A systems administrator is investigating a significant performance degradation on a database server. The server, which hosts its database on a hardware RAID 10 array, has been experiencing extremely slow I/O write speeds since a brief, unexpected power outage in the data center last week. The server rebooted without any explicit errors in the OS event logs. System resources like CPU and memory utilization are within normal ranges. Which of the following is the MOST likely cause of the slow write performance?
The filesystem on the array has become corrupted.
The array was rebuilt using mismatched drives.
There is insufficient disk space on the logical drive.
The RAID controller's write-back cache is turned off.
The correct answer is that the RAID controller's write-back cache is turned off. RAID controllers often use a volatile memory cache to improve write performance by acknowledging a write operation as complete before the data is physically written to the disks (write-back caching). This cache is typically protected by a Battery Backup Unit (BBU). If the server experiences a power outage and the BBU is old or has failed, the controller will detect this upon reboot. As a safety precaution to prevent data loss from the unprotected cache in case of another power failure, the controller will disable write-back caching and switch to a much slower, but safer, write-through mode, where data is written directly to the disks. This behavior perfectly matches the symptoms of a sudden, drastic drop in write performance following a power event.