Verification Of Data Integrity Following Dbtier Upgrade Or Maintenance Actions (Doc ID 2970393.1)

Last updated on SEPTEMBER 01, 2023

Applies to:

Goal

 Upgrades to cnDBTier and similar cnDBTier maintenance actions may involve rolling restarts of the cnDBTier pods.  The cnDBTier data, management, and mysql interface pods are deployed in pairs to preserve data across maintenance activities like upgrade, and the upgrade itself is designed to not begin restart of the next pod in a paired set until the previous has completed and recovered from its restart activity.  However it is possible that other activities or errors happening at the same time as the maintenance actions could lead to a situation where two paired pods are down/restarting at the same time.  If data pods enter this condition, recovery should still occur however data may be lost.  At the conclusion of the maintenance action, all pods may show all containers up & in Running state, but if data loss occurred no outward sign may be immediately evident.  This is not a normal, typical, or frequent condition to encounter however to assuage concerns about data integrity over cnDBTier maintenance actions some additional steps may be prudent prior to reinstating traffic or returning the node to service.

Solution

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.