Had an unclean shutdown yesterday and noticed 4 drives throwing errors afterwards. The four drives are all connected to my MB onboard LSI3108 controller (which is in JBOD mode). There are four other drives on this controller that is working fine. This did not happen until this latest shutdown but only presents itself during a parity check so may have been an issue for longer. The server otherwise seems to run without issues.

 

Dec 3 03:25:06 Mordor kernel: sd 1:2:3:0: [sde] tag#2097 BRCM Debug mfi stat 0x2d, data len requested/completed 0x80000/0x0

 

and eventually into these and then write errors

Dec 3 03:34:18 Mordor kernel: sd 1:2:2:0: [sdd] tag#2876 Abort request is for SMID: 2877

Dec 3 03:34:18 Mordor kernel: sd 1:2:2:0: attempting task abort! scmd(0x00000000e7b2f6ea) tm_dev_handle 0x2

Dec 3 03:34:18 Mordor kernel: megaraid_sas 0000:01:00.0: megasas_disable_intr_fusion is called outbound_intr_mask:0x40000009

Dec 3 03:34:18 Mordor kernel: megaraid_sas 0000:01:00.0: megasas_enable_intr_fusion is called outbound_intr_mask:0x40000000

Dec 3 03:34:18 Mordor kernel: sd 1:2:2:0: [sdd] tag#2876 task abort FAILED!! scmd(0x00000000e7b2f6ea)

 

The fact that it is only these 4 makes me think it may be a cable problem... but I have used the same setup for many months now through many different parity checks without issue. Wondering if anyone has input into other causes for these errors before I buy new cabling as they are fairly spendy. Thanks!

mordor-diagnostics-20231203-1306.zip

Edited December 3, 20232 yr by bs.king

5 hours ago, bs.king said:

The fact that it is only these 4 makes me think it may be a cable problem

Could be, try swapping the cables around and see if the problem follows the cable.