Hi all,
We have multiple Pega clusters for different environmets, we are trying to connect multiple Pega Clusters to a single Cassandra ring, can someone clarify the potential issues of this implementation ?
For example, Cross-cluster data corruption, query failures, etc.
@MiguelHenriquez Pega platform uses a defined set of keyspaces, so OOTB, if you would connect multiple platform installations to the same external Cassandra cluster, then the platform instances would overwrite each other’s data (using the same keyspaces).
However, you can prevent that from happening by setting a prconfig on each environment: dnode/keyspaces_prefix. By defining a unique prefix for each environment, each one would create its own set of keyspaces and would not access each other’s data (keyspace names will no longer clash).
Implication to keep in mind is the potential impact of one environment on another, i.e. if there’s a significant load on Cassandra generated by one of the environments, it can affect query performance of another.
@wtekiela
I have set the prefix in the prconfig as you suggest however i see unauthorized exception in the logs .. User xyz has no CREATE permission on or any of its parents. I have verified that the user has all permissions on the custom keyspaces. Any thoughts ?