Scheduling database maintenance

Database maintenance tasks; for example, defragmentation, index rebuilding , and statistics recalculation, should not be scheduled at the same time as auto discovery , automated user administration , automatic assignment or other instance processes which update a lot of records in many tables. If the tasks and processes run at that same time they will lock each other’s processes, resulting in the instance process, database maintenance, or both, to hang.

When performing long-lasting database maintenance; for example, index rebuilding, OS or database patching that is expected to take more than 10-20 minutes it is recommended to stop updating the database.

Do not schedule database maintenance at the same time on all databases which serve the same application.

If database maintenance lasts more than a few minutes, schedule the Bravura Security Fabric nodes that use that database to be taken out of the load balancer and all Database services from the nodes connecting to that database must be stopped .

It is a good idea to perform maintenance (for example OS upgrades) on the Bravura Security Fabric node itself at the same time as it is done on its database (if so, take down the Bravura Security Fabric node before the database one, and bring up the database before the Bravura Security Fabric ).