Latest release: v1.14.0
Changelog
Stay up to date with the latest QueueHub releases, feature additions, bug fixes, and improvements to our queue management platform.
v1.17.0
Latest✨ Features
- Live worker concurrency control — adjust any connected worker's concurrency directly from the dashboard's live worker view, with in-flight counts updating in real time as the change takes effect
- Worker utilization heatmap — per-worker view of concurrency headroom vs actual in-flight jobs over time, making I/O-bound and CPU-bound worker profiles distinguishable at a glance
- Throughput-vs-concurrency overlay — job duration and throughput charts can now be plotted against worker concurrency changes, so the effect of a concurrency adjustment is measurable, not guessed
🐛 Bug Fixes
- Fixed worker list showing stale concurrency values after a runtime adjustment made directly in application code via worker.concurrency
- Resolved job duration percentiles being skewed by jobs counted across worker restarts, showing inflated p99 on long-running queues
- Fixed the live worker view disconnecting and reconnecting in a loop when a worker's Redis connection pool rotated keys mid-session
v1.16.0
✨ Features
- Job retention policy inspector — the queue detail panel now shows each queue's configured removeOnComplete/removeOnFail settings (age, count) next to live state counts, so unbounded history is visible before it becomes memory pressure
- One-click queue cleanup — run paced batch cleanup (completed/failed, configurable age threshold) directly from the queue view, with a dry-run preview of how many jobs would be removed
- Retention drift detection — flags queues that share a name across environments but have different retention settings, catching the classic staging-works-production-grows failure
🐛 Bug Fixes
- Fixed completed-job counts on the overview dashboard growing without bound for queues with no retention policy — counts now reflect post-cleanup reality within one refresh cycle
- Resolved queue cleanup operations timing out on very large state sets when run during peak traffic; cleanup batches are now paced server-side
- Fixed job detail view failing to load for jobs whose data contained very large string payloads (over 1 MB) by paginating the payload display
v1.15.0
✨ Features
- DLQ depth alerting — SQS queues with a redrive policy can now alert on dead-letter queue depth and oldest-message age, with the same threshold-rule engine used for queue counts
- Redrive policy panel — the queue detail view surfaces maxReceiveCount and the DLQ target next to live receive-count distributions, so miscounted slow messages (visibility-timeout redeliveries) are distinguishable from true failures
- DLQ replay — replay dead-lettered messages back to their source queue in paced batches from the dashboard, with per-message receive history shown before you commit
🐛 Bug Fixes
- Fixed DLQ depth showing zero for FIFO dead-letter queues whose messages were all in the in-flight state rather than waiting
- Resolved alert notifications for SQS queues not including the queue's redrive policy context, making DLQ alerts harder to triage than count alerts
- Fixed an edge case where a queue's oldest-message-age calculation used the DLQ's own creation time for messages that predated it
v1.14.0
✨ Features
- SQS visibility timeout monitor — the queue detail panel now shows each queue's configured visibility timeout, current inflight count, and oldest inflight message age, so a message looping back for redelivery is visible before it double-processes work
- Per-message redelivery timeline — SQS job details now show the full receive history with timestamps for each visibility-timeout expiry, so you can tell a poison message from a merely slow one at a glance
- SQS dead-letter queue insights — DLQ-connected queues surface their redrive policy (source queue, maxReceiveCount) and a live count of messages awaiting redrive in the queue header
🐛 Bug Fixes
- Fixed SQS inflight counts drifting after a consumer deleted a message whose visibility timeout had already expired (double-delete race between two consumers)
- Resolved job detail not reflecting lease extensions made via ChangeMessageVisibility from your own workers, showing a stale remaining-time value
- Fixed SQS FIFO receive counts being double-counted in metrics when a message was redelivered within the same batch
v1.13.0
✨ Features
- Flow tree visualization — the queue detail panel now renders BullMQ parent-child flows as an interactive tree, with per-child status, attempt count, and run duration at a glance
- Orphan child detection — flags child jobs whose parent never enqueued (from a partially-failed flow.add), with a one-click promote action to release them from waiting-children
- Parent wait-time metrics — per-flow time from enqueue to parent release shown on the queue overview, so a slow batch is traceable to the exact child queue
