I noticed that scrub will only run if there are change in array, and will only scrub after sync completed.

According documents from snapraid, the scrub job should run every week at least even there are nothing change in array to proect data from damaged(particularly silent error), and only scrub in synced array.

That's means we should run scrub jobs even there are nothing change in array if it's a synced array.

I am also going to implemet this in scripts and submit to previous pull request.

Scrub absolutely needs to be run regularly even if there has been no change, but if there has been changes but it's below the threshold set for a sync, then a scrub won't have the information for the current data and it could end up trying to fix something that was changed. I think that the script should be changed to run a sync after a predetermined time regardless if there are no, or so few changes that it didn't trigger a sync earlier. That way you can make sure the scrub is run regularly sync data that reflects the current state of the disk.

Hi,
This is exactly what I implemented on the new version scripts. And I have changed to using new scripts without any problem for a few weeks.

But according to the author's' suggestion, we may find another volunteer who is also using SnapRaid to test it and ensure everything works as expected. So, @thenebular could you have a test?

The usage is simple, just disable cron of old version, and upload the new version to anywhere you want, just remember to give it execute permission. Then add scripts to cron according to your settings. And we may go on to that PR.

Scripts action that I implemented currently just for your reference.

1. Try to sync our array and will exit with an email if hit our threshold.
2. Regardless of whether the array is changed or not, once we have synced array and meet our scrub interval, scrub array as we configured.
3. Send an email with full logs the scripts ran.

Scrub absolutely needs to be run regularly even if there has been no change, but if there has been changes but it's below the threshold set for a sync, then a scrub won't have the information for the current data and it could end up trying to fix something that was changed. I think that the script should be changed to run a sync after a predetermined time regardless if there are no, or so few changes that it didn't trigger a sync earlier. That way you can make sure the scrub is run regularly sync data that reflects the current state of the disk.

Closing due to age.