[Plugin] Parity Check Tuning

[dopeytree]

Ah ok sorry about that, swapped to testing for the attached.

 

moulin-rouge-diagnostics-20230912-1519.zip

Edited September 12, 2023 by dopeytree

[itimpi]

Thanks - I see what is the issue now and have reproduced it?     That notification is erroneously being generated as you are stopping the array without doing a full shutdown or reboot as I am getting an event from Unraid I did not expect during the stopping sequence.   All my recent testing concentrated on using the Stop or Reboot buttons to get a tidy shutdown or deliberately forcing an unclean shutdown.  I had been a bit lazy and skipped doing the simple array stop/start thinking it was not relevant to unclean shutdowns .  

 

Can you confirm that you are getting the notifications while stopping the array?

 

I will work on an update that will fix this issue.

 

 

 

[Ruato]

14 minutes ago, itimpi said:

Thanks - I see what is the issue now and have reproduced it?     That notification is erroneously being generated as you are stopping the array without doing a full shutdown or reboot as I am getting an event from Unraid I did not expect during the stopping sequence.   All my recent testing concentrated on using the Stop or Reboot buttons to get a tidy shutdown or deliberately forcing an unclean shutdown.   Can you confirm that you are getting the notifications while stopping the array?

 

I will issue an update shortly that will fix this issue.

 

Great that you found something. For me, the unnecessary notification came after updating my server to 6.12.4. That is,
1) Run the update to 6.12.4
2) Shutdown
3) Boot up the server
4) Start array manually
--> notification

[itimpi]

17 minutes ago, Ruato said:

 

Great that you found something. For me, the unnecessary notification came after updating my server to 6.12.4. That is,
1) Run the update to 6.12.4
2) Shutdown
3) Boot up the server
4) Start array manually
--> notification

Going through a reboot sequence should not generate the notification spuriously so it is possible in your case the notification was genuine.    The issue I have found relates to stopping and then restarting the array without going through a shutdown or reboot sequence.

[Ruato]

14 hours ago, itimpi said:

Going through a reboot sequence should not generate the notification spuriously so it is possible in your case the notification was genuine.    The issue I have found relates to stopping and then restarting the array without going through a shutdown or reboot sequence.

 

I will test the this with the upcoming new version as soon as I get back home. I will stop the array before shutdown to guarantee clean shutdown.

[itimpi]

4 hours ago, Ruato said:

 

I will test the this with the upcoming new version as soon as I get back home. I will stop the array before shutdown to guarantee clean shutdown.

 

Just posted the update.   If it still goes wrong then try and get me diagnostics with the Testing mode logging active in the plugin.

[itimpi]

On 9/12/2023 at 3:20 PM, dopeytree said:

Ah ok sorry about that, swapped to testing for the attached.

 

moulin-rouge-diagnostics-20230912-1519.zip 205.93 kB · 1 download

 

Just posted the update which should fix this issue.   If it still goes wrong then try and get me diagnostics again with the Testing mode logging active in the plugin.

[itimpi]

@Ruato @dopeytree

Just thought I would check whether the lack of new reports is good sign and means your issues are now resolved since the last release of the plugin, or is that you have simply not been in a position where a check is running so you would notice if there is still a problem.

[dopeytree]

Seems good have been powering down and restarting the array a lot lately while tuning some settings. Thanks for the fix 🙂

[itimpi]

6 minutes ago, dopeytree said:

Seems good have been powering down and restarting the array a lot lately while tuning some settings. Thanks for the fix 🙂

Thanks for the feedback.   

 

Always glad to have confirmation that the issue seems to have been resolved and something new has not immediately popped up  

 

[Ruato]

On 9/21/2023 at 11:29 AM, itimpi said:

@Ruato @dopeytree

Just thought I would check whether the lack of new reports is good sign and means your issues are now resolved since the last release of the plugin, or is that you have simply not been in a position where a check is running so you would notice if there is still a problem.

 

Have been travelling (too much) and have not been able to test this. Very good that @dopeytree confirmed the fix. Great work @itimpi!

[SirCadian]

Quick question.  I've got my Parity Check set to run once a month and Parity Check Tuning ensures it runs between 1am and 9am each day.  It's also set to pause when Appdata Backup is running (@4am each day) but it does not automatically resume once Appdata Backup concludes (usually around 5am each day).  The tooltip suggests that it should resume as it is still within the increment window of 1am to 9am.  Have I misunderstood?

[itimpi]

10 hours ago, SirCadian said:

Quick question.  I've got my Parity Check set to run once a month and Parity Check Tuning ensures it runs between 1am and 9am each day.  It's also set to pause when Appdata Backup is running (@4am each day) but it does not automatically resume once Appdata Backup concludes (usually around 5am each day).  The tooltip suggests that it should resume as it is still within the increment window of 1am to 9am.  Have I misunderstood?

