Disk issues and ethernet card question

[MrBrac]

I have gigabit internet , I get poor download speeds , You think it could be my tp link pci ethernet card? the eth0 on mobo died.

 

Right now doing rebuild on disk 11 with bad sectors. But disk 10 has crc issues and now i Can tell its i/o error and now it needs to have its XFS fs cleaned again

 

 

bracservultra-diagnostics-20221014-0127.zip

[JorgeB]

Disk10 dropped offline, cancel the rebuild, replace cables on the disk10 and try again.

[MrBrac]

checked cables and then ran xfs_repair all good on disk 10 rebuild now.

[MrBrac]

so I swapped my sas card to pci slot 2 and now I got massive errors everywhere ?

 

 

bracservultra-diagnostics-20221015-0613.zip

[JorgeB]

Swap it back, it's not liking that slot, or it's not well seated, it stopped being detected, you are also having issues with disks connected to the onboard SATA, so there could be for example a power problem.

[MrBrac]

I am buying a new SAS , at dont trust this sas , just swapped it back.

 

disk 10 back to being a jack. ITs using same cables the 7 other disks are using , all other disk issues non sas are not issue with my swap back.

Edited October 15, 2022 by MrBrac

[MrBrac]

I tried everything I can not rebuild disk without it freezing , tried do it in maintenance.

bracservultra-diagnostics-20221016-1354.zip

[JorgeB]

Swap cables/slot between disk10 and a different disk, if issues continue it could be a disk problem.

[MrBrac]

I got the new sas attempting to redo parity , I formated 10 and 11. Right now 10 is btrfs and rest are xfs is that okay?

 

I went to switch disk 10 back to XFS , need format it . But then the parity is going insanely fast roughly 6gbs a second and its almost done.

bracservultra-diagnostics-20221019-1446.zip

 

Oct 19 12:59:30 bracservultra kernel: Buffer I/O error on dev md10, logical block 1953506608, async page read
Oct 19 12:59:30 bracservultra  emhttpd: shcmd (197): mkfs.xfs -m crc=1,finobt=1 -f /dev/md10
Oct 19 12:59:30 bracservultra root: mkfs.xfs: pwrite failed: Input/output error
Oct 19 12:59:30 bracservultra root: libxfs_bwrite: write failed on (unknown) bno 0x3a3812948/0x100, err=5
Oct 19 12:59:30 bracservultra root: mkfs.xfs: Releasing dirty buffer to free list!
Oct 19 12:59:30 bracservultra root: found dirty buffer (bulk) on free list!

 

 

forget everything said and lets just look at current daigs , cuz stuff is running good. Somehow parity was done in 45 mins .

 

 

bracservultra-diagnostics-20221019-1446.zip

Edited October 20, 2022 by MrBrac

[JorgeB]

Still read errors on disk10, rebuild will skip ahead on errors, hence the speed.

[MrBrac]

23 hours ago, JorgeB said:

Still read errors on disk10, rebuild will skip ahead on errors, hence the speed.

I never did a rebuild , I formated it and said screw the rebuild I moved data off emulated disk and formatted it. I also formated 11. I then did parity check and somehow I got 5.6 GB speed instead of my usual 143 mb speed.

 

... This was a parity check. I ran extended smart test on 9 10 11 , so far 9 and 10 came back 100% good.

 

 

Ok Idk anymore

 

Took out DIsk 1 and 10. 10 had a good smart health.

 

I think 1 is old and dropping out.

 

Now it says my parity is X

 

 

Says disk 9 has errors but I ran xfs repair and nothing

 

I cant sleep with this crap , so I redid parity in maintenance its now doing a parity sync. 

 

Before it was trying to do a check but I removed 2 disks so that would cause confusion right?

 

so the parity is totally invalid and has be redone. + the disk is brand new I can not believe I have this many disks with issues.

 

I keep running xfs repair on  so called issue disks and it has no problems.

bracservultra-diagnostics-20221021-0258.zip

Edited October 21, 2022 by MrBrac

[JorgeB]

9 hours ago, MrBrac said:

I never did a rebuild ,

I meant the parity sync, it cannot be completed successfully with all the errors on disk10, any errors are skipped.

[MrBrac]

I mean I dont really know at this point , I checked disk 10 its good empty disk. at the moment I am just trying get back my parity and save disk 9. It seems crazy that I got this much disk issues.

[JorgeB]

On 10/17/2022 at 10:17 AM, JorgeB said:

Swap cables/slot between disk10 and a different disk, if issues continue it could be a disk problem.

Have you done this?

[MrBrac]

3 hours ago, JorgeB said:

Have you done this?

gonna swap disk 1 and 10 after parity is fixed.

[JorgeB]

Parity cannot be fixed while disk10 keeps giving read errors.

 

5 hours ago, MrBrac said:

at the moment I am just trying get back my parity and save disk 9

What do you meant by this? Parity can cannot be synced to save a failing disk, also dsk9 looks healthy, errors might have the same root cause as disk10.

[MrBrac]

I know disk 9 and 10 are healthy well IO errors happen on 9 when parity starts. Disk 1 is really old but has passed smart stats. I think i will pull that disk. I replaced the sas cables before I can reswap the old cables. There standard blue cables from amazon.

Edited October 21, 2022 by MrBrac

[JorgeB]

Problems are likely due to bad SATA cables, bad power cables/splitter or bad PSU.

[MrBrac]

I bought some nice power splitters off amazon the Y types and I got cables off amazon. My PSU is EVGA brand I trust them.

 

What about running a memory test?

[JorgeB]

RAM won't cause disk read errors.

[MrBrac]

Disk 1 as Dev 1 smart test keeps saying I interrupted it which is a lie. I think the disk is old and dead.

 

BTW what do you think of my Ethernet options should I upgrade?

 

swapped cables for all disks on sas and now disk 10 is now disk 1 lets see

Edited October 22, 2022 by MrBrac

[MrBrac]

no more errors on disk 1(10) with older cable swapped.

[MrBrac]

Why is unraid disabling my Disk 1 (10) when its 100% healthy with no read sector errors?

 

disk 11 now failing extended smart test , Man am I cursed?

 

 

bracservultra-diagnostics-20221026-2335.zip

Edited October 27, 2022 by MrBrac

[JorgeB]

13 hours ago, MrBrac said:

Disk 1 (10)

Is this disk1 or disk10? No disk10 is assigned and there are no disk related errors on the syslog.

 

Disk11 is failing and needs to be replaced.

[itimpi]

16 hours ago, MrBrac said:

Why is unraid disabling my Disk 1 (10) when its 100% healthy with no read sector errors?

If Unraid disables a disk it thinks a write to it has failed.  
 

Disk1 has CRC errors, so maybe one of them could not be recovered and caused a write error.   CRC errors are more often than not related to cabling (SATA or power) rather than the disk itself.