🐛 Bug Fixes
- Fixed flow tree rendering collapsed when a child job's data contained circular references during serialization
- Resolved parent jobs stuck in waiting-children forever after a child was force-removed from the failed queue
- Fixed parent completion timestamps showing the last child's finish time instead of the parent's actual processed time
v1.12.0
✨ Features
- Repeatable job insights — new repeatable jobs tab on the queue detail panel showing every schedule's pattern or interval, timezone, next run time, and total completed runs
- Upcoming runs preview — hover any repeatable job to see its next 10 scheduled run times, computed with the configured timezone and DST rules instead of server-local time
- Schedule drift detection — warns when two queues register the same job name with different repeat configs, so a botched "fix" that creates a duplicate schedule surfaces immediately
🐛 Bug Fixes
- Fixed next-run time showing the previous run's timestamp when a worker restarted between scheduled executions
- Resolved duplicate repeatable job rows when two queues shared the same job name and pattern
- Fixed schedule preview overstating the next run time by ignoring jobs still delayed in the queue
v1.11.0
✨ Features
- Rate limiter state view — the queue detail panel now shows each queue's active limiter config (max, duration, groupKey) plus its current throttle position, so you can see exactly why waiting counts are climbing
- Webhook alert channels — deliver alert notifications to any HTTP endpoint with HMAC-SHA256 signing, custom headers, and automatic retry with exponential backoff
- Job retention policies — per-queue auto-pruning of completed and failed jobs older than a configurable age, with a dry-run preview before the policy takes effect
- SQS FIFO attributes in job details — inspect MessageGroupId, deduplication ID, and receive count directly in the job detail panel
🐛 Bug Fixes
- Fixed rate-limit throughput chart lagging one refresh cycle behind actual job processing on high-traffic queues
- Resolved webhook deliveries stalling indefinitely when the destination responded with a 3xx redirect loop
- Fixed retention policy purge skipping jobs still in the delayed state at scan time
v1.10.0
✨ Features
- Stalled job alerts — new alert rule that triggers when stalled job counts exceed a per-queue threshold, with configurable cooldown to avoid noise during deploys
- Retry schedule preview — see the computed attempt timeline and backoff delays before retrying a failed job
- Queue comparison view — side-by-side throughput, latency, and failure-rate charts for up to four queues
- CSV export for job history — export filtered job lists from the queue detail panel with column selection
🐛 Bug Fixes
- Fixed stalled job counter drifting negative after a worker reconnected and re-claimed expired jobs
- Fixed retry action resetting a job's priority to the default instead of preserving the original value
- Resolved job detail panel crash when payloads contained multi-megabyte strings
v1.9.0
✨ Features
- Priority queue depth indicators — per-priority-band waiting counts on the queue detail panel
- Agent Relay latency SLO dashboard — p50/p95/p99 per-agent latency with configurable alert thresholds
- WebSocket reconnect tuning — reduced reconnect interval from 5s to 500ms for agent relay connections
- Workspace-level API rate limits — per-workspace request quotas with configurable burst and cooldown windows
🐛 Bug Fixes
- Fixed priority lane starvation not reflected in queue depth metrics when high-priority jobs dominated
- Resolved WebSocket reconnection storm after Redis restart causing temporary 502 spikes
- Fixed worker count display showing stale values after backend reconnection
v1.8.0
✨ Features
- Agent Relay v2: WebSocket-based tunneling with automatic reconnection, end-to-end encryption, and sub-50ms latency
- Amazon SQS Beta — fully integrated queue monitoring for SQS Standard and FIFO queues
- Multi-region agent support: deploy agents across US, EU, and APAC regions
- Agent health dashboard with per-agent latency, throughput, and connection status
🐛 Bug Fixes
- Fixed memory leak in agent relay connection pool after extended uptime (24h+)
- Resolved queue name collision when two backends share identical queue names
- Fixed pagination reset when filtering jobs by status
- Improved WebSocket reconnection backoff to prevent thundering herd on Redis restart
v1.7.0
✨ Features
- Team Collaboration — invite team members, assign roles (Admin, Editor, Viewer), share dashboards
- Bulk Job Actions — retry, remove, or promote hundreds of jobs in a single operation with progress tracking
- Audit log with user actions, timestamps, and IP addresses for compliance
- Dashboard sharing via public links with expiration dates and passphrase protection
🐛 Bug Fixes
- Fixed job detail panel not refreshing after bulk retry actions
