Understand reasons for performance issues in huge SVC environments (part 1)
mipi 270004DGB0 Visits (1902)
I just finished an performance bottleneck analysis for an 8 nmode cluster with several 100 TB and found one thing which I think might be a commonly made mistake in many of the bigger SVC environments. This picture shows the MDisk performance of SSD storage which is not acceptable. We have 37 ms latency on 37000 IOPS which is a bad value for this technology.
This picture shows one common reason, for bad storage performance whilst SVC and the backend storage look bored.
This picture shows that the "in queue latency of the mdisks" are far too high. The best value for the "in queue latency" is zero. It just should not exist at all or it should be less than 0.04 ms. There are only some reasons, why this in queue latency is too high.