I'm running Unraid 6.3.5 on a Supermicro 2U Server X8DTN+ 2x Xeon X5675 3.06ghz Hexcore with 48 Gigs of RAM. I have a Supermicro 933T which I am using as an expansion chassis which is connected from the backplane of the 933T to a SATA extender card. I am currently running two Marvel SATA cards in the Main Supermicro, but have two LSI cards on the way. I get daily random system crashes. The system will be unusable and when i view the console from IPMI I can see the login screen but it is also unresponsive. I get multiple sector errors in two 8TB drives which are installed in the expansion Chassis. I have removed these drives and done both SMART test and low-level block checks, with no errors. I have also changed the SATA cables. Any help would be appreciated. Again I'm hoping the LSI cards will rectify any issues. Thanks. 

sam-the-eagle-diagnostics-20171213-0930.zip

FCPsyslog_tail.txt

There are checksums errors on your cache device:

 

Dec 13 09:14:28 Sam-The-Eagle kernel: BTRFS warning (device sdj1): csum failed ino 22600422 off 0 csum 2566472073 expected csum 3881293769
Dec 13 09:14:28 Sam-The-Eagle kernel: BTRFS warning (device sdj1): csum failed ino 22600422 off 0 csum 2566472073 expected csum 3881293769

Possibly a corrupt docker image, running a scrub will tell which files are affected.

I did a scrub, fixed the affected files. The system ran great for about 8 hours. Until it crashed again. 

sam-the-eagle-diagnostics-20171213-2138.zip

sam-the-eagle-diagnostics-20171213-2047.zip

 

FCPsyslog_tail.txt

Edited December 14, 20178 yr by xcsuperfly

These files are corrupt, since you have s single device cache scrub can't fix them, you'll need to replace or delete those files:

 

 

Dec 13 11:53:37 Sam-The-Eagle emhttp: cmd: /usr/local/emhttp/plugins/dynamix/scripts/tail_log syslog
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS warning (device sdj1): checksum error at logical 58057940992 on dev /dev/sdj1, sector 65176304, root 5, inode 22600422, offset 0, length 4096, links 1 (path: appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Cache/autotag_wordlist.json)
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS error (device sdj1): bdev /dev/sdj1 errs: wr 0, rd 0, flush 0, corrupt 1, gen 0
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS warning (device sdj1): checksum error at logical 58057945088 on dev /dev/sdj1, sector 65176312, root 5, inode 22600422, offset 4096, length 4096, links 1 (path: appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Cache/autotag_wordlist.json)
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS error (device sdj1): bdev /dev/sdj1 errs: wr 0, rd 0, flush 0, corrupt 2, gen 0
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS warning (device sdj1): checksum error at logical 58057949184 on dev /dev/sdj1, sector 65176320, root 5, inode 22600422, offset 8192, length 950, links 1 (path: appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Cache/autotag_wordlist.json)
Dec 13 12:02:50 Sam-The-Eagle kernel: BTRFS error (device sdj1): bdev /dev/sdj1 errs: wr 0, rd 0, flush 0, corrupt 3, gen 0
Dec 13 12:10:07 Sam-The-Eagle kernel: BTRFS warning (device sdj1): checksum error at logical 80486334464 on dev /dev/sdj1, sector 115273216, root 5, inode 22436957, offset 552960, length 4096, links 1 (path: appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Logs/Plex Media Server.2.log)
Dec 13 12:10:07 Sam-The-Eagle kernel: BTRFS error (device sdj1): bdev /dev/sdj1 errs: wr 0, rd 0, flush 0, corrupt 4, gen 0
Dec 13 12:10:07 Sam-The-Eagle kernel: BTRFS error (device sdj1): bdev /dev/sdj1 errs: wr 0, rd 0, flush 0, corrupt 5, gen 0

 

Yesterday appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Cache/autotag_wordlist.json was corrupt and I deleted it. Could there be a reason these files continue to become corrupt?

Sent from my Pixel XL using Tapatalk

Not really, disk seems fine, you should be using ECC, so maybe try using a different disk for cache if you have one.

The server crashed again. Hasn't for the past few days. all seemed well. Attached the logs.

sam-the-eagle-diagnostics-20171223-2049.zip

FCPsyslog_tail.txt

You likely have a very bad SATA cable/connection on disk12:

 

199 UDMA_CRC_Error_Count    0x000a   198   198   000    Old_age   Always       -       13417933

Most likely the reason for these:

 

Dec 19 14:11:17 Sam-The-Eagle kernel: sd 8:0:1:0: [sdk] tag#1 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Dec 19 14:11:17 Sam-The-Eagle kernel: sd 8:0:1:0: [sdk] tag#1 Sense Key : 0xb [current]
Dec 19 14:11:17 Sam-The-Eagle kernel: sd 8:0:1:0: [sdk] tag#1 ASC=0x47 ASCQ=0x3
Dec 19 14:11:17 Sam-The-Eagle kernel: sd 8:0:1:0: [sdk] tag#1 CDB: opcode=0x88 88 00 00 00 00 00 05 c7 7f c0 00 00 04 00 00 00
Dec 19 14:11:17 Sam-The-Eagle kernel: blk_update_request: I/O error, dev sdk, sector 96960448
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960384
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960392
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960400
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960408
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960416
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960424
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960432
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960440
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960448
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960456
Dec 19 14:11:17 Sam-The-Eagle kernel: md: disk12 read error, sector=96960464