When I run a status display on a given snapraid setup, I see a line near the bottom that says:
The oldest block was scrubbed 112 days ago, the median 40, the newest 0.
I'm obvioulsy misreading the meaning of this statement, but what I'm looking for is an easy wayt to
run a scrub with the paramter -o 100 and reduce theoldest block scrubbed to 100 days old.
In this case, I tun this, and the 112 days ago value doesn't change, thus, this doesn't mean what I think it means -- but how do I validate that everything has been scrubbed wihtout doing a 100% scrub?
On a setup this big (14 2 TB volume + 2 parity drives) takes days...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
-p, --percentage PERC
Selects the part of the array to process in "scrub". PERC is a numeric value from 0 to 100, default is 12. When specifying 0, only the blocks marked as bad are scrubbed. This option can
be used only with "scrub".
So I think no matter if you put "-o 100" still only (or rather at most) 12% of the array will be processed, no matter how much older data you have, be it 20% or 50%.
I think you can easily check this theory by running multiple times the same -o 100 scrub. First of all you might notice after each run the "oldest block" maybe decreasing, up to 100 days at which point probably will stop and won't do anything on subsequent runs.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
When I run a status display on a given snapraid setup, I see a line near the bottom that says:
I'm obvioulsy misreading the meaning of this statement, but what I'm looking for is an easy wayt to
run a scrub with the paramter -o 100 and reduce theoldest block scrubbed to 100 days old.
In this case, I tun this, and the 112 days ago value doesn't change, thus, this doesn't mean what I think it means -- but how do I validate that everything has been scrubbed wihtout doing a 100% scrub?
On a setup this big (14 2 TB volume + 2 parity drives) takes days...
I think you are hit by the "-p" parameter:
-p, --percentage PERC
Selects the part of the array to process in "scrub". PERC is a numeric value from 0 to 100, default is 12. When specifying 0, only the blocks marked as bad are scrubbed. This option can
be used only with "scrub".
So I think no matter if you put "-o 100" still only (or rather at most) 12% of the array will be processed, no matter how much older data you have, be it 20% or 50%.
I think you can easily check this theory by running multiple times the same -o 100 scrub. First of all you might notice after each run the "oldest block" maybe decreasing, up to 100 days at which point probably will stop and won't do anything on subsequent runs.
Is there a way to formulate the command such that any data older than a specified number of days old will be scrubbed?
You should be able to use "-p 100". That will tell it to scrub 100% of the blocks that are older than your "-o 100" setting (100 days).
I personally just scrub once a week (12% of the drive) and its only takes 2 months for everything to be scrubed