Going for a cluster

[Oscar Besga]

Hi all !

I'm planning to migrate from a redis-sentinel servers to a clustered one.

Our keys have the format  

pool:charge:{id.XXXXXX} 

so I do not expect problems for this side.

But one operations rely on a complete SCAN for all the keys with the pattern

 pool:charge:{id.*}

and I think this will not work in a cluster

Here I found some code that might work

https://www.adictosaltrabajo.com/2019/03/05/escaneo-de-claves-en-un-cluster-redis/

(spanish)

Basically, iterate for every node doing the SCAN operation there and aggregating results.

What do you think of this code ?

Tks !