According to this Microsoft article, caching is not recommended on disks hosting SQL/Server log files. If the workload is heavily utilizing the underlying disks (any other application than SQL/Server), disk caching should also be set to none.
Please refer to https://stackoverflow.com/questions/39623778/why-do-best-practices-tell-you-to-disable-caching-on-the-log-drive-of-sql-server for more information.
To change the value:
- Go to the specific instance
- Select disks
- Click on the name of the data disk
- Change host caching to the specific value
- Click save
When tested on both cluster nodes with large numbers of files copied onto a 100GB volume, no degradation was detected. We also performed switchovers and paused and continued the mirror and no degradation was detected.
Thanks for your feedback.
Post your comment on this topic.