What is Unraid doing during array health test (array health report notification)?

[Nostromo]

Hi guys,

 

New Unraid user here, coming from TrueNAS. I wanted to get more from my server so moved on recently.

 

After basic setup I am trying to focus on the proper schedule maintenance but still a bit lost. Going to the point, I am trying to understand what Unraid is checking when I receive the "array health report". I know how to schedule but despite checking forum and documentations could not find what is happening. Could you help me?

 

Extending my question, I would like to prepare a good maintenance schedule. As the scrub test is not existing in xfs and it is not clear for me if it is worth to do regular extended SMART test, understanding array health check / report can help me to decide.

 

Thanks in advance!

 

 

[itimpi]

6 minutes ago, Nostromo said:

Extending my question, I would like to prepare a good maintenance schedule. As the scrub test is not existing in xfs and it is not clear for me if it is worth to do regular extended SMART test, understanding array health check / report can help me to decide

For the array disks you normally want a periodic parity check (Settings->Scheduler).  A monthly schedule is not atypical, and it is recommended that you set such checks to be non-correcting to avoid a disk exhibiting problems ending up meaning parity is not completely valid (only run correcting checks if you are reasonably certain you do not have any drives mis-behaving).  You may also want the Parity Check Tuning plugin to give more control over such checks.

 

I think the

11 minutes ago, Nostromo said:

After basic setup I am trying to focus on the proper schedule maintenance but still a bit lost. Going to the point, I am trying to understand what Unraid is checking when I receive the "array health report". I know how to schedule but despite checking forum and documentations could not find what is happening. Could you help me?

 

I think that basically all this is doing is telling you that no drive is currently showing read or write errors.   I am reasonably certain it does NOT mean that all drives would  pass short or extended SMART tests if you ran them.

[Nostromo]

Thank you very much! Then I guess that I should add some SMART test to the schedule for completing the parity check.

[JorgeB]

Just now, Nostromo said:

Then I guess that I should add some SMART test

Parity check is basically the same as an extended SMART test, you can add some SMART shorts test in between the recommended monthly check but unlikely those would find any issues without you been notifying of something like pending sectors or similar.

[itimpi]

33 minutes ago, Nostromo said:

Thank you very much! Then I guess that I should add some SMART test to the schedule for completing the parity check.

I would not bother - the parity check already reads every sector on all the drives so is equivalent to the SMART long test

[Nostromo]

Oh ok good, that simplifies a lot the procedure.

 

Many thanks to both for the fast replies