September 3, 20241 yr Hello everyone. I have restarted my server and found that one of my ZFS pools is marked as degraded. I have looked at the pool detail, and although Scrub is running (no apparent errors), one of the disks has a message stating “was /dev/sdm1”. Because of all this, the pool is marked as DEGRADED. How can I solve this problem? This is the first time I have had such an error using ZFS. Thank you very much.
September 3, 20241 yr Community Expert Not sure how that happened but you have a device from one pool being used to replace a device from another pool: NAME STATE READ WRITE CKSUM rocinante ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 replacing-0 ONLINE 0 0 0 sdc1 ONLINE 0 0 0 sdl1 ONLINE 0 0 0 (resilvering) sde1 ONLINE 0 0 0 sdh1 ONLINE 0 0 0 sdd1 ONLINE 0 0 0 sdf1 ONLINE 0 0 0 sdg1 ONLINE 0 0 0 cache sdc FAULTED 0 0 0 corrupted data Wait for this resilver to finish then post new diags, there's also a problem with the the cache vdev.
September 3, 20241 yr Author OMG! Well, it was all due to a server reboot... I'll wait for the Scrub process to finish and I'll post again the diagnosis.
September 3, 20241 yr Author Now that I look in detail that cache drive that FAULTED, it is an SSD disk that on reboot has been disconnected. Maybe the whole problem comes from there. When I finish resilvering, I will check the data and power connection to see why it was not detected during the boot process. I will report the progress.
September 4, 20241 yr Author Good morning again. The resilvering process is finished. Inside the GUI of unRAID I keep on marking the pool as DEGRADED. I attach the diagnostic files again. What should I do now? Thank you very much. midiam.local-diagnostics-20240904-0829.zip
September 4, 20241 yr Community Expert This is an very confused state, cache vdev from the rociante pool was assigned twice to the same pool, also as a min device, not sure how you did that, I would recommend manually removing the cache vdev from that pool, then re-importing both pools with the actual devices they have assigned at the moment, then replace the first device from the rocinante pool to correct that one and free up the device that it's using from the other pool, finally use that device to correct the tycho pool. Post back if you need help to do the above.
September 4, 20241 yr Author It is very strange what happened to me. But it was all due to a server restart and, maybe, the cache disk of the rocinante pool disappeared. I understand that the steps that you indicate to me would have to be done from the CLI and not from the GUI. I am not very familiar with ZFS management from the CLI (I added the cache disk to test and see if it was worth using an old SSD to speed up disk access following a guide in the forum). How would you remove the cache disk from the rocinante pool first? "zpool detach rocinante sdc"? Would I have to do the rest of the steps you indicate from the CLI or could I do it from the unRAID interface using the ZFS Master plugin? Thank you very much for the support you are giving me, JorgeB.
September 4, 20241 yr Community Expert To remove the cache vdev you need to use the CLI, the rest should be possibly to do with the GUI, with the array running type: zpool remove rocinante /dev/sdc Then post the complete zpool status output or new diags.
September 4, 20241 yr Author After deleting the missing cache disk with the command you indicated, I have reassigned again the disks to each of the pools and now it tells me that the pool is unmountable and the option it gives me is to format it all... T_T I don't know if it can be mounted by hand from the CLI somehow? I send the current logs (I have configured unRAID so that it does not launch docker and can interfere with something) Any suggestions? Thanks! PS: I think I'm going to cry.... midiam.local-diagnostics-20240904-1637.zip
September 4, 20241 yr Community Expert 33 minutes ago, Coque said: I have reassigned again the disks to each of the pools You have to assign the currently devices the pool is using, not the ones you want in the end. Since you dind't post the zpool status output as asked, now post the output from zpool import and maybe it's not too late.
September 4, 20241 yr Author Sorry for the inconvenience. Here it is the zpool import: pool: tycho id: 10901411189181231896 state: DEGRADED status: One or more devices contains corrupted data. action: The pool can be imported despite missing or damaged devices. The fault tolerance of the pool may be compromised if imported. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J config: tycho DEGRADED raidz1-0 DEGRADED 13502292040317525666 FAULTED corrupted data sdk1 ONLINE sdm1 ONLINE sdn1 ONLINE sdo1 ONLINE sdp1 ONLINE pool: donnager id: 16552885184057344388 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: donnager ONLINE raidz1-0 ONLINE sdi1 ONLINE sdj1 ONLINE sdr1 ONLINE sdq1 ONLINE sds1 ONLINE sdt1 ONLINE pool: tycho id: 5924509355974799305 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: tycho ONLINE raidz1-0 ONLINE sdl1 ONLINE sde1 ONLINE sdh1 ONLINE sdd1 ONLINE sdf1 ONLINE sdg1 ONLINE pool: disk1 id: 8733512297879271467 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk1 ONLINE sda1 ONLINE pool: cache id: 3766673087186651326 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: cache ONLINE nvme0n1p1 ONLINE Thanks again.
