Change Details

As we move towards implementing the new solution, we should reappraise our storage requirements. Currently we have 9 Postgres nodes and 9 Cassandra instances in each datacentre. This is almost certainly overkill and we should reexamine the capacity we require, and how those hosts are configured. We should also examine capacity requirements for moving to Redis - can we simply migrate to using the RDB clusters in existence or will our significant volume of data cause issues?

As we move towards implementing the new solution, we should reappraise our storage requirements. Currently we have 9 Postgres nodes and 9 Cassandra instances in each datacentre. This is almost certainly overkill and we should reexamine the capacity we require, and how those hosts are configured. We should also examine capacity requirements for moving to Redis - can we simply migrate to using the RDB clusters in existence or will our significant volume of data cause issues? ==== Acceptance Criteria - [] Documentation is available on wiki under https://wikitech.wikimedia.org/wiki/Maps

As we move towards implementing the new solution, we should reappraise our storage requirements. Currently we have 9 Postgres nodes and 9 Cassandra instances in each datacentre. This is almost certainly overkill and we should reexamine the capacity we require, and how those hosts are configured. We should also examine capacity requirements for moving to Redis - can we simply migrate to using the RDB clusters in existence or will our significant volume of data cause issues? ==== Acceptance Criteria - [] Documentation is available on wiki under https://wikitech.wikimedia.org/wiki/Maps