Jump to content
Sign in to follow this  
ars92

New parity drive add syncing really slow

13 posts in this topic Last Reply

Recommended Posts

I know there are many topics on this but most who have complained were using kind of old hardware. 

 

I'm running a threadripper 1920x at 4ghz which should support avx2 so Q calculation shouldn't be too slow but it's currently running at speeds like 1mbps at times and sometimes even lower. 

 

It's a new 10tb red which is an addition to another 10tb red which was added around April last year but don't think that ran this slow at the time though. 

 

Attached diagnostics. Apologies if I don't reply fast enough as it's currently close to midnight where I am and am about to doze off. 

 

It looks like it's currently building parity of the third disk (i can only see the read count of disk 3-6 increasing now and no more on disk 1 and 2) so could that mean disk 3 is having issues? 

tower-diagnostics-20191203-2325.zip

Share this post


Link to post
6 minutes ago, ars92 said:

It looks like it's currently building parity of the third disk (i can only see the read count of disk 3-6 increasing now and no more on disk 1 and 2) so could that mean disk 3 is having issues? 

That's because it's past the 3TB mark, nothing on the log, you should run the diskspeed docker to see if al disks are performing normally.

Share this post


Link to post
15 minutes ago, johnnie.black said:

That's because it's past the 3TB mark, nothing on the log, you should run the diskspeed docker to see if al disks are performing normally.

Thanks Johnnie, am running it on each disk one by one now. 

Share this post


Link to post

Yeah it does seem like disk 3 has some issues at a certain sector of the disk. 

 

I guess for now I just have to wait it out and once its done with those  sectors the speed should pick back up. 

 

Will try replacing the SATA cable but I don't think that'll help. It really looks like the disk had issues at a certain sector of a certain platter. 

SmartSelect_20191204-003546_Chrome.jpg

Share this post


Link to post

Yes, disk3 is likely the problem, for WD drives these slow sector zones can usually be seen on SMART on this attribute:

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   193   193   051    -    46077

 

This should be 0 on a healthy WD, and in my experience these slow sectors turn to bad sectors sooner or later.

Share this post


Link to post
6 hours ago, johnnie.black said:

Yes, disk3 is likely the problem, for WD drives these slow sector zones can usually be seen on SMART on this attribute:


ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   193   193   051    -    46077

 

This should be 0 on a healthy WD, and in my experience these slow sectors turn to bad sectors sooner or later.

Welp the raw value for that attribute is now at 68k and increasing. 

 

And on unraid the read error count is at 9048 for that disk. 

 

I don't think it's worth it to continue the Q calculation as it would most probably be not accurate for disk 3 in case a rebuild is needed anyway. 

 

Would it be better if I cancel the Q calculation, reuse disk 6 (brand new 8tb with only 10gb data in it which I can move over to another disk) and do a new config with disk 6 as the new disk 3,do a sync using the P parity to rebuild disk 3, and then redo the Q calculation. 

 

What do you think? And I think when I do that new config I won't even add parity 2 to the new config at first just to ensure it foesn try to do anything as its technically not readyto rebuild disk 3 array drive. 

Share this post


Link to post
7 hours ago, ars92 said:

And on unraid the read error count is at 9048 for that disk. 

That means there are already bad sectors.

 

Please post current diags to see best option(s) going forward.

Share this post


Link to post
1 hour ago, johnnie.black said:

That means there are already bad sectors.

 

Please post current diags to see best option(s) going forward.

OK here it is. Do note I've stopped the Q calculation this morning after my last reply. 

tower-diagnostics-20191204-1557.zip

Share this post


Link to post

OK, IMHO these are the best 2 options:

 

1) let parity2 sync finish, it will take more time than normal but as long as there are no read errors on another disk at the same time as the ones on disk3 parity1 can correct the read errors on disk3, and parity2 will be valid once synced, then replace disk3 when possible.

 

2) use parity2 disk to replace disk3 now, to do that unassign parity2, start array, stop array, assign it as disk3, start array to rebuild.

 

P.S. disk2 also shows some read errors, but they were some time ago and it should be fine for now, but it's a good idea to monitor the raw_read_error_rate attribute, if it keeps increasing it will likely fail soon.

Share this post


Link to post
1 hour ago, johnnie.black said:

OK, IMHO these are the best 2 options:

 

1) let parity2 sync finish, it will take more time than normal but as long as there are no read errors on another disk at the same time as the ones on disk3 parity1 can correct the read errors on disk3, and parity2 will be valid once synced, then replace disk3 when possible.

 

2) use parity2 disk to replace disk3 now, to do that unassign parity2, start array, stop array, assign it as disk3, start array to rebuild.

 

P.S. disk2 also shows some read errors, but they were some time ago and it should be fine for now, but it's a good idea to monitor the raw_read_error_rate attribute, if it keeps increasing it will likely fail soon.

Thanks Johnnie I think for now ill go with option 1 as my plan is to get another 8tb EFAX Red in January as the prices are not too bad(~260usd In my area). 

 

I just cant bring myself to use a 10tb for disk drive now as I've always wanted the parity disks to be a pair of 10tb and disk drive to slowly move to 8tb as time passes, just a personal preference to make the disk sizes look nice :D

 

Have resumed parity for now. At least the dip seems to be on a pretty narrow sector. So the slow sync may be for a few GBs rather than hundreds of GBs. Will update here once it's done. 

Share this post


Link to post

So the parity sync ended couple of hours ago and there’s like 220000 read errors on unraid for disk 3

 

So just to confirm, I should run a parity check, but UNCHECK “write corrections to parity” right? Or else those errors on the disk would pretty much screw up the parity?

Share this post


Link to post

If you want to run a parity check run with no correct, but it will take a wile because of disk3, and it can even get disabled, I would probably wait until disk3 is replaced.

 

You could also unassign disk3 and work with the emulated disk for now, with dual parity it still protects you if another disk fails before disk3 gets replaced.

Share this post


Link to post
50 minutes ago, johnnie.black said:

If you want to run a parity check run with no correct, but it will take a wile because of disk3, and it can even get disabled, I would probably wait until disk3 is replaced.

 

You could also unassign disk3 and work with the emulated disk for now, with dual parity it still protects you if another disk fails before disk3 gets replaced.

Hmm Guess I’ll stay put for now and exclude disk3 from all shares. 

 

Or maybe even slowly move data out of disk 3 and get rid of it while waiting for a new disk. 

 

Thanks so much for your help man :)

Share this post


Link to post

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this