How z4j stacks up.
Most task-queue dashboards cover one engine and stop at 'view.' z4j covers six engines and keeps going: actions, schedules, RBAC, audit trail, redaction. Here is how we compare, honestly.
vs. Flower
Celery only Flower
The classic Celery viewer. Built in 2011.
- Positioning
- Read-only viewer
- License
- BSD-3, open source
vs. RQ Dashboard
RQ only RQ Dashboard
The simple web view for RQ.
- Positioning
- Lightweight viewer with basic requeue
- License
- MIT, open source
vs. Grafana + Prometheus
Any engine, via custom exporters Grafana + Prometheus
The DIY metrics approach. Great for SRE, wrong tool for task ops.
- Positioning
- Metrics and alerting, not task-level control
- License
- AGPL-3 (Grafana) + Apache-2 (Prometheus)
How we compare
One table, the whole field
A high-level capability row. Click any dashboard above for the deep dive.
| Capability | Flower | RQ Dashboard | Grafana + Prometheus | z4j |
|---|---|---|---|---|
| Engine coverage | Celery only | RQ only | Whatever you export yourself | 6 engines |
| Persistent history | In-memory only, lost on restart | Only what RQ keeps in Redis (TTL-bound) | - | Postgres-backed |
| Retry / cancel actions | None, strictly a viewer | Requeue failed only | None - read-only dashboards | Universal |
| RBAC / multi-user | Single basic-auth user | Basic-auth or none | - | Built-in |
| Audit log | None | None | None | HMAC-chained |
| Historical trends | None | None | - | Built-in |
This is a summary. Each dashboard's page has the full capability matrix with honest trade-offs.
Still on one of these?
Install z4j alongside your existing dashboard. Run both for a week. Keep what you prefer.