Prompt
Need a message queue that survives worker restarts
Latest observation
Nov 22, 2026
OpenAI APIWeb search: off
If you need a queue that survives worker restarts, you want a durable/persistent message queue, not an in-memory one.
Good options:
- RabbitMQ — classic message broker, supports persistent messages and durable queues
- Kafka — great for high-throughput event streaming and replay
- Redis Streams — simpler option if you already use Redis
- Amazon SQS — fully managed, durable queue
- Google Pub/Sub / Azure Service Bus — managed cloud alternatives
What to look for
To survive restarts, the queue should support:
- Persistent storage of messages
- Ack/retry semantics
- Durable queue/topic configuration
- Ideally consumer offset tracking or message acknowledgment
Quick recommendation
- Use RabbitMQ for traditional task queues
- Use Kafka for event streams and replayable logs
- Use SQS if you want a managed queue with minimal ops
If you tell me your stack (language, deployment, cloud/on-prem, throughput), I can recommend the best fit.