Hi all,
We've got 3RDS servers that are using FSLOGIX and I've split the Office 365 and Profile containers onto different RAIDS on one of our servers and for a while we've been getting the blue circle on nearly everything (Outlook, Excel, SharePoint etc) and this lasts anything up to a couple of minutes so I've checked the disks and found that mostly (but not always) when this happens that the disk queue length spikes to around the high 70's (sometimes 200's) and then after a while it drops.
I've changed disks and servers, made sure this is the only VM on the physical server, changed RAID types from 10 to 5 to mirrored to test and it's the same all the time however the office 365 disks never go above 0.02 ? The CPU never goes above 30% and the same with the RAM.
The profiles disks queue length sit about 0.2 or 0.5 when they're not spiking and when they do spike I've noticed the top "talkers" still change (i.e. so it's not one profile that sits at the top spot until the spike goes) and the read/writes don't correlate with the spikes either (i.e it could show a massive amount of read/writes and there'd be no spike but then the read/writes could be really low and it might then spike)
The problem is that as the profiles are VHDX then i can't see exactly what is happening within the disks at the time in that i suspect that it's something within the profile disks with an action the users are doing but i can't see it.
I could do with knowing if anyone else has a spike on their disks (every 20 minutes or so) and if their setup seems ok so i know it's "normal" (and yes, i know each setup is different)
Any advice on where to look to see whats causing the blue circle so much ?