November 17, 20232 yr Community Expert Hi All I am in the process of building a new production server... at the moment it is just a blank shell - so this is not important and if its going to be hard to fix, I will just wipe everything and start again, although a solution may help others where this is critical to them Anyway, this is whats happened - I have built the server and assigned a 6Tb parity disk (xfs), 5 x 6Tb array disks(xfs) and a 500Gb SSD(btrfs) cache disk. Set everything up and moved /appdata, /domains and /system on the cache disk. All 'was' working fine. I then changed the RAID card and re-assigned all the disks in the correct slots and used tools->new config to clear the array and reassign them. I ticked the "parity is valid" box as no data has changed. Anyway, now when I start the array it all comes online fine - but the cache disk shows as File System "auto" and shows as unmountable: unsupported or no file system! obviously my /appdata /domains and /system shares are missing as they should be on the cache disk. Is there a fix for this, or am I better just wiping the USB key and starting from scratch again? Thanks in advance Edited November 18, 20232 yr by SliMat
November 17, 20232 yr Author Community Expert Apologies for missing the diags off @JorgeB Edited November 17, 20232 yr by SliMat
November 17, 20232 yr Community Expert Looks like there's no valid filesystem on cache, possibly changing the controller affected that, possibly will need to re-format, post the output of: fdisk -l /dev/sdh and blkid
November 17, 20232 yr Author Community Expert 1 hour ago, JorgeB said: fdisk -l /dev/sdh Disk /dev/sdh: 465.76 GiB, 500107862016 bytes, 976773168 sectors Disk model: Seagate BarraCud Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes 1 hour ago, JorgeB said: blkid /dev/sda1: LABEL_FATBOOT="UNRAID" LABEL="UNRAID" UUID="272C-EBE2" BLOCK_SIZE="512" TYPE="vfat" /dev/loop1: TYPE="squashfs" /dev/sdf1: UUID="9051d512-d9ae-4296-9a37-03b80b732e25" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="7eacda8e-e23a-4e72-b54a-cb4680c4de78" /dev/sdd1: UUID="36f9802c-1dd0-4fb9-b20f-5af44818ea0f" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="a354f918-4690-4575-a405-c3ed9d54d91e" /dev/md2p1: UUID="9051d512-d9ae-4296-9a37-03b80b732e25" BLOCK_SIZE="512" TYPE="xfs" /dev/sdb1: UUID="4edd947d-64d4-49a6-b0b8-dafe25718840" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="782589f1-ff67-4caa-af79-d0ea0e09695a" /dev/md5p1: UUID="a8556926-4df7-42b6-9ea2-0e160924f708" BLOCK_SIZE="512" TYPE="xfs" /dev/md1p1: UUID="158d392e-3dce-49d7-9aa3-ee4ccac37aeb" BLOCK_SIZE="512" TYPE="xfs" /dev/sdg1: UUID="55ad914b-d093-4fe8-9c81-63e8a5fdc189" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="b8d05b0f-fd3c-4cb1-90cd-b9b06adef593" /dev/md4p1: UUID="4edd947d-64d4-49a6-b0b8-dafe25718840" BLOCK_SIZE="512" TYPE="xfs" /dev/loop0: TYPE="squashfs" /dev/sde1: UUID="158d392e-3dce-49d7-9aa3-ee4ccac37aeb" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="195693c8-ad13-4462-8f02-d16fbbaa4219" /dev/sdc1: UUID="a8556926-4df7-42b6-9ea2-0e160924f708" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="8e2e2148-ca38-46f6-b2cd-ac2dfb4bd205" /dev/md3p1: UUID="55ad914b-d093-4fe8-9c81-63e8a5fdc189" BLOCK_SIZE="512" TYPE="xfs"
November 17, 20232 yr Community Expert 2 hours ago, SliMat said: parity disk (xfs) Just thought I would comment on this to clear up misconceptions. Parity has no filesystem so it is not xfs or any other. It is all just bits and no files.
November 17, 20232 yr Author Community Expert 1 hour ago, JorgeB said: possibly will need to re-format If I do end up having to format the cache disk, is there a way to recreate the default system as the /appdata, /domains & /system shares are now missing? There are no VMs or dockers installed yet - so a "reset" to 'as new' will not be a problem.
November 17, 20232 yr Author Community Expert 5 minutes ago, trurl said: Just thought I would comment on this to clear up misconceptions. Parity has no filesystem so it is not xfs or any other. It is all just bits and no files. Absolutely correct... I missed this - thanks @trurl
November 17, 20232 yr Community Expert Disable VM Manager (and keep Docker also disabled) until you have cache for system share to live on. appdata share will be created when you enable Docker.
November 17, 20232 yr Author Community Expert 6 minutes ago, trurl said: Disable VM Manager (and keep Docker also disabled) until you have cache for system share to live on. appdata share will be created when you enable Docker. Done
November 18, 20232 yr Community Expert There's no partition o cache, you can try a recovery util like testdisk but probably faster to just re-format and recreate.
November 18, 20232 yr Author Community Expert 48 minutes ago, JorgeB said: There's no partition o cache, you can try a recovery util like testdisk but probably faster to just re-format and recreate. So, Docker disabled itself when the cache disk disappeared and I have disabled VM Manager, as suggested by @trurl. How do I recreate the missing information, /appdata, /system. /domains, etc, so I can start with an 'as new' system again? Thanks
November 18, 20232 yr Community Expert Solution 44 minutes ago, SliMat said: How do I recreate the missing information, /appdata, /system. /domains, etc, so I can start with an 'as new' system again These are automatically created if docker and VMs are enabled.
November 18, 20232 yr Author Community Expert 21 minutes ago, itimpi said: These are automatically created if docker and VMs are enabled. Thanks guys... formatted and re-enabled and now missing shares are re-created. Learning point - be careful swapping RAID cards... luckily this was a test server
November 18, 20232 yr Community Expert 1 hour ago, SliMat said: be careful swapping RAID cards Yep, one of the reasons we don't recommend using RAID controllers with Unraid, swapping regular SATA controllers or SAS/SATA HBAs is not a problem.
November 18, 20232 yr Author Community Expert 6 hours ago, JorgeB said: Yep, one of the reasons we don't recommend using RAID controllers with Unraid, swapping regular SATA controllers or SAS/SATA HBAs is not a problem. Agreed, although I'm using an HPE Proliant DL380 and have updated the chassis from G8 to G9 (before putting the machine in service), hence the swap in RAID card. In case anyone else with a Proliant reads this, the best set up I have found is to set the RAID card to HBA mode first Interesting you say swapping HBA controllers is not an issue as the G8 chassis had a P420i in HBA mode and the G9 chassis uses the p840ar in HBA mode too - so in this instance it was just swapping an HBA controller and I still lost the cache disk! Edited November 18, 20232 yr by SliMat
November 19, 20232 yr Community Expert 15 hours ago, SliMat said: Interesting you say swapping HBA controllers is not an issue as the G8 chassis had a P420i in HBA mode I meant true HBAs, like an LSI in IT mode, controllers that have an HBA mode may not always work correctly, because they may not be totally transparent, like a true HBA.
November 19, 20232 yr Author Community Expert 1 minute ago, JorgeB said: I meant true HBAs, like an LSI in IT mode, controllers that have an HBA mode may not always work correctly, because they may not be totally transparent, like a true HBA. Ah, fair enough - at least its working again now and luckily it was not in production
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.