Issues

RateLimiter::rollback_increment uses get_mut, promoting the rolled-back key to most-recently-used and skewing LRU eviction/cleanup toward failed-admission keys

area: logic

LOW

#333

· erskingardner opened

on Jul 17, 2026

Internal-trigger shutdown arms force-quit without having consumed a signal, so a single routine SIGTERM aborts the graceful drain (asymmetric with the OS-signal path)

area: logic

LOW

#332

· erskingardner opened

on Jul 16, 2026

GlitchTip Event items are scrubbed twice (before_send + transport), and redact() heap-allocates once per redaction pattern even when only one matches

area: performance

LOW

#331

· erskingardner opened

on Jul 16, 2026

FCM OAuth success-body Parse/TooLarge is classified Permanent, dropping the push with no retry — the sibling variants #248 left uncovered

area: logic

LOW

#330

· erskingardner opened

on Jul 16, 2026

record_push_retry re-lowercases a static service name (heap allocation) on every retry attempt, on the provider-storm hot path

area: performance

LOW

#329

· erskingardner opened

on Jul 16, 2026

Initial gift-wrap subscription lookback ignores max_notification_age_secs / future_skew, under-fetching on startup relative to the server's own acceptance window and the reconnect path

area: logic

LOW

#328

· erskingardner opened

on Jul 16, 2026

Lagged subscription-refresh monitor performs no catch-up when a relay flaps entirely within the dropped-notification window, silently losing that relay's gift wraps

area: logic

LOW

#327

· erskingardner opened

on Jul 16, 2026

event_processing_permits is an identity pass-through with #[must_use], a doc block, and two tests — a dead abstraction that reads as if it enforces an invariant it does not

LOW

#326

· erskingardner opened

on Jul 15, 2026

run_periodic_cleanup uses the default Burst missed-tick behavior and an immediate startup tick, unlike the relay-status refresher — catch-up bursts compound the cleanup write-lock contention

area: performance

LOW

#325

· erskingardner opened

on Jul 15, 2026

OS signal handlers are installed only when wait_for_signal is first polled (after client + health-server init), so a SIGTERM during startup exits ungracefully — contradicting the comment at app.rs:544

area: logic

LOW

#324

· erskingardner opened

on Jul 15, 2026

server.max_rate_limit_cache_size has no upper bound, so a large value OOM/aborts at startup when the sharded RateLimiter preallocates its LRU stripes

area: logic

LOW

#323

· erskingardner opened

on Jul 15, 2026

append_seen_locked reopens the dedup file and issues a redundant chmod on every persisted event, inside the write-lock critical section

area: performance

LOW

#322

· erskingardner opened

on Jul 15, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

All issues

RateLimiter::rollback_increment uses get_mut, promoting the rolled-back key to most-recently-used and skewing LRU eviction/cleanup toward failed-admission keys

Internal-trigger shutdown arms force-quit without having consumed a signal, so a single routine SIGTERM aborts the graceful drain (asymmetric with the OS-signal path)

GlitchTip Event items are scrubbed twice (before_send + transport), and redact() heap-allocates once per redaction pattern even when only one matches

FCM OAuth success-body Parse/TooLarge is classified Permanent, dropping the push with no retry — the sibling variants #248 left uncovered

record_push_retry re-lowercases a static service name (heap allocation) on every retry attempt, on the provider-storm hot path

Initial gift-wrap subscription lookback ignores max_notification_age_secs / future_skew, under-fetching on startup relative to the server's own acceptance window and the reconnect path

Lagged subscription-refresh monitor performs no catch-up when a relay flaps entirely within the dropped-notification window, silently losing that relay's gift wraps

event_processing_permits is an identity pass-through with #[must_use], a doc block, and two tests — a dead abstraction that reads as if it enforces an invariant it does not

run_periodic_cleanup uses the default Burst missed-tick behavior and an immediate startup tick, unlike the relay-status refresher — catch-up bursts compound the cleanup write-lock contention

OS signal handlers are installed only when wait_for_signal is first polled (after client + health-server init), so a SIGTERM during startup exits ungracefully — contradicting the comment at app.rs:544

server.max_rate_limit_cache_size has no upper bound, so a large value OOM/aborts at startup when the sharded RateLimiter preallocates its LRU stripes

append_seen_locked reopens the dedup file and issues a redundant chmod on every persisted event, inside the write-lock critical section

Uh oh!

Search results