January 8Jan 8 Hello,I updated my UnRAID server from 7.2.0 to 7.2.3. The update completed successfully with no issue but when I restarted I was unable to start the Array and was given an error message stating "Wrong Pool State Cache - missing devices"I foolishly referenced ChatGPT which assured me that this was a known issue and simply needed the cache device manually mounted to a temp_cache directory then to use the New Config tool to rebuild the cache pool.Upon running the New Config all the devices in my Array were unmounted along with a variety of UDMA CRC errors flagging. I do not believe the UDMA CRC errors are hardware related as the system had been running stably for months prior. With all my drives now unmounted and no clear way to address the cache pool issue I was hoping to get some help putting everything back together. I have a recent diagnostics log below.Thanks,Kurttowerarchives-diagnostics-20260108-1303.zip
January 9Jan 9 2 hours ago, HatSolo said:Upon running the New Config all the devices in my Array were unmountedProbably you mean Unassigned, not unmounted. No drives are mounted until the array is started. 'mount' is when the OS loads the filesystem to allow access to the contents.2 hours ago, HatSolo said:a variety of UDMA CRC errors flagging. I do not believe the UDMA CRC errors are hardware relatedUDMA CRC errors are logged in the drive firmware when it detects an error in the data received (checksum). These are not drive problems, they are problems communicating with the drive, connections, etc. Often these won't show up as an error since the data is resent. And some problems that do show up as errors will not log UDMA CRC error since the drive never received any data to checksum.These will not reset, but you can acknowledge these on the Dashboard page by clicking on the SMART warning (👎) and it will warn again if it increases.Assign all drives as they were, check the Parity Valid box and start the array.
January 9Jan 9 Also, since you reset the array, it will have considered all the previously counted UDMA CRC errors as new and so that is why you would get notifications about that.
January 9Jan 9 1 hour ago, trurl said:Assign all drives as they were, check the Parity Valid box and start the array.Do you know your drive assignments?
January 9Jan 9 Author I have the drive assignments written down from when I last added a drive to the array but I believe it's from over a year ago. Is it possible to pull the drive assignments from the diagnostics file from prior to them being unassigned? I'm fairly confident my notes are still accurate but validating would be nice if possible.Thanks for all the help
January 9Jan 9 Yes, if the assignments were correct at boot:disk0 is parity, disk29 is parity2Jan 8 11:48:00 TowerArchives kernel: md: import disk0: (sde) ST16000NT001-3LV101_ZRS15WZ7 size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk29: (sdh) ST16000NT001-3LV101_ZRS15KP8 size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk1: (sdj) ST16000NT001-3LV101_ZRS0R455 size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk2: (sdf) ST14000NM001G-2KJ103_ZTM0HSQ5 size: 13672382412 Jan 8 11:48:00 TowerArchives kernel: md: import disk3: (sdc) ST14000NM001G-2KJ103_ZTM0HL71 size: 13672382412 Jan 8 11:48:00 TowerArchives kernel: md: import disk4: (sdg) ST14000NM001G-2KJ103_ZTM0HZN0 size: 13672382412 Jan 8 11:48:00 TowerArchives kernel: md: import disk5: (sdd) ST12000VN0008-2YS101_ZRT1A0FF size: 11718885324 Jan 8 11:48:00 TowerArchives kernel: md: import disk6: (sdi) ST16000NT001-3LV101_ZRS1NZ48 size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk7: (sdl) ST16000NT001-3LV101_ZR5G4L1Q size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk8: (sdk) ST16000NT001-3LV101_ZR5G5BR5 size: 15625879500 Jan 8 11:48:00 TowerArchives kernel: md: import disk9: (sdn) ST16000NT001-3LV101_ZRS1NXQA size: 15625879500 Jan 8 11:48:00 TowerArchives emhttpd: import 30 pool device: no deviceJan 8 11:48:00 TowerArchives emhttpd: import 31 pool device: (sdm) Samsung_SSD_870_EVO_1TB_S75BNL0Y906869EJan 8 11:48:00 TowerArchives emhttpd: import 32 pool device: (sdb) Samsung_SSD_870_EVO_500GB_S6PXNU0X311391M
January 9Jan 9 Author Excellent, the assignments you listed match what I have from my notes so high confidence that they're correct. Unfortunately, when I tried to start the array (with drives assigned and Parity Valid box clicked) I still get the "Wrong Pool State Cache - missing devices" error. Unsure if this is related but my Pool Drives does list an unused Cache device with nothing assigned to it. I'm unsure what it is and can remove it if we think that may address the issue.Here is the output from btrfs fi showroot@TowerArchives:~# btrfs fi showLabel: none uuid: 1a45a4a6-246b-4f7c-83cf-9388f7c27b6c Total devices 1 FS bytes used 125.70GiB devid 1 size 465.76GiB used 159.02GiB path /dev/sdb1Label: none uuid: 4ec52a98-d674-4da6-997e-0a4b9d027f30 Total devices 1 FS bytes used 144.00KiB devid 1 size 931.51GiB used 2.02GiB path /dev/sdm1
January 10Jan 10 Author Here's the updated diagnostics, thanks for the help! towerarchives-diagnostics-20260109-1600.zip
January 10Jan 10 Solution If nothing is supposed to be assigned as cache, change its slots to zero and see if you can start.
January 10Jan 10 Author Haha well that ended up being a simple fix. Array started without issue. I've spot checked a few things and seem to have access to everything with no issues but let me know if you'd recommend checking anything else. Again I really appreciate the help you all are lifesavers!Below is the btrfs if show results if that is still helpful.root@TowerArchives:~# btrfs fi showLabel: none uuid: 4ec52a98-d674-4da6-997e-0a4b9d027f30 Total devices 1 FS bytes used 144.00KiB devid 1 size 931.51GiB used 2.02GiB path /dev/sdm1Label: none uuid: 1a45a4a6-246b-4f7c-83cf-9388f7c27b6c Total devices 1 FS bytes used 128.59GiB devid 1 size 465.76GiB used 160.02GiB path /dev/sdb1Label: none uuid: f7effdea-9855-4791-831c-8501b27a9d69 Total devices 1 FS bytes used 9.34GiB devid 1 size 20.00GiB used 17.02GiB path /dev/loop2
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.