It think the answer comes down to what type of disk you have deployed and sort of disk I/O your are experiencing.
Fragmentation mostly affects mechanical disk, where latency is added by additional physical effort related to the reading a fragmented file from disk. The file size and how much it is fragmented also affect the observed performance. I have never experienced fragmentation related performance issues on SSD disks.
End-user triggered calculation and interaction runs in-memory, so slower disk will not typically impact end-user perception. Qlik Sense will access files when apps are opened, edited or reloaded. So it might be worth making some observations about response times, and evaluate if you perceive it as an issue or not. Then defrag and compare the results.
No, because Qlik Sense is a In Memory BI 🙂 Acceleration of the disk subsystem will not have a strong impact on the operation of the business analysis system
It depends on the serer side activities. You have to keep in mind that there is still a significant amount of disk I/0 for app and QVD files during opening closing a reload. Repository database also utilizes disk. I would not neglect the risk with mechanical disk, but also recommend SSD disks to performance optimize disks.
