Help! Drive fail AFTER parity check

[PSteward]

Hello!

 

I finally completed the parity drive addition and everything was a success. I was curious what happens when I pressed the check parity button after this completion (to see how long it would take) and suddenly a drive got an x and started having errors (newer never had issues before).

I don't want to lose any data and am not sure how to proceed from here to get it back into the array (without out an X).

Edit: I read a ton and removed/re-added and now it's rebuilding but I don't understand what happened as the drive was fine up until I clicked on parity check (I saw write errors after it was disabled, unfortunately didn't save log before rebooting).

There were no physical changes, the drive has been fine for it's 1.5 years of existence, and it was perfect when added to the array, data moved to it, during parity creation. What could cause this by clicking "parity check"??

 

Please let me know!

 

Thanks,

 

--Phil

homeserver-diagnostics-20230512-1832.zip

Edited May 12, 2023 by PSteward

[PSteward]

Ok I don't know what the hell is going on, now my disk 1 is getting read errors during this rebuild....what is happening??

First disk 2 now disk 1 (both 16tb Seagate).

 

I removed disk 2 again and am copying all data off, and I am copying disk 1 off to another drive too temporarily. This is nuts all this time and copying and copying, why if I never had issues with these drives EVER does this OS have issues? I am copying ALL data from both these drives with ZERO errors but the moment parity check or build is involved they have read/write errors.

 

homeserver-diagnostics-20230512-1920.zip

Edited May 13, 2023 by PSteward

[PSteward]

Ok, more info.

During the rebuild it gave me read errors on disk1 (one of tons):

May 12 19:16:53 HOMESERVER kernel: md: disk1 read error, sector=221271952

SO I plugged this into badblocks (badblocks -v /dev/md1/ RANGE OF SECTORS) AND there are zero 0 bad blocks found and no read errors.

Data is good, Smart test is good, drive is good, badblock read is good only conclusion I have is that is a system bug with Unraid parity?

[JonathanM]

Random(ish) errors after adding a drive are typically caused by power issues, either poor connections or lack of capacity to service the extra load. Parity checks are a point of maximum power use, especially when first initiated and multiple drives are spun up.

 

A high resistance connection caused by a splitter can also be an issue, as it causes a voltage sag under load.

[PSteward]

Thanks for the response.

I will investigate power issues, but to be clear there is more usage during a parity check/rebuild than during a parity build? My parity build had all drives running full for 1.5 days without issue.

[JonathanM]

16 hours ago, PSteward said:

the drive was fine up until I clicked on parity check

Starting a parity operation spins up all drives. Spin up events are far and away the highest current draw, but only for a few moments.

[itimpi]

3 hours ago, PSteward said:

I will investigate power issues, but to be clear there is more usage during a parity check/rebuild than during a parity build?

From the same start conditions they should be identical.  As was pointed out it is the current surge when spinning up multiple drives simultaneously.    I would think that when you do the initial parity build that is likely that at least some drives are likely to already be spun up.

[PSteward]

I stopped what I was doing and did new config and then selected the option to re-use the existing parity before starting the array.

Now I am running the parity check again (error correcting) and here I am:

Total size:16 TB

Elapsed time:3 hours, 32 minutes

Current position:2.41 TB (15.1 %)

Estimated speed:179.5 MB/sec

Estimated finish:21 hours, 2 minutes

Sync errors corrected:3

 

As a note, I haven't physically changed anything (not in the same city as my server anyways!), BUT I did spin down ALL off my un-assigned drives. So I am leaning towards not getting enough power and already ordered a 1200w power supply for absolute overkill just in case.

Edited May 13, 2023 by PSteward

[PSteward]

So to conclude to this, I haven't had any more read errors.

I upgraded my power supply and assume all previous READ errors were related to running all drives at once since all other times everything worked fine but obviously didn't pull that amount of wattage.