JorgeB

Yes, it in the release notes:

No need to convert system separately, if you convert metadata (-mconvert) system will use the same profile. Some good info about using raid5 with btrfs from the mailing list: https://lore.kernel.org/linux-btrfs/[email protected]/

Changed Status to Closed Changed Priority to Other

You are using an SSD for cache, spin down is now only done for non rotational devices, it's on the release notes.

You can type "diagnostics" but this still works:

I was suggested one workaround, which is scrubbing each disk individually, but in a quick test total time looked that it would be about the same, still need to test better.

It did look like a power/connection problem. Once a disk gets disabled it needs to be rebuilt, if the emulated disk is mounting and contents looks correct you can rebuild on top. https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive

Yep, that's a known issue, but no fix for now.

Also, if safe mode doesn't help please post the diagnostics (downloaded when using safe mode).

It doesn't, btrfs saves all info on the device metadata, so you can even clone one or more devices with dd and the pool would start without complaining. Same or similar if the device is missing, though now with the latest beta and multiple pool support you can more easily add, remove or replace devices using the GUI. Yes, if the metadata is redundant, still files might be split on multiple drives (there's no way to prevent that), and the pool won't mount, you'll need to use recovery options. I'm already using v6.9 for my pools, good news i that you can import any pool create on UD.

Yes, the IPMI plugin also worked fine on v6.8.3, that was just an example, there might be other plugins affecting pool spin down, so first thing to do is to try in safe mode.

Disks are mounting: Filesystem Size Used Avail Use% Mounted on /dev/md12 1.9T 1.7T 221G 89% /mnt/disk12 /dev/md13 1.9T 1.3T 539G 72% /mnt/disk13 It should be, but with xfs you'd need previously created checksums to be sure, check is used space looks correct and if there's no lost+found folder all should be fine. https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive

Did you test in safe mode? There's a known issue with the IPMI plugin that causes that.

I missed that this started after a trim, in that case you should definitely connect the SSD to one of the Intel ports, there have been issues before with Samsung devices failing to trim on an Asmedia controller.

Most likely will need to use v6.9 latest beta for your NIC to work.

SSD is on an Asmedia controller, and those are usually fine, start by replacing cables since Samsung SSDs can be very picky with cable quality. You can also swap it to one of the Intel ports if issues persist.

It's under legacy HBAs on the download site, current firmware is 20.00.07.00

That suggests a damaged superblock, connect the disk to an onboard SATA port just to make sure it's not controller related and please post xfs_repair output.

Please don't crosspost:

Run xfs_repair again.

Please post new diags after array start.

Run it again without -n or nothing will be done.

Also note that both disks will still need to be rebuilt, but only if the emulated disks are mounting and showing the correct data.

To rebuild a failed disk.

