Getting a call trace alert in Fix Common Problems. Additionally, my log is showing the following entry:

BTRFS warning (device loop2): csum failed root 5 ino 25348 off 5787648 csum 0xfc87e656 expected csum 0xacee4560 mirror 1

Running btrfs dev stats /mnt/cache shows 0s across the board even when checked every 5 min over 15 min.

 

Any help is greatly appreciated. Diagnostics are attached.

starkillerbase-diagnostics-20171005-2050.zip

Lots of ATA errors on your OCZ SSD, these are usually the result of a bad cable/connection, after fixing that run a correcting scrub on the cache pool, make sure there were no uncorrectable errors and delete and recreate your docker image.

 

 

12 hours ago, johnnie.black said:

Lots of ATA errors on your OCZ SSD, these are usually the result of a bad cable/connection, after fixing that run a correcting scrub on the cache pool, make sure there were no uncorrectable errors and delete and recreate your docker image.

 

 

Hey Johnnie, How can one do a scrub? 

And can you do it at any point just to double check your pools?

1 minute ago, SavellM said:

How can one do a scrub? 

 

On the main page click on the first cache device.

 

1 minute ago, SavellM said:

And can you do it at any point just to double check your pools?

 

Yes, but pool performance will take a hit for the duration.

1 minute ago, johnnie.black said:

 

On the main page click on the first cache device.

 

 

Yes, but pool performance will take a hit for the duration.

Thanks man,

 

Is it worth doing this periodically?

And shall I tick the option Repair corrupted blocks?

11 minutes ago, SavellM said:

Is it worth doing this periodically?

 

If all is working well it's usually not needed but not a bad idea to do it monthly just to make sure.

 

11 minutes ago, SavellM said:

And shall I tick the option Repair corrupted blocks?

 

Might as well to avoid a second run if errors are found.

 

Also, don't forget to keep an eye on your syslog to confirm the ATA errors are gone, they look like this:

 

Sep 30 22:07:08 StarkillerBase kernel: ata5.00: status: { DRDY }
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: failed command: WRITE FPDMA QUEUED
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: cmd 61/00:c0:40:31:07/07:00:0c:00:00/40 tag 24 ncq dma 917504 out
Sep 30 22:07:08 StarkillerBase kernel:         res 40/00:70:c0:94:f6/00:00:1a:00:00/40 Emask 0x50 (ATA bus error)
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: status: { DRDY }
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: failed command: WRITE FPDMA QUEUED
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: cmd 61/60:c8:40:38:07/04:00:0c:00:00/40 tag 25 ncq dma 573440 out
Sep 30 22:07:08 StarkillerBase kernel:         res 40/00:70:c0:94:f6/00:00:1a:00:00/40 Emask 0x50 (ATA bus error)
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: status: { DRDY }
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: failed command: WRITE FPDMA QUEUED
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: cmd 61/00:d0:e8:78:48/02:00:0f:00:00/40 tag 26 ncq dma 262144 out
Sep 30 22:07:08 StarkillerBase kernel:         res 40/00:70:c0:94:f6/00:00:1a:00:00/40 Emask 0x50 (ATA bus error)
Sep 30 22:07:08 StarkillerBase kernel: ata5.00: status: { DRDY }
Sep 30 22:07:08 StarkillerBase kernel: ata5: hard resetting link
Sep 30 22:07:09 StarkillerBase kernel: ata5: SATA link down (SStatus 0 SControl 300)
Sep 30 22:07:11 StarkillerBase kernel: ata5: hard resetting link
Sep 30 22:07:11 StarkillerBase kernel: ata5: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Sep 30 22:07:14 StarkillerBase kernel: ata5.00: configured for UDMA/133
Sep 30 22:07:14 StarkillerBase kernel: ata5: EH complete

 

Edited October 6, 20178 yr by johnnie.black

On 10/6/2017 at 12:24 AM, johnnie.black said:

Lots of ATA errors on your OCZ SSD, these are usually the result of a bad cable/connection, after fixing that run a correcting scrub on the cache pool, make sure there were no uncorrectable errors and delete and recreate your docker image.

 

 

 

If there were no errors is the docker image step needed? Is it a good idea just in case?

Yes

Okay, I've done that. Thank you.

 

Could I ask for the reasoning behind that step? It is the only one that doesn't seem obvious to me.

Syslog if filled with errors like these:

 

Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1
Oct  5 04:41:16 StarkillerBase kernel: BTRFS warning (device loop2): csum failed root 5 ino 40214 off 462848 csum 0x2659a9b5 expected csum 0xc35f94a8 mirror 1

 

 

loop2 is your docker image, it was corrupt.

Ahh, that makes more sense. Thank you!