Data Replication best practices in the B2C Commerce platform

Data Replication Best Practices

Note: Also see additional best practices in Replication Best Practices in B2C Commerce.

1. Replications should be run during low traffic hours since this is a database intensive process. We recommend running this during late night or early morning hours.
2. No manual edits, updates or changes should be made in business manager on target or source system when replication process is running. Also, make sure there are no running jobs on the target system on which the data is being replicated.
3. Scheduled search index rebuild jobs should be disabled on target and source system if you are replicating search indexes (incremental indexing can run on production.)
4. All data must be staged to development first, then data should be verified. For example, when replicating a search index to production, verify that search is working after replication. If replicating content or products make sure they are working as designed after replication.
5. The same data also needs to be verified after it is staged to production.
6. Only one replication process should be run at a time. Multiple overlapping processes will cause issues.
7. There is an undo feature that can roll back the last replication that was run. (Note this can only undo the last process, which is another reason to verify data.) If your replicated data is not working as designed, you can revert the application back to the state it was in before replication was run by using this process.
8. Staging logs should also be checked during replication runs. They exist on both the target and source system. They will provide details on root cause if your replication is in a failed state.
9. Users should hit Create to create a staging process instead of start in the wizard.
10. If replicating multiple objects within a task and the process fails, try to create a process with only one object and see if it completes. The staging logs may also shed light on the group that caused the failure.
11. Search the Salesforce Commerce Cloud knowledge base for the root cause of any replication failure error you obtain from the logs.
12. Cache should never be cleared directly after a replication finishes. The last step of replication automatically clears the cache so this should not be done manually.
13. No manual edits on the cartridge path or code should be done directly in production. Both code and preferences should always be replicated from Staging to Production. Manual edits can result in unforeseen behavior such as two different versions of a page.
14. Schedule replication after 30 mins of maintenance window completes to avoid any issues.