June 24, 20233 yr why do i keep getting random errors on my drives? theres no rhyme or reason. I just replaced a HDD that was giving i/o errors. then replaced a cache drive also giving errors that looked like it was a failing drive. and now one of my parity drives have just dropped offline with io errors while rebuilding data on that first data drive. am i chasing ghosts here? this all happening within 3 days, am i having some sort of hardware issue that i cant pinpoint? with one of my 2 parity drives down if i lose one more drive during this data rebuild i am screwed. At this rate Ill have another failure before this rebuild finishes. Is this maybe RAM going bad? I have ECC Example of error message on the latest issue on one of my parity drives: Jun 23 23:16:52 Snuts kernel: I/O error, dev sdi, sector 2269569328 op 0x1:(WRITE) flags 0x0 phys_seg 128 prio class 0 Jun 23 23:16:52 Snuts kernel: sd 1:0:7:0: [sdi] tag#2497 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=0s Jun 23 23:16:52 Snuts kernel: sd 1:0:7:0: [sdi] tag#2497 Sense Key : 0x2 [current] Jun 23 23:16:52 Snuts kernel: sd 1:0:7:0: [sdi] tag#2497 ASC=0x4 ASCQ=0x0 Jun 23 23:16:52 Snuts kernel: sd 1:0:7:0: [sdi] tag#2497 CDB: opcode=0x8a 8a 00 00 00 00 00 87 46 e5 30 00 00 04 00 00 00 Jun 23 23:16:52 Snuts kernel: I/O error, dev sdi, sector 2269570352 op 0x1:(WRITE) flags 0x0 phys_seg 128 prio class 0 ran a smart test on the drive and it came back no errors. Diagnostics attached. snuts-diagnostics-20230623-2348.zip Edited June 24, 20233 yr by DontWorryScro formatting
June 24, 20233 yr Community Expert It's not logged as a disk problem, and SMART looks fine, so most likely power/connection issue.
June 24, 20233 yr Author 7 minutes ago, JorgeB said: It's not logged as a disk problem, and SMART looks fine, so most likely power/connection issue. so am i just looking at buying a higher wattage PSU? heres the other thing. the three drives that have popped offline this week are all connected differently. I was getting btrfs errors on the cache drive which was a m.2 nvme on a pcie adapter plugged directly into the motherboard on the main NAS. i swapped that out for an enterprise SSD today. the initial drive that was giving similar errors to this latest parity drive dropoff was in a different case with its own PSU in an HP expander connected DAS. This parity drive that fell off tonight during a data rebuild of the initial HDD is connected via an LSI HBA on the main NAS. seems like an odd confluence of events in the span of 3 days. all separate. all related to power and cables? i mean of course its possible. just seems like something else is going on and its killin me.
June 24, 20233 yr Community Expert It's difficult to say what the problem could be, only that doesn't look disk related.
July 3, 20233 yr Author On 6/24/2023 at 4:33 AM, JorgeB said: It's difficult to say what the problem could be, only that doesn't look disk related. OK I managed to re-enable the parity drive referenced in my original post that had dropped off (sdi) and after 3 days of rebuilding and finishing without issue it was fine for a couple days. Today I checked on the server to find that same drive had dropped off with i/o errors. Seems like this has been happening during scheduled mover operations overnight. Can you take a look at the latest diagnostics I've attached and tell me if it is exhibiting the same behavior as before? Do I need to just completely re-cable the sata power and mini SAS cables to my Unraid NAS to rule that out? The PSU is an EVGA SuperNOVA 1000 G+ 80 Plus Gold so I merely assume the power itself is probably adequate. snuts-diagnostics-20230703-1034.zip
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.