During peak business hours, users report that copying files to a Windows Server 2022 SMB share now takes several minutes instead of seconds. The rack-mount server uses a hardware RAID controller with four 10 000 RPM SAS drives in a RAID 10 virtual disk. Windows Resource Monitor shows an average disk queue length above 50 and throughput never exceeds 20 MB/s, while CPU, memory and network utilization remain low. The RAID management utility logs the message "Controller write cache policy changed to write-through; battery charge capacity below threshold."
Which action should you take first to restore normal disk I/O performance without jeopardizing data integrity?
Reformat the volume with a 64 KB NTFS allocation unit size to reduce fragmentation.
Disable SMB signing on the file server to lower protocol overhead.
Run an offline disk defragmentation and disable 8.3 filename generation on the file server.
Replace the RAID controller's battery pack and re-enable write-back caching on the virtual disk.
The controller switched from write-back to write-through mode when its battery lost the ability to protect cached data. Write-through forces every write to be committed to the disks before completion, severely limiting throughput. Replacing the failed battery (or allowing it to complete a relearn cycle) lets you safely re-enable write-back cache, immediately restoring the array's expected performance. Changing allocation unit size, defragmenting the volume, or disabling SMB signing might produce marginal gains in specific scenarios but will not overcome the fundamental write-through bottleneck caused by the disabled cache.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
What is the difference between write-back and write-through caching?
Open an interactive chat with Bash
Why does a RAID controller need a battery for write-back caching?