// queue observability
See why a job is stuck, not just that it is.
Pallet sits next to your existing queue (SQS, Redis, RabbitMQ) and gives you per-job traces, retry history, and consumer lag — without shipping your payloads anywhere.
Start free — 14 daysRead the docs$ pallet trace job_4f21a9
✓ enqueued 12:04:01.221
✓ consumed 12:04:01.804 (worker-07)
↻ retry 1/3 12:04:03.912 timeout: upstream 502
✓ completed 12:04:06.331
total latency: 5.11s across 2 attempts
Built for the 2am page, not the demo
Three things a queue dashboard should actually tell you, and usually doesn't.
01
Per-job traces
Every enqueue, retry, and dead-letter for a single job, without grepping five services' logs.
02
Consumer lag alerts
Know a worker pool fell behind before your queue depth graph even moves.
03
No payload storage
We index metadata only — job IDs, timings, error classes. Your data stays yours.
4ms
median trace ingest overhead
3
supported queue backends today
0
payload bytes ever stored
10min
typical setup time
Point it at one queue. See if it earns the second.
No sales call required for the first workspace. Bring your own SQS or Redis connection string and you're tracing jobs in minutes.
Get an API key