Taking care of your Chef Server Best practices and Recommendations.

This page should eventually be published on the docsite.

Routine tasks for the best results from infra server:

1. Limit the number of cookbook versions uploaded to the server.
2. Clean up the orphaned authz_ids that might be left behind as a result of creating and deleting chef objects. Before the version 14.15, only the authz_ids for the orphaned users were deleted from the bifrost tables. But multiple biforst ids were created as a part of the object creating process. Run chef-server-ctl cleanup-bifrost as a one time operation to delete these orphaned ids. 14.15 onwards the deletion of a chef user will cause all of the related authz ids to be deleted as a cascade effect.
3. Vacuum full the postgres database on upgrade or on deleting a large number of chef-objects.
4. For optimum performance - number of various pool workers
5. Scheduling of client runs.