Skip to content

fix(envd): return error when concurrent /init races on NFS mount #3460

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
AdaAibaby wants to merge 3 commits into from
Closed

fix(envd): return error when concurrent /init races on NFS mount #3460

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
ac63647
fix(envd): return error when concurrent /init races on NFS mount
adababys Jul 30, 2026
eaba58c
chore(envd): bump version to 0.6.14
adababys Aug 4, 2026
851478c
fix(envd): return 503 on concurrent NFS init so orchestrator retries
adababys Aug 4, 2026
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
9 changes: 8 additions & 1 deletion packages/envd/internal/api/init.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,7 @@ var pinMMDSWarnLimit = ratelimit.New(10 * time.Second)
var (
ErrAccessTokenMismatch = errors.New("access token validation failed")
ErrAccessTokenResetNotAuthorized = errors.New("access token reset not authorized")
ErrConcurrentNFSInit = errors.New("NFS mount already in progress")
)

const (
Expand Down Expand Up @@ -334,6 +335,12 @@ func writeInitError(w http.ResponseWriter, logger zerolog.Logger, err error) {
switch {
case errors.Is(err, ErrAccessTokenMismatch), errors.Is(err, ErrAccessTokenResetNotAuthorized):
w.WriteHeader(http.StatusUnauthorized)
case errors.Is(err, ErrConcurrentNFSInit):
// Transient: a previous /init is still mounting NFS (setupNFS uses
// context.WithoutCancel so it outlives a request cancelled by EnvdInitRequestTimeout).
// The orchestrator retries on 503.
logger.Warn().Err(err).Msg("NFS mount already in progress, retrying")
w.WriteHeader(http.StatusServiceUnavailable)
case errors.Is(err, host.ErrCAInstallInProgress):
// Not a failure, a concurrent install still holds the CA lock.
logger.Warn().Err(err).Msg("CA initialization still in progress, retrying")
Expand Down Expand Up @@ -371,7 +378,7 @@ func (a *API) setupNFS(ctx context.Context, logger zerolog.Logger, lifecycleID *
if !a.isMountingNFS.CompareAndSwap(false, true) {
logger.Debug().Msg("NFS volumes already mounting")

return e
return ErrConcurrentNFSInit
Comment thread
AdaAibaby marked this conversation as resolved.
Resolved
Comment thread
AdaAibaby marked this conversation as resolved.
Resolved
}
defer a.isMountingNFS.Store(false)

Expand Down
2 changes: 1 addition & 1 deletion packages/envd/pkg/version.go
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,3 @@
package pkg

const Version = "0.6.13" // x-release-please-version
const Version = "0.6.14" // x-release-please-version
9 changes: 8 additions & 1 deletion packages/orchestrator/pkg/sandbox/envd.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -118,7 +118,14 @@ func (s *Sandbox) doRequestWithInfiniteRetries(
cancel()

if err == nil {
return response, requestCount, nil
// Treat 503 as a transient failure and retry: envd returns 503 when a
// concurrent /init is still in progress (e.g. an NFS mount outlived the
// previous per-request timeout). Discard the body and loop.
if response.StatusCode == http.StatusServiceUnavailable {
response.Body.Close()
} else {
return response, requestCount, nil
}
}

select {
Expand Down