Hi there,

 

i got a problem to get my unraid server to sleep due to not understandable disk activity.

I also asked for help in the "Dynamix Plugin-Section" - but due to a lot of new post my request seems to be overseen.

 

Goal to achieve: unRAID server should go to sleep

Problem: Disk activity prevents sleep and no way (for me) to specify whats the cause of this

 

What I've done:

• Set a global spin down to one hour
• Set a specific spin down for all my disks within my array to one hour, too.
• Moved the docker image location to cache only! (see screenshot docker_image_settings.png)
• Set the docker parameters for each docker to cache only! (see screenshot docker_settings.png)
• Activate the Dynamix Sleep Plugin with the settings shown in my attached screenshot (sleep_plugin_settings.png)
  • Excluded no days
  • Excluded hours from 08:00 until 24:00

 

To make it even more complicate for me the dashboard shows me an inactivity of the complete array at about 23:50.

10 Minutes before requested sleep -- but it dont went to sleep

And - for complete confusion - sometimes it works completly correct (see sleep_works.png) or sometime just particially (see sleep_part.png) from 04:41 to 10:00 (RTC start)

 

Hopefully someone can give me some hints or tipps how to solve this riddle.

Thanks in advance.

No one any idea? Not even a possibility to check what cause the disk activity?

This not understandable process persists ... from 27th to 28th (perfect sleep) - form 28th to 29th or 29th to today (no sleep).

I have never changed anything... drives me crazy. 😕

As an experiment, have you tried disabling your Dockers and see if it will go to sleep properly?

Mhh to be honest not yet due to the correct WebGui information (all disks showing spun down) and the fact that all dockers running on the cache drive. 
 

But if you think it can cause the problem I’ll try it today evening and report tomorrow. 
Thank you and have a nice day

16 minutes ago, Maddeen said:

Mhh to be honest not yet due to the correct WebGui information (all disks showing spun down) and the fact that all dockers running on the cache drive. 
 

But if you think it can cause the problem I’ll try it today evening and report tomorrow. 
Thank you and have a nice day

Just as a part of process of elimination. It's not clear to me that the server will go to sleep with an active docker. That seems dangerous for the server to go to sleep if a docker in the middle of writing to disk.

It's not clear for me either - there is no documentation about that.

And - it already worked on some days with the dockers running - I only have PLEX and Pi-Hole.

 

Sadly there's also a missing of a real analysis tool for this.

The sleep plugin just telling me that "sdd" got "activity" - sorry, but that is not a "debug log" - that information is completly useless due to the missing of details.

 

Anyway - I'll stopp the two dockers before I go to bed and we'll see tomorrow if that was the problem.

Hopefully imho that isn't the solution because then I have to work out a solution that automatically stopping my dockers before going to sleep and start them right after wakeing up.

And for me - it crashes my logic. Because I set up "if the disks of the array (exclude cache drive!) are spun down --> go to sleep.
And all my dockers running on the cache drive".

So where's the problem just working like I set up?  

 

 

Edited June 2, 20206 yr by Maddeen

Update: Yesterday Bonienl (the dev of the plugin) gave me the hint to change the Device activity monitoring to "counters only" and it seems to work. 🙏

I'll validate this within the next days to be sure that my server is going to sleep reliably and will give an update here.

 

Jun  3 00:03:26 UNRAID s3_sleep: All monitored HDDs are spun down
Jun  3 00:03:26 UNRAID s3_sleep: Extra delay period running: 0 minute(s)
Jun  3 00:03:26 UNRAID s3_sleep: Check TCP/SSH/TTY/IP activity
Jun  3 00:03:26 UNRAID s3_sleep: Communication state is idle
Jun  3 00:03:26 UNRAID s3_sleep: Enter sleep state now
Jun  3 00:03:26 UNRAID kernel: PM: suspend entry (deep)
Jun  3 00:03:26 UNRAID kernel: PM: Syncing filesystems ... done.
Jun  3 10:00:10 UNRAID kernel: Freezing user space processes ... (elapsed 0.001 seconds) done.
Jun  3 10:00:10 UNRAID kernel: OOM killer disabled.
Jun  3 10:00:10 UNRAID kernel: Freezing remaining freezable tasks ... (elapsed 0.000 seconds) done.
Jun  3 10:00:10 UNRAID kernel: Suspending console(s) (use no_console_suspend to debug)