the relocation of util.UseClient here is intentional and that omitting its call at the original location does not lead to any side effects or missed initialization.

Suggested change
util.UseClient(m.kubeInteractor.clientset)

Copilot uses AI. Check for mistakes.
var candidatePod *v1.Pod
for _, pod := range availablePods {
Expand Down Expand Up @@ -406,7 +418,6 @@ Allocate:
return nil, fmt.Errorf("failed to update pod annotation %v", err)
}

util.UseClient(m.kubeInteractor.clientset)
klog.V(3).Infoln("Releasing lock: nodeName=", m.kubeInteractor.nodeName)
err = util.ReleaseNodeLock(m.kubeInteractor.nodeName, "gpu")
if err != nil {
Expand Down
Loading