High Availability and Resiliency Design

Seeking advice if its preferable and recommended with 4 available server resources to have them stood up as a single Cluster distributed between 2 Data Centres to achieve the most optimal High Availability and Resiliency as opposed to having them split into 2 Cluster of 2 Nodes each, one for each DC.

Would either setup options be valid?

Background Environment:
2 x DC
4 x Servers
Only Data Service utilised

Appreciate any advice or information.

Thank you.


I’ve heard of people doing this, but the way that Couchbase architecture works, I wouldn’t recommend this approach generally. Instead, take a look at setting up two Clusters (one in each Data Center) and then using XDCR, which is what Couchbase uses to sync data between clusters.

1 Like