How to setup SSDB for HA/failover?

[0xgeert]

It seems SSDB can do replication in a master/slave setup (like Redis). However, I can't find any docs on this. For instance, does SSDB support Sentinel to check for Master failure and do Slave -> Master promotion?

In general I'm looking to implement this setup (which contains Sentinel + HaProxy). Would that work for SSDB?

If not, how do companies using SSDB in production handle this?

[ideawu]

Hi, SSDB is however a single-node storage, it does not concern about sharding, distributing, proxying...

ssdb-cluster is under development.

[ss3ds]

@ideawu clustering is very important for all types of databases

i wish you make it soon

[yehosef]

HA doesn't require clustering, per se. You just need to assign the master and slave and reassign as needed dynamically.

[wangqiang8511]

@ideawu Is ssdb-cluster still under development? Is there anywhere we can find the documentation or design roadmap