SMART Poll CPU spikes, SAS Drives wont Spin Down


jbeezy
Go to solution Solved by trurl,

Recommended Posts

First time noob, first time setup. I've gotten v6.11.3 running on an R610, with a LSI 9200-E controller (IT mode), connected to a SC200 array of drives. Basic functionality seems to be working fine, except every 30 seconds my CPU cores would spike.

 

I figured out what was causing this- it was the SMART polling. I figured this out by changing the `Tunable (poll_attributes)` from the default of 30 seconds, to 60. Then I could clearly and immediately see, based on the timing of the CPU spikes, that this was the activity making the CPU spike.

 

When I noticed that the drives weren't spinning down, I realized there was something else going on. I discovered in some old forum posts that unRAID was unable to command SAS drives to spin down. So, my drives are spinning 24/7, and since they're awake, the SMART Polling is occurring every 30 seconds, which for some reason spikes the CPU.

 

My main concern now, is that the drives won't spin down. My guess is that if the drives were spun down, the SMART polling would not be occurring during that time anyways, so I'm ignoring the CPU spike at the moment and just trying to get the drives to spin down on their own.

 

I have a clean environment with VMs and extras turned off, to make easier troubleshooting.

 

I am a little unclear from old forum posts as to what is going on. There was an old plugin that addressed the situation but from what I read, I think the newer unRAID versions incorporated the ability to talk to the SAS drives. I also haven't found that plugin updated since about August 2020, so I stayed away from that.

 

Came here hoping someone can tear apart my diagnosis files and offer some advice on what might be happening here.

 

 

 

 

storage-diagnostics-20230225-1819.zip

Link to comment
11 minutes ago, jbeezy said:

First time noob, first time setup. I've gotten v6.11.3 running on an R610, with a LSI 9200-E controller (IT mode), connected to a SC200 array of drives. Basic functionality seems to be working fine, except every 30 seconds my CPU cores would spike.

 

I figured out what was causing this- it was the SMART polling. I figured this out by changing the `Tunable (poll_attributes)` from the default of 30 seconds, to 60. Then I could clearly and immediately see, based on the timing of the CPU spikes, that this was the activity making the CPU spike.

 

When I noticed that the drives weren't spinning down, I realized there was something else going on. I discovered in some old forum posts that unRAID was unable to command SAS drives to spin down. So, my drives are spinning 24/7, and since they're awake, the SMART Polling is occurring every 30 seconds, which for some reason spikes the CPU.

 

My main concern now, is that the drives won't spin down. My guess is that if the drives were spun down, the SMART polling would not be occurring during that time anyways, so I'm ignoring the CPU spike at the moment and just trying to get the drives to spin down on their own.

 

I have a clean environment with VMs and extras turned off, to make easier troubleshooting.

 

I am a little unclear from old forum posts as to what is going on. There was an old plugin that addressed the situation but from what I read, I think the newer unRAID versions incorporated the ability to talk to the SAS drives. I also haven't found that plugin updated since about August 2020, so I stayed away from that.

 

Came here hoping someone can tear apart my diagnosis files and offer some advice on what might be happening here.

 

 

 

 

storage-diagnostics-20230225-1819.zip 90.49 kB · 0 downloads

As trurl has posted link you need to install doron spin down plugin for sas it was updated in Aug 2022.

Link to comment

Posting back for the benefit of others -- this plugin worked to resolve the SAS Disks not spinning down (which I got off the Community Applications after confirming it matched the latest github version).

 

I set the spin down delay to 15 minutes, then installed the plugin, then rebooted the server.

 

I waited until 15 minutes elapsed, and then I saw the shutdown command via the syslog and now I can see visual confirmation in the GUI.

 

I can also see that the CPU spikes from the SMART polling will stop once the disks are spun-down. 

 

Thanks all!

 

 

Screen Shot 2023-02-25 at 7.01.28 PM.png

Screen Shot 2023-02-25 at 7.01.56 PM.png

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.