OOF seemed a bit slow earlier, so I just assumed the hamster in its yellow jacket had escaped and was rioting in Paris all weekend.
We're currently back on SAS spinny drives for a bit, following a controller glitch on the SSDs (which are SATA)
Scratch chin, why would SSDs make a difference as I would expect the reverse with slower HDDs with possible queue buffer overflow?
It looked like the SATA controller lockup, rather than the SSD, but hard to prove, hence having non critical stuff on the SATA SSDs to get confidence back. Some of HPE's firmwares over the past couple of years do leave a lot to be desired
So we're on the SAS controllers, well away from anything SATA, for a bit. None of the SAS stuff has shown any issues.
We do not use SATA hard drives, because they are universally rubbish when it comes to performance - I guarantee if I swapped out the SAS for SATA, we'd see disk latencies measured in seconds, once the caches were exhausted.