Databases@Powered by Passion
The DR site faced dual disk failure, causing database and cluster failures. ASM issue triggered rebalancing, leading to additional disk failures. Recovery attempts failed, resulting in loss of standby databases and DATA disk group.
On a typically busy Monday at the office, a storage server experienced multiple disk failures, causing a reboot and loss of grid disks. Investigation revealed cacheContentLost status due to a flashdisk failure and import failure. The team resolved the issue by recreating the grid disks and replacing the faulty disks. To prevent future occurrences, they updated the Exadata server image. The problem has not recurred since.
The MDRT_$ tables in Oracle Spatial are crucial for spatial indexes and should not be moved, altered, or replicated. It’s possible to drop unrelated tables if using Oracle DB version 12cR2 or later. While MDXT_# tables, which supported spatial index statistics, can be dropped after use, changes to the tablespace for MDRT_# tables are permitted. Queries are provided to identify and clean up unnecessary tables, resulting in a significant reduction in table count.
After applying release update 19.13, a standby site cluster experienced Node Evictions and Reboot Problems. Investigation revealed high CPU usage by the CVU (Cluster Verification Utility) as the root cause. A related bug was identified and fixed in 19.20 Release Update.
Osman’s DBlog 