#Filesystem Analytics and Performance
1 messages · Page 1 of 1 (latest)
Depending on the size of the volume and amount of data it can cause an impact.
As an example, we have a volume that is about 22TB in size, serving around 130million files, mostly small files under 3k each.
When we enabled FSA on this volume, it did take a hit, not a huge one that was noticed by end users, but the system noticed.
Turn them on one volume at a time if possible, and for larger ones it'd be good to do it in a test environment if possible or during a maintenance of some kind
we have had customers see massively increased latency (with peaks up to 1000ms) where NetApp support told us it is because of the initial scan and we should disable that. After disabling FSA the latencies returned to normal. It was some time ago so it might have been a bug or issue with an older ONTAP release but yeah, basically do proper monitoring and if people complain that everything is slow, the first thing you should try is turning it off again.
after the initial scan is completed there is no noticeable impact on performannce anymore
Thanks. Just what i was looking for. We have a very large volume (100TB) everyone uses and the one we could benefit most from the analytics as we break it up into nested volumes. I had turned it on briefly on a volume with much less traffic last month on 9.8 and my spidey senses tingled over how widely it could be used.
yeah, just turn it on and see, if the perf impact is too much you can always cancel it and re-run the scan on a weekend or so, when there is less impact. We also have many customers who enabled it on their (large) volumes and who didn't see any impact even though the scan took days to complete. I guess it depends on many factors 🤷♂️
9.14.1 fixes FSA performance. @distant heron thank you for the pointer.