That's not what the diags show, initially: Aug 10 01:10:27 Tower emhttpd: shcmd (85): mkdir -p /mnt/disk12 Aug 10 01:10:27 Tower emhttpd: shcmd (86): mount -t xfs -o noatime,nodiratime /dev/md12 /mnt/disk12 Aug 10 01:10:27 Tower kernel: XFS (md12): Metadata CRC error detected at xfs_sb_read_verify+0x114/0x15e [xfs], xfs_sb block 0xffffffffffffffff Aug 10 01:10:27 Tower kernel: XFS (md12): Unmount and run xfs_repair Aug 10 01:10:27 Tower kernel: XFS (md12): First 128 bytes of corrupted metadata buffer: Aug 10 01:10:27 Tower kernel: 00000000d3e59167: 58 46 53 42 00 00 10 00 00 00 00 00 1d 1c 11 0e XFSB............ Aug 10 01:10:27 Tower kernel: 00000000680eb2cf: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ Aug 10 01:10:27 Tower kernel: 000000007cb90224: 29 d1 16 8f 8d 1c 42 6b b6 a7 42 d1 f4 88 c4 75 ).....Bk..B....u Aug 10 01:10:27 Tower kernel: 00000000c3566992: 00 00 00 00 10 00 00 05 00 00 00 00 00 00 00 60 ...............` Aug 10 01:10:27 Tower kernel: 00000000323821d9: 00 00 00 00 00 00 00 61 00 00 00 00 00 00 00 62 .......a.......b Aug 10 01:10:27 Tower kernel: 00000000f1b45957: 00 00 00 01 07 47 04 44 00 00 00 04 00 00 00 00 .....G.D........ Aug 10 01:10:27 Tower kernel: 00000000d994568a: 00 03 a3 82 b4 b5 02 00 02 00 00 08 00 00 00 00 ................ Aug 10 01:10:27 Tower kernel: 0000000019056dcd: 00 00 00 00 00 00 00 00 0c 09 09 03 1b 00 00 05 ................ Aug 10 01:10:27 Tower kernel: XFS (md12): SB validate failed with error -74. Aug 10 01:10:27 Tower root: mount: /mnt/disk12: mount(2) system call failed: Structure needs cleaning. Aug 10 01:10:27 Tower emhttpd: shcmd (86): exit status: 32 Aug 10 01:10:27 Tower emhttpd: /mnt/disk12 mount error: No file system Aug 10 01:10:27 Tower emhttpd: shcmd (87): umount /mnt/disk12 Aug 10 01:10:27 Tower root: umount: /mnt/disk12: not mounted. Aug 10 01:10:27 Tower emhttpd: shcmd (87): exit status: 32 Aug 10 01:10:27 Tower emhttpd: shcmd (88): rmdir /mnt/disk12 Aug 10 01:10:27 Tower emhttpd: shcmd (89): mkdir -p /mnt/disk13 Aug 10 01:10:27 Tower emhttpd: shcmd (90): mount -t xfs -o noatime,nodiratime /dev/md13 /mnt/disk13 Aug 10 01:10:27 Tower root: mount: /mnt/disk13: mount(2) system call failed: Structure needs cleaning. Aug 10 01:10:27 Tower kernel: XFS (md13): Metadata CRC error detected at xfs_sb_read_verify+0x114/0x15e [xfs], xfs_sb block 0xffffffffffffffff Aug 10 01:10:27 Tower kernel: XFS (md13): Unmount and run xfs_repair Aug 10 01:10:27 Tower kernel: XFS (md13): First 128 bytes of corrupted metadata buffer: Aug 10 01:10:27 Tower kernel: 000000006f8c05c4: 58 46 53 42 00 00 10 00 00 00 00 00 1d 1c 11 0e XFSB............ Aug 10 01:10:27 Tower kernel: 00000000ce15d76b: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ Aug 10 01:10:27 Tower kernel: 00000000204e19e6: 84 8f 09 65 2b b2 41 fc 85 8c e1 ba fb 86 15 3f ...e+.A........? Aug 10 01:10:27 Tower kernel: 00000000ef363133: 00 00 00 00 10 00 00 05 00 00 00 00 00 00 00 60 ...............` Aug 10 01:10:27 Tower kernel: 000000008480b48b: 00 00 00 00 00 00 00 61 00 00 00 00 00 00 00 62 .......a.......b Aug 10 01:10:27 Tower kernel: 000000004876be3a: 00 00 00 01 07 47 04 44 00 00 00 04 00 00 00 00 .....G.D........ Aug 10 01:10:27 Tower kernel: 000000002d34a656: 00 03 a3 82 b4 b5 02 00 02 00 00 08 00 00 00 00 ................ Aug 10 01:10:27 Tower kernel: 00000000188e404a: 00 00 00 00 00 00 00 00 0c 09 09 03 1b 00 00 05 ................ Aug 10 01:10:27 Tower kernel: XFS (md13): SB validate failed with error -74. Aug 10 01:10:27 Tower emhttpd: shcmd (90): exit status: 32 Aug 10 01:10:27 Tower emhttpd: /mnt/disk13 mount error: No file system Aug 10 01:10:27 Tower emhttpd: shcmd (91): umount /mnt/disk13 Aug 10 01:10:27 Tower root: umount: /mnt/disk13: not mounted. Then: Aug 10 01:36:43 Tower emhttpd: shcmd (430): mkdir -p /mnt/disk12 Aug 10 01:36:43 Tower emhttpd: shcmd (431): mount -t xfs -o noatime,nodiratime /dev/md12 /mnt/disk12 Aug 10 01:36:43 Tower kernel: XFS (md12): Mounting V5 Filesystem Aug 10 01:36:44 Tower kernel: XFS (md12): Starting recovery (logdev: internal) Aug 10 01:36:44 Tower kernel: XFS (md12): Ending recovery (logdev: internal) Aug 10 01:36:44 Tower emhttpd: shcmd (432): xfs_growfs /mnt/disk12 Aug 10 01:36:44 Tower root: meta-data=/dev/md12 isize=512 agcount=4, agsize=122094660 blks Aug 10 01:36:44 Tower root: = sectsz=512 attr=2, projid32bit=1 Aug 10 01:36:44 Tower root: = crc=1 finobt=1, sparse=0, rmapbt=0 Aug 10 01:36:44 Tower root: = reflink=0 Aug 10 01:36:44 Tower root: data = bsize=4096 blocks=488378638, imaxpct=5 Aug 10 01:36:44 Tower root: = sunit=0 swidth=0 blks Aug 10 01:36:44 Tower root: naming =version 2 bsize=4096 ascii-ci=0, ftype=1 Aug 10 01:36:44 Tower root: log =internal log bsize=4096 blocks=238466, version=2 Aug 10 01:36:44 Tower root: = sectsz=512 sunit=0 blks, lazy-count=1 Aug 10 01:36:44 Tower root: realtime =none extsz=4096 blocks=0, rtextents=0 Aug 10 01:36:44 Tower emhttpd: shcmd (433): mkdir -p /mnt/disk13 Aug 10 01:36:44 Tower emhttpd: shcmd (434): mount -t xfs -o noatime,nodiratime /dev/md13 /mnt/disk13 Aug 10 01:36:44 Tower kernel: XFS (md13): Mounting V5 Filesystem Aug 10 01:36:44 Tower kernel: XFS (md13): Starting recovery (logdev: internal) Aug 10 01:36:44 Tower kernel: XFS (md13): Ending recovery (logdev: internal) Aug 10 01:36:44 Tower emhttpd: shcmd (435): xfs_growfs /mnt/disk13 Aug 10 01:36:44 Tower root: meta-data=/dev/md13 isize=512 agcount=4, agsize=122094660 blks Aug 10 01:36:44 Tower root: = sectsz=512 attr=2, projid32bit=1 Aug 10 01:36:44 Tower root: = crc=1 finobt=1, sparse=0, rmapbt=0 Aug 10 01:36:44 Tower root: = reflink=0 Aug 10 01:36:44 Tower root: data = bsize=4096 blocks=488378638, imaxpct=5 Aug 10 01:36:44 Tower root: = sunit=0 swidth=0 blks Aug 10 01:36:44 Tower root: naming =version 2 bsize=4096 ascii-ci=0, ftype=1 Aug 10 01:36:44 Tower root: log =internal log bsize=4096 blocks=238466, version=2 Aug 10 01:36:44 Tower root: = sectsz=512 sunit=0 blks, lazy-count=1 Aug 10 01:36:44 Tower root: realtime =none extsz=4096 blocks=0, rtextents=0 Please post new diags after array start.