Let's say I am replacing my parity drive with a larger one, and party is re syncing.
All of a sudden the power goes out, but I have my server on a UPS with 5 minutes of run time left

If I pause the party sync before the server shuts down, then, when power is restored and the server boots backup, can I resume the parity sync from where I paused it, or would I have to restart the party sync all over again?

 

My guess, is I would have to start all over again.

With basic Unraid you would have to start again.    Normally all array level operations need to be restarted from the beginning if they do not run to completion before stopping the server.

 

However if you have the Parity Check Tuning plugin installed and you are on Unraid 6.9.0 (or later) and the following conditions are met:

• you have set the option to resume array operations next time array starts in the plug-in settings [EDIT:   I have been thinking of making this the default behaviour so that the user does not need to explicitly request it - any feedback on this idea is welcomed]
• you pause the sync if you want although this should not be necessary.
• you successfully stop the array (so that you will not end up with an unclean shutdown).

then you can shutdown the server and when you next boot the server (presumably after the power is OK) the parity operation should be resumed from the point it had already reached when the array is started.

 

Note that I used the word ‘should’ because although I have tested what I think is this scenario on my development environment I have not yet had any feedback as to whether it has successfully (or not) been handled in a real world situation.  If it does not work for any reason then it can be restarted from the beginning so it does not hurt to try.   I would love to get some feedback from anyone who has used this feature.

 

EDIT:  If anyone finds themselves about to use this feature in anger I would be very grateful if they could also set the ‘testing’ mode of logging in the plugins settings and then send me their diagnostics/syslog after restarting the array so I can check if the restart was handled as expected.