If no replication is used, consistency is guaranteed. Scaling up and down won't affect consistency neither. However it will affect your availability, if one of the masters is gone, your Cluster is down if you use default configuration. This can happen from time to time especially network partition is not a rare event.
Sent from my iPhone
> On Aug 27, 2017, at 1:05, Max Holland <
m...@hollandhome.co.uk> wrote:
>
> We have a slightly odd use case, we are using redis as a data store but don't care too much about the occasional data loss, so currently we are using a single redis instance with no replication. We need a lot of throughput, we are now reaching the point of saturating the network on the biggest redis instance you can provision in AWS elasticache, so we have to look at redis partitioning options.
>
> I've been reading this tutorial on redis cluster
https://redis.io/topics/cluster-tutorial, I'm wondering if we're not using any replication, will we be able to achieve strong read write consistency? Or do you still run into consistency issues without replication at times of scaling up/down for example?
>
> --
> 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+u...@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.