September 4, 20241 yr Community Expert Type: echo 1 > /sys/block/sdl/device/delete This will make this disk go offline, to see if it's interfering with the other pool being detected, it won't damage anything, and if you reboot later it will come back online, after typing the command wait 5 secs and post the output from zpool import again
September 4, 20241 yr Author This is the result: pool: tycho id: 10901411189181231896 state: DEGRADED status: One or more devices contains corrupted data. action: The pool can be imported despite missing or damaged devices. The fault tolerance of the pool may be compromised if imported. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J config: tycho DEGRADED raidz1-0 DEGRADED 13502292040317525666 FAULTED corrupted data sdk1 ONLINE sdm1 ONLINE sdn1 ONLINE sdo1 ONLINE sdp1 ONLINE pool: donnager id: 16552885184057344388 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: donnager ONLINE raidz1-0 ONLINE sdi1 ONLINE sdj1 ONLINE sdr1 ONLINE sdq1 ONLINE sds1 ONLINE sdt1 ONLINE pool: tycho id: 5924509355974799305 state: DEGRADED status: One or more devices contains corrupted data. action: The pool can be imported despite missing or damaged devices. The fault tolerance of the pool may be compromised if imported. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J config: tycho DEGRADED raidz1-0 DEGRADED sdl1 UNAVAIL sde1 ONLINE sdh1 ONLINE sdd1 ONLINE sdf1 ONLINE sdg1 ONLINE pool: disk1 id: 8733512297879271467 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk1 ONLINE sda1 ONLINE pool: cache id: 3766673087186651326 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: cache ONLINE nvme0n1p1 ONLINE sdl1 sais that unavailable, but the disk in tycho still sais 13502292040317525666 FAULTED corrupted data
September 4, 20241 yr Community Expert Problem is that it's showing two pools named tyco and it's missing the rocinante pool, so big mess, use the same command above to offline the remaining tyco pool devices, replacing sdl with the devices below sdk sdm sdn sdo sdp Then output from zpool import again
September 4, 20241 yr Community Expert P.S. about to go out in a few minutes, so my take a couple o hours to reply.
September 4, 20241 yr Author Here is the new zpool import after deleting the devices you mentioned: pool: donnager id: 16552885184057344388 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: donnager ONLINE raidz1-0 ONLINE sdi1 ONLINE sdj1 ONLINE sdr1 ONLINE sdq1 ONLINE sds1 ONLINE sdt1 ONLINE pool: tycho id: 5924509355974799305 state: DEGRADED status: One or more devices contains corrupted data. action: The pool can be imported despite missing or damaged devices. The fault tolerance of the pool may be compromised if imported. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J config: tycho DEGRADED raidz1-0 DEGRADED sdl1 UNAVAIL sde1 ONLINE sdh1 ONLINE sdd1 ONLINE sdf1 ONLINE sdg1 ONLINE pool: disk1 id: 8733512297879271467 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk1 ONLINE sda1 ONLINE pool: cache id: 3766673087186651326 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: cache ONLINE nvme0n1p1 ONLINE
September 4, 20241 yr Community Expert For now I think both pools can be recovered, but it still appears with the wrong name, lets try a different way, reboot to bring all the disks back, don't start the array, so if auto-start is enabled, disable that first, then post again zpool import to confirm all disks for the second pool are present and we will try to import it based on the ID.
September 4, 20241 yr Author This is the new zpool import without starting the pool on reboot pool: tycho id: 10901411189181231896 state: DEGRADED status: One or more devices contains corrupted data. action: The pool can be imported despite missing or damaged devices. The fault tolerance of the pool may be compromised if imported. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J config: tycho DEGRADED raidz1-0 DEGRADED 13502292040317525666 FAULTED corrupted data sdl1 ONLINE sdm1 ONLINE sdn1 ONLINE sdo1 ONLINE sdp1 ONLINE pool: donnager id: 16552885184057344388 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: donnager ONLINE raidz1-0 ONLINE sdi1 ONLINE sdj1 ONLINE sdr1 ONLINE sdq1 ONLINE sds1 ONLINE sdt1 ONLINE pool: disk1 id: 8733512297879271467 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk1 ONLINE sda1 ONLINE pool: tycho id: 5924509355974799305 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: tycho ONLINE raidz1-0 ONLINE sdk1 ONLINE sde1 ONLINE sdh1 ONLINE sdd1 ONLINE sdf1 ONLINE sdg1 ONLINE pool: cache id: 3766673087186651326 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: cache ONLINE nvme0n1p1 ONLINE
September 4, 20241 yr Community Expert Type zpool import 5924509355974799305 And then output of zpool status
September 4, 20241 yr Author The name is wrong, because the content inside belongs to rocinante, but seems well mounted. pool: tycho state: ONLINE scan: resilvered 12.9T in 1 days 19:11:37 with 0 errors on Wed Sep 4 04:23:27 2024 config: NAME STATE READ WRITE CKSUM tycho ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 sdk1 ONLINE 0 0 0 sde1 ONLINE 0 0 0 sdh1 ONLINE 0 0 0 sdd1 ONLINE 0 0 0 sdf1 ONLINE 0 0 0 sdg1 ONLINE 0 0 0 errors: No known data errors Following step? Edited September 4, 20241 yr by Coque
September 4, 20241 yr Author I have obfuscated the name of the folders because some of them have sensitive names.
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.