What is the recommended replication strategy for OpsCenter keyspace?

0
votes

I'm using OpsCenter to monitor and configure my Cassandra cluster (It's actually a DSE cluster) and I have a keyspace that spans multiple datacenters. The OpsCenter keyspace, which is created and maintained by OpsCenter, use SimpleStrategy as the default replication strategy, which prevents me from turning on its repair service (mentioned in OpsCenter's document).

As the Isolating OpsCenter Performance Data blog says that using a dedicated datacenter requires us to manually monitor and scale the OpsCenter nodes, I was wondering what is the recommended replication strategy and factor for the OpsCenter keyspace so that storing OpsCenter data has limited performance impact on my production nodes while requires minimal tuning when I scale my production datacenters?

Suppose my production nodes use NetworkTopologyStrategy with two datacenters 'Cassandra' and 'Solr' (in a DSE setting) where 'Cassandra' datacenter supports OLTP and 'Solr' datacenter is dedicated for searching. Is it a valid solution to set the replication of OpsCenter keyspace to { 'class' : 'NetworkTopologyStrategy', 'Cassandra': 1}?

Thanks,

Ziju

1
datastax-enterprisedatastaxopscenter

1 Answers

2
votes

NetworkTopologyStrategy is the recommended one for cases like this, so your proposed solution is valid. RF of 1 is debatable, but since OpsCenter keyspace doesn’t usually store anything very important, it should work okay, although I’d bump that to 2.