It should, and if it is not that is a bug.   It has been resuming in my test environment.

 

if you want me to try and work out why it is not working for you then enable the testing mode logging in the plugin settings and then post your diagnostics after you have managed to recreate the issue.

[SirCadian]

3 hours ago, itimpi said:

It should, and if it is not that is a bug.   It has been resuming in my test environment.

 

if you want me to try and work out why it is not working for you then enable the testing mode logging in the plugin settings and then post your diagnostics after you have managed to recreate the issue.

Will do. Thanks. 

[itimpi]

@SirCadian  I have just pushed out release 2023-10-07 as a typo in the last release was identified that might explain why the check was not resuming.    Please update to this one to see if your problem is still occurring

[SirCadian]

18 hours ago, itimpi said:

@SirCadian  I have just pushed out release 2023-10-07 as a typo in the last release was identified that might explain why the check was not resuming.    Please update to this one to see if your problem is still occurring

Will do.  I ran with test logging enabled last night to provide you with the diagnostics.  Parity Check Tuning updated overnight at 5.15am but not until after Appdata Backup concluded at 4.50am.  Parity Check did not resume but that's probably because PCT hadn't been updated at the point Appdata Backup finished.  I'll leave everything as is and see if the new update fixes things and if not, I'll dm you the diagnostics.  I'll post here either way.  Thanks.

[itimpi]

1 minute ago, SirCadian said:

Will do.  I ran with test logging enabled last night to provide you with the diagnostics.  Parity Check Tuning updated overnight at 5.15am but not until after Appdata Backup concluded at 4.50am.  Parity Check did not resume but that's probably because PCT hadn't been updated at the point Appdata Backup finished.  I'll leave everything as is and see if the new update fixes things and if not, I'll dm you the diagnostics.  I'll post here either way.  Thanks.

If you still have them then the diagnostics would probably tell me if you had encountered the problem that yesterdays release fixed.

[SirCadian]

3 minutes ago, itimpi said:

If you still have them then the diagnostics would probably tell me if you had encountered the problem that yesterdays release fixed.

Sent via DM.  Thanks for this.

[Zerginator]

Parity check tuning is not stopping anymore on disk overheat. Everything is activated, but the array rebuild just keeps going.

Led to 4 completely overheated disks with Smart failure, as disks are very slowly overheating after some hours and cooling cannot be improved unfortunately

[itimpi]

1 hour ago, Zerginator said:

Parity check tuning is not stopping anymore on disk overheat. Everything is activated, but the array rebuild just keeps going.

Led to 4 completely overheated disks with Smart failure, as disks are very slowly overheating after some hours and cooling cannot be improved unfortunately

 

Not sure why that should have broken as nothing has changed there for some time, but it is very difficult to test that option on my system as my drives never overheat so it is always possible that something slips through.   

 

If you can enable the Testing mode of logging in the plugin settings then get diagnostics at a point where you think the check should have been paused but it has not I should be able to see why it has not paused.

[Zerginator]

52 minutes ago, itimpi said:

 

If you can enable the Testing mode of logging in the plugin settings then get diagnostics at a point where you think the check should have been paused but it has not I should be able to see why it has not paused.

 

This is the line where the temperature value is triggered

[itimpi]

I can see that something has gone wrong in that the script has exited prematurely but not exactly why.    If you can go to Tools->PHP Settings and set the error reporting to All Categories then it will hopefully give me the exact line in the code that is causing the problem and why it is a problem.

[Zerginator]

9 hours ago, itimpi said:

I can see that something has gone wrong in that the script has exited prematurely but not exactly why.    If you can go to Tools->PHP Settings and set the error reporting to All Categories then it will hopefully give me the exact line in the code that is causing the problem and why it is a problem.

 

This is what the PHP Settings error log tells me:

[itimpi]

2 hours ago, Zerginator said:

 

This is what the PHP Settings error log tells me:

Thanks - that was exactly what I needed.    On that line it should be ‘$temp’ rather than simply ‘temp’.   At some point the $ must have been accidentally removed because at one point that code was definitely working.    It makes sense that it shows up for you as that is a bit of code that is only executed when drives are detected to actually overheat.    
 

I will make a quick fix; do a quick test; and push out an update later today.   When I have done so it would be very useful if you could confirm the plugin is now working as expected or if there is something else for me to track down.

[Zerginator]

1 minute ago, itimpi said:

I will make a quick fix; do a quick test; and push out an update later today.   When I have done so it would be very useful if you could confirm the plugin is now working as expected or if there is something else for me to track down.

 

Sure, I will update as soon as you have pushed it and give it a test run.