- Resolved Safari-specific rendering bug on queue metrics charts
- Fixed permission escalation edge case when modifying team member roles
- Addressed slow page loads on workspaces with 500+ queues
v1.6.0
✨ Features
- Alerting & Notifications — configure email, Slack, and Discord alerts for queue thresholds
- Custom alert rules: job failure rate, queue depth, stalled workers, delayed jobs
- On-call schedules with escalation policies and auto-resolution tracking
- Notification history dashboard with filterable event timeline
🐛 Bug Fixes
- Fixed alert trigger delay averaging 90 seconds instead of the configured 30-second window
- Resolved duplicate notifications when multiple rules matched the same event
- Fixed timezone offset display on alert timeline for non-UTC users
- Addressed notification channel test button not validating credentials
v1.5.0
✨ Features
- Self-Hosted Deployment — deploy QueueHub on your own infrastructure via Docker Compose and Helm charts
- PostgreSQL and SQLite support for QueueHub application data storage
- Environment-based configuration with comprehensive .env documentation
- Health check endpoints and readiness probes for Kubernetes deployments
- Database migration system with rollback support
🐛 Bug Fixes
- Fixed Docker entrypoint not honoring custom Redis port configuration
- Resolved Helm chart ingress template conflict with cert-manager annotations
- Fixed migration ordering issue when upgrading from v1.3 directly to v1.5
- Addressed health check timeout being too aggressive for cold-start scenarios
v1.4.0
✨ Features
- Job Inspection Overhaul — new job detail panel with syntax-highlighted data, stack traces, and timing waterfall
- Job search with Elasticsearch-like filtering: by status, queue, priority, timestamp range, and custom data fields
- Job comparison view to diff two jobs side-by-side
- Export job data as JSON or CSV with column selection
🐛 Bug Fixes
- Fixed job data truncation at 10KB in the detail panel (now supports 1MB payloads)
- Resolved infinite loading spinner when inspecting jobs with circular references in data
- Fixed timestamp display showing incorrect timezone for completed jobs
- Addressed search filter OR/AND logic mismatch in advanced queries
v1.3.0
✨ Features
- Multi-Backend Support — connect BullMQ, BeeQueue, and Bull queues simultaneously from a single dashboard
- Unified queue list that aggregates queues across all connected backends
- Backend health monitoring with per-connection status indicators and latency metrics
- Connection profiles with test-connection workflow before saving
🐛 Bug Fixes
- Fixed BeeQueue job state mapping where completed jobs incorrectly appeared as failed
- Resolved WebSocket reconnection loop when one of multiple backends becomes unreachable
- Fixed queue count mismatch between BullMQ v2 and v3 backends
- Addressed backend connection timeout not displaying user-friendly error messages
v1.2.0
✨ Features
- Agent Relay — securely connect to queues behind firewalls via WebSocket tunnel proxy
- Zero-trust architecture: agents authenticate via signed tokens and never expose internal ports
- Automatic TLS termination with Let's Encrypt certificate provisioning
- Agent CLI with setup wizard, logs, and status commands
🐛 Bug Fixes
- Fixed agent registration timeout for high-latency network connections (>500ms RTT)
- Resolved tunnel multiplexing contention when handling 50+ concurrent connections
- Fixed log rotation on agent side consuming excessive disk space for verbose mode
- Addressed agent reconnection exponential backoff starting at too-high initial delay
v1.1.0
✨ Features
- Dashboard Customization — drag-and-drop widget layout, custom dashboard creation, and saved views
- Widget gallery: job throughput chart, queue depth gauge, worker activity heatmap, failed jobs timeline
- Custom date range selectors with preset quick-picks (last 1h, 24h, 7d, 30d)
- Auto-refresh interval configuration per widget (5s, 15s, 30s, 60s)
🐛 Bug Fixes
- Fixed widget resize handler not triggering on touchscreen devices
- Resolved saved view order randomly shuffling after page refresh
- Fixed chart tooltip clipping on small viewports
- Addressed auto-refresh interval resetting to default on theme change
v1.0.0
Launch✨ Features
- Public Launch — QueueHub goes live with real-time BullMQ queue monitoring
- Live dashboard with auto-refreshing queue metrics (waiting, active, delayed, failed, completed counts)
- Job browser with pagination, status filtering, and detailed job inspection
- Queue management actions: pause, resume, empty, and delete queues
- Dark mode and light mode themes
- Responsive design for desktop and tablet
🐛 Bug Fixes
- Initial release — no prior bug history. All feedback welcome!
Stay in the loop
Subscribe to our release announcements and never miss an update. We ship every 6–8 weeks.
Get release notifications