December 28, 20178 yr Due to moving, I can't run my unRAID server for the near future. Most of my data is on two of the drives in my array though, and my workstation has two "hot swap tray" bays (not really hot swap, but means no side panels have to be messed with). Figured this was lucky, as I know all my drives in the pool are just xfs, and I know exactly which of my drives is which physically and in unraid. Threw them into the workstation, Ubuntu 16.04 LTS, and found that it won't mount the damn things. Running `syslog | tail` after trying shows the following, for both sdb, sdb1, sdc and sdc1 (the two drives in question, with one partition each). ┌─[ $ ]─[ k2trf@desktop ]───[ Wed Dec 27 | 23:43:58 ]───[ ~ ] └─> dmesg | tail [ 41.296933] userif-3: sent link up event. [ 43.913444] userif-3: sent link down event. [ 43.913451] userif-3: sent link up event. [ 79.096560] sdb: sdb1 [ 79.098099] sdc: sdc1 [ 372.615530] SGI XFS with ACLs, security attributes, realtime, no debug enabled [ 381.557248] sdb: sdb1 [ 381.559387] sdc: sdc1 [ 667.199182] XFS (sdb1): Invalid superblock magic number [ 681.848472] XFS (sdb): Invalid superblock magic number I am running (or I suppose was since it isn't currently up...) a 6.4 prerelease. I'd just updated as much as I could before the final shutdown, I want to say 6.4_rc17? That may not be right, but it's a close ballpark. I know for fact all the array drives were xfs (the system was initialized well back, sometime just prior to 6.2 IIRC), and all my drives have a relatively low pwron hours, with no SMART issues, nominal health, etc. before the final shutdown. They're both 1TB HGST 2.5" drives (R610 only slots 2.5" sadly), so I'm not entirely sure what to make of it. Installed xfsdump and xfsprogs, as I realized I didn't have them, and rebooted, but the tail shows the same. I've not tried reading the disks directly like this prior, but was aware that it was a fairly simple matter, especially with ubuntu on the workstation. The ACL/security flag has me a bit concerned, as I didn't recall setting up anything of the sort under unraid. Hoping I'm just missing something, possibly related to being in a 6.4 pre-release! If anyone has any insight into what might be tripping me up, I'm happily all ears!
December 28, 20178 yr Community Expert They should mount directly, I only see an error about sdb, maybe your parity drive?
December 28, 20178 yr Author In the case of the tail, I only attempted to mount sdb, since both drives produce the same output it seemed redundant. I'm fairly sure its not the parity -- I set the disks in unRAID to follow the numbering on the R610 face; 0, 1, 2, 3 were my four data drives (1-4 respectively), with 4 and 5 being parity one and two. Can always try the others when I have time later today, but I simultaneously don't think that likely, but also hear your logic, since even gparted shows them as unknown fs types. :\
December 28, 20178 yr Community Expert You cannot rely on the sdX Number staying constant across boots. Linux assigns them dynamically on each boot according to the order in which it sees them.
December 28, 20178 yr Author The sdX I'm referencing is when I pop them into my workstation -- I know which drives are which in unRAID before they are mounted, not paying attention to the sdX, but rather the drive assignment (Parity 1, Parity 2, Array 1, etc.), and that hasn't ever changed on me through reboots, shutdowns, etc. My workstation seems to just sort by UUID, alphabetically. Which makes my boot ssd sda, my two extracted drives from unRAID sdb and sdc, and my /home hdd sdd. Kinda funny in a way! EDIT: I'll be able to try throwing the other drives in when I have time later this evening. They're not very difficult to mount to the sleds, though it isn't toolless for 2.5" drives. Once I do that, regardless of the outcome, I'll come back, so we can hopefully start to get a bigger picture, because johnnie and you have made good points; I may be trying to look at two parity drives. Just because I'm pretty sure of what I'm doing doesn't mean I'm right! Edited December 28, 20178 yr by k2trf
December 29, 20178 yr Author You guys had it! Somehow, 0 and 1 must be my parity drives, because 2 and 3 immediately mounted with their xfs partitions, and I seem to have the bulk of my stuff! Definate *doh* moment! Thanks guys
Archived
This topic is now archived and is closed to further replies.