CPU goes to 100% waiting for drive(s) to spin up?


Recommended Posts

Folks,

 

Recently I've been seeing random "freezes" in my Unraid server performance. They look like momentary "choking" when streaming from the server halts for a few seconds. Then it resumes, problem goes away, until next time.

 

It finally became annoying enough to look into. It seems as if from time to time the server goes to 100% CPU for no apparent reason (no real load, no VMs, couple dockers that do backup work in the wee hours, no RAM shortage), for several seconds, and then it goes back to to normal (which is like 1%-6%). 

I seem to be mapping these spikes to spin-up of HDDs. It appears as if the system waits for the drive to spin up in a busy loop. I mean this does not make a lot of sense, but it looks that way.

Now: it does not happen every time a drive needs to spin up. In fact when the disks spin up for an SMB or NFS request - no problem (and CPU stays at single digits). Might be happening when a docker spins up a drive. 

 

My environment: Unraid 6.6.5 running as a VM under ESXi. HDDs passed through. Two parity drives, three data drives, all 4TB. Array is LUKS encrypted.

 

Thoughts, anyone?

 

Thanks!

Link to comment
9 hours ago, trurl said:

Moved to the proper forum

@trurl, I'm not sure Virtualizing Unraid is the "proper forum". The fact that my server runs as a VM may or may not be related to the problem at hand. Analogously, I'm sure that the fact that someone's drives are WD wouldn't make you move her posts to a WD-specific forum...

Link to comment
4 hours ago, doron said:

The fact that my server runs as a VM may or may not be related to the problem at hand.

If the problem is documented while running bare metal, then general support IS the proper forum. If you are only running virtual, then this is the proper place for troubleshooting.

 

If you wish to get support from limetech, your first troubleshooting step is to switch to running bare metal and replicate the issue. Until you can document that it happens bare metal, it's a possible virtualization issue which limetech doesn't directly support. You've been here long enough to know that.

 

 

 

 

Edited by jonathanm
  • Like 1
Link to comment
1 hour ago, jonathanm said:

You've been here long enough to know that.

I've been here long enough to remember being able to ask the community for assistance and insights, even for configurations that are not directly supported by Limetech. I should know, I've been there, on both ends.

 

I guess things have changed. Which is fine, your forum - your rules. Thanks for clarifying.

Link to comment

This subforum is still part of the community, and all other forum users have full access to it.

 

More importantly it is the community of other users who are virtualizing Unraid. I have moved other posts here in the past so they would get more visibility to that community.

 

I personally am not confident I have any ability to provide support for those users.

Link to comment
24 minutes ago, uldise said:

Which ESXi version? how your drives are connected to the system? how you pass-through them?

ESXi 5.5. Drives are SATA, connected to an onboard SAS controller on the X10SL7.

Passed through to the VM (done awhile ago, twas unRAID 5...) by creating an RDM VMDK for the real drive and passing it to the VM.

Edited by doron
Link to comment

your ESXi looks a bit old, may be there are some incompatibilities with newest unraid Linux kernel, but i'm not sure..

if you can, try to update it to version 6.5, but keep in mind that legacy c# ESXi Client will not work anymore and you need to use browser to administer your server..

Link to comment
Just now, uldise said:

keep in mind that legacy c# ESXi Client will not work anymore and you need to use browser to administer your server..

Yep - one of the reasons I kept it at 5.5 until now 🙂  Plus that it wasn't broke.

It's gonna be an interesting challenge to move the whole setup to the new ESXi (several guests in there). Thanks for the suggestion.

Link to comment
19 minutes ago, uldise said:

my ESXi server are on version Imageprofile ESXi-6.5.0-20170404001-standard (Build 5310538), and i can still use old Client. if you switch to something newer, then your legacy client is gone..

Hmm I don't seem to be able to download this particular version (aka 6.5.0d). Only lets me download an earlier or later one. Oh well.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.