Skip to content

fix: avoid deleting peers on graceful close #14165

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

Merged
merged 14 commits into from
Aug 14, 2024
Merged

fix: avoid deleting peers on graceful close #14165

Changes from 1 commit
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
e2f3c31
fix: avoid deleting peers on graceful close
sreya Aug 4, 2024
82c6dce
lint
sreya Aug 5, 2024
aeca9b3
Merge branch 'main' into jon/pgc
sreya Aug 5, 2024
b5c6d96
set all peers to lost on shutdown
sreya Aug 11, 2024
b1f6aea
add heartbeat failure test
sreya Aug 11, 2024
1c905c7
delete stale comments
sreya Aug 11, 2024
fadde34
lint
sreya Aug 11, 2024
159b509
extra checks
sreya Aug 12, 2024
04472e3
pr comments
sreya Aug 12, 2024
eb52aa2
wait for workers to exit
sreya Aug 13, 2024
3173e44
simplify ctx management
sreya Aug 13, 2024
4754554
delete peers once
sreya Aug 14, 2024
0913392
instantiate channel
sreya Aug 14, 2024
8a65a2f
wrong ctx
sreya Aug 14, 2024
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
Prev Previous commit
Next Next commit
wait for workers to exit
  • Loading branch information
@sreya
sreya committed Aug 13, 2024
commit eb52aa21423a2b75ddcfb5a1cee6f1ccb5f3c6ce
29 changes: 16 additions & 13 deletions enterprise/tailnet/pgcoord.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -164,11 +164,7 @@ func newPGCoordInternal(
closed: make(chan struct{}),
}
go func() {
// when the main context is canceled, or the coordinator closed, the binder, tunneler, and
// handshaker always eventually stop. When the
// binder stops it updates all the peers handled
// by this coordinator to LOST.
c.binder.workerWG.Wait()
c.binder.wait()
c.tunneler.workerWG.Wait()
c.handshaker.workerWG.Wait()
querierCancel()
Expand Down Expand Up @@ -518,14 +514,7 @@ func (b *binder) handleBindings() {
for {
select {
case <-b.ctx.Done():
b.logger.Debug(b.ctx, "binder exiting, updating peers to lost", slog.Error(b.ctx.Err()))
err := b.store.UpdateTailnetPeerStatusByCoordinator(context.Background(), database.UpdateTailnetPeerStatusByCoordinatorParams{
CoordinatorID: b.coordinatorID,
Status: database.TailnetStatusLost,
})
if err != nil {
b.logger.Error(b.ctx, "update peer status to lost", slog.Error(err))
}
b.logger.Debug(b.ctx, "binder exiting")
return
case bnd := <-b.bindings:
b.storeBinding(bnd)
Expand Down Expand Up @@ -637,6 +626,20 @@ func (b *binder) retrieveBinding(bk bKey) binding {
return bnd
}

func (b *binder) wait() {
b.workerWG.Wait()

b.logger.Debug(b.ctx, "binder exiting, updating peers to lost", slog.Error(b.ctx.Err()))

err := b.store.UpdateTailnetPeerStatusByCoordinator(context.Background(), database.UpdateTailnetPeerStatusByCoordinatorParams{
CoordinatorID: b.coordinatorID,
Status: database.TailnetStatusLost,
})
if err != nil {
b.logger.Error(b.ctx, "update peer status to lost", slog.Error(err))
}
}

// mapper tracks data sent to a peer, and sends updates based on changes read from the database.
type mapper struct {
ctx context.Context
Expand Down
Loading