Redis Dead Letter Queue

[Paulo Roberto Silva]

Hello!

I would like to know if Redis supports dead letter queue, similar to RabbitMQ https://www.rabbitmq.com/dlx.html.

Thank you

[Pierre Chapuis]

Redis is not a background job system, so no.

Background job systems implemented using Redis can support such a mechanism, e.g. Resque does.

Disque, which is written by the same author as Redis, uses a different approach.

[Paulo Roberto Silva]

Thanks Pierre!

Could I do something like that using Celery + Redis?

--
You received this message because you are subscribed to the Google Groups "Redis DB" group.
To unsubscribe from this group and stop receiving emails from it, send an email to redis-db+unsubscribe@googlegroups.com.
To post to this group, send email to redi...@googlegroups.com.
Visit this group at https://groups.google.com/group/redis-db.
For more options, visit https://groups.google.com/d/optout.

--

Paulo Roberto Silva  

(12) 99221-8616 - Twitter | Linkedin | Github | Blog | Izepa

[Salvatore Sanfilippo]

Hello,

Basically Redis has primitives in order to create a dead letter
feature, but the fact it is available or not depends on the library
implementing queues that you are using.

Libraries based on BRPOPLPUSH command or that in general do not remove
the job from the queue before acknowledged processing may have such
feature (sometimes queue systems on top of Redis are basically
implemented in Lua and do many funny things to create new features).

In the case of the new Stream data type of Redis 5, the consumer
groups have explicit dead letter features, you can grep the Redis
streams intro (https://redis.io/topics/streams-intro) for dead letter,
but it's better that you read it all if you plan to use streams.

Cheers,
Salvatore


--
Salvatore 'antirez' Sanfilippo
open source developer - Redis Labs https://redislabs.com

"If a system is to have conceptual integrity, someone must control the
concepts."
— Fred Brooks, "The Mythical Man-Month", 1975.

[Salvatore Sanfilippo]

> Disque, which is written by the same author as Redis, uses a different approach.

Yep, and the same idea was borrowed for the stream data type basically.