HPE Storage Users Group

I am curious if 16.0 is high for PD total queue length? I am seeing in system reporter the host total queue length as 3.0?

That depends on several factors and what report you are looking at.

For example: Daily/Hourly/Hi_res granularity is important to note as the data (16 PD queue length) gets averaged as you roll up.... obviously a 16 QD on an Daily report is very concerning, but a spike on a hi-res collection is much less concerning.

Also what is the "Compare:" set to? If you are comparing PDs, its important to note whether its just one or two disks occasional spiking, versus a large group of drives all maintaining a high queue depth... if your "Compare" is default at none, then you need to understand that 16 is the sum of all PDs queue depth, which comparatively would be pretty low.

Thank you for the response. So I am just getting use to System Reporter, but I went and ran a, "Latest Hourly Summary - Dynamic Charts (experimental)" just now and saw PD Total Queue Length of 167.0?!

We only have 64, 600TB FC disks in it, but that just seems insane to me?! I am not seeing the compare option when running this report, so I am going to play around with running my own custom report more....

Anything else I should be watching for that could play into this?

EDIT: I just ran a high resolution for past 24 hours, with all PDs - Dynamic Charts (experimental) with compare = none....

I am seeing a lot of peaks and valleys with most going up and down from ~100 to 0, but with 2-3 spikes over 200!