tdunks Posted April 11, 2021 Share Posted April 11, 2021 (edited) A few weeks ago I started getting read errors on three new-to-me drives that I recently picked up. These drives are all older 8tb Seagate Archive drives which I picked up off of someone in an unused state - they only have <2 months of power-on hours. I am only getting these read errors near the end of parity checks and only on these drives. The drives were pre-cleared without issue before having them be put in. I am currently using a supermicro SC846 case with a SAS2-EL1 backplane and an LSI 9205-8i HBA with both cables plugged into the backplane. I thought the added drives might be stressing the HBA so I added a 40mm fan to it with no change. I have also tried reseating the HBA and the cables, as well as changing the slots the drives are in. The odd thing is that all 3 usually have identical levels of read errors, sometimes only a few and sometimes tens or hundreds of thousands. The drives all pass extended SMART tests and don't seem to have any other issues. I have no idea what the issue could be here - I really doubt all 3 drives are failing in the same way at the same time but I don't see another issue? Could it be the HBA? if so, why only these drives? tower-diagnostics-20210411-1628.zip Edited April 11, 2021 by tdunks Diagnostics Quote Link to comment
Squid Posted April 11, 2021 Share Posted April 11, 2021 Diagnostics might tell the story Quote Link to comment
akawoz Posted April 11, 2021 Share Posted April 11, 2021 (edited) I had exactly this happening with my SAS drives under 6.9.1 when I started using the SAS spindown plugin, see my post here: Same case, same backplane, but using SM motherboard with integrated HBA. Edited April 11, 2021 by akawoz more info on my config Quote Link to comment
tdunks Posted April 11, 2021 Author Share Posted April 11, 2021 54 minutes ago, Squid said: Diagnostics might tell the story I added to OP Quote Link to comment
JorgeB Posted April 12, 2021 Share Posted April 12, 2021 Log is full of task abort errors, and not just for those disks, any change if you disable disk spin down? Quote Link to comment
tdunks Posted April 13, 2021 Author Share Posted April 13, 2021 On 4/11/2021 at 4:29 PM, akawoz said: I had exactly this happening with my SAS drives under 6.9.1 when I started using the SAS spindown plugin, see my post here: Same case, same backplane, but using SM motherboard with integrated HBA. I just did this fix and tried a parity check again with the same results. I'm not sure what else to try at this point. Quote Link to comment
JorgeB Posted April 13, 2021 Share Posted April 13, 2021 Dis you disable disk spin down? Quote Link to comment
tdunks Posted April 13, 2021 Author Share Posted April 13, 2021 30 minutes ago, JorgeB said: Dis you disable disk spin down? Will test next, but a basic feature line this as part of unraid shouldn't cause these issues, that's a lot of wasted power keeping them spin up all the time Quote Link to comment
JorgeB Posted April 13, 2021 Share Posted April 13, 2021 7 minutes ago, tdunks said: but a basic feature line this as part of unraid shouldn't cause these issues It shouldn't, and maybe it isn't, but it doesn't hurt to test. Quote Link to comment
tdunks Posted April 15, 2021 Author Share Posted April 15, 2021 On 4/13/2021 at 2:09 PM, JorgeB said: It shouldn't, and maybe it isn't, but it doesn't hurt to test. It seems to have worked - I have attached both Diagnostics from after the EPC fix with errors and after disabling spindown to see if a cause can be identified. tower-diagnostics-20210415-1343.zip tower-diagnostics-20210413-1711.zip Quote Link to comment
JorgeB Posted April 15, 2021 Share Posted April 15, 2021 Logs looks clean now, for some reason your hardware is not liking spin down, hough it's not the first time an LSI HBA has some issues with spin down. Quote Link to comment
tdunks Posted April 15, 2021 Author Share Posted April 15, 2021 Is it likely to be fixed by switching to a newer SAS3 HBA? Quote Link to comment
PeteAron Posted April 16, 2021 Share Posted April 16, 2021 arent LSI _the_ recommended cards? Forgive my ignorance but it is very surprising to read that a common, recommended controller doesnt play nicely with unraid. What's the deal with this? (I have two LSI cards and _extremely_ happy that I havent yet moved on from 6.8.3, jesus) Quote Link to comment
JorgeB Posted April 16, 2021 Share Posted April 16, 2021 5 hours ago, PeteAron said: arent LSI _the_ recommended cards? Yes, and they are usually the most reliable controller you can use with Unriad, other than the onboard Intel SATA ports, but with some specific issues/hardware there can be issues. Quote Link to comment
Lameth Posted April 25, 2021 Share Posted April 25, 2021 I had this error when I upgraded to 6.9.2 from 6.9.1. It seems that the spin up wasn't working properly and I got lots of read errors. I've performed a fallback, and it solved my problem. Quote Link to comment
tdunks Posted April 25, 2021 Author Share Posted April 25, 2021 I changed the controller to a 9311 and no longer have errors with spindown on. It seems so be an issue only with the older controllers. Quote Link to comment
Recommended Posts
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.