wobblewoo Posted December 31, 2023 Share Posted December 31, 2023 Hello all. I installed a new USB PCI card and when i rebooted the array wouldn't start reporting one of my cache disks was missing. The cache was a RAID1 array with two 480Gb SanDisk SSD Plus drives. At first i thought i'd knocked something so checked Power/Sata etc, thats all fine. I then swapped around the leads and the error repeats to the same drive, so i know its not a lead. The faulty drive shows in Unassigned Devices as an odd drive (Sandisk Milpitas) @ 16.4 KB that won't format. I can get the array to start by stopping the array and setting pool devices to 1 but my VM (which was on the Cache drive) is missing. Can i get help with the following please? How do i check the faulty SSD? Fix the cache array if i buy a new identical SSD? Change the Cache to a single drive? Diagnostics attached. Many thanks - i'm a bit lost! tower-diagnostics-20231231-1239.zip Quote Link to comment
JorgeB Posted December 31, 2023 Share Posted December 31, 2023 If the pool was redundant you can still start the array with the remaining device by checking "I want to do this" next to array start button. Quote Link to comment
wobblewoo Posted December 31, 2023 Author Share Posted December 31, 2023 Thank you. I've got the array started with only one SSD but on the main screen it says under Size: Unmountable: Unsupported or no file system My Docker containers and VMs are all gone. I thought being RAID1 this was mirrored so was recoverable. Quote Link to comment
JorgeB Posted January 1 Share Posted January 1 18 hours ago, wobblewoo said: I thought being RAID1 this was mirrored so was recoverable. It would be unless there are other issues, post the diagnostics. Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 Hello again, thanks for the reply. Happy new year. I did on the first post but ive just done another and its attached. The only thing i may have done whilst checking leads is to attach the working SSD drive to a SATA connector different to it's original - but theres only two options so i can easily put it back. I hope that makes sense. A replacement SSD is arriving tomorrow. Wx tower-diagnostics-20240101-1145.zip Quote Link to comment
JorgeB Posted January 1 Share Posted January 1 Dec 31 15:09:48 Tower emhttpd: /mnt/cache: no btrfs or device /dev/sdg1 is not single Looks like the pool a single slot, that won't work, stop the array, unassign the pool device, start array, stop array, change pool slots to 2, assign the pool device to the 1st slot, start array, post new diags. 1 Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 Stopped array & unassigned pool device (so it now says 'no device') but i cannot start the array without checking 'Yes, I want to do this' - so i did. Stopped array once started, changed pool to 2 slots and restarted. Yay, dockers and VMs are back - i can't thank you enough. I'm hoping the process of rebuilding the RAID1 is easy once the new drive arrives. New diags attached. What i now need to do is learn out how to backup the VMs & Dockers (apart from in RAID1) tower-diagnostics-20240101-1320.zip Quote Link to comment
itimpi Posted January 1 Share Posted January 1 5 minutes ago, wobblewoo said: What i now need to do is learn out how to backup the VMs & Dockers (apart from in RAID1) Users typically use the Appdata Backup and VM Backup plugins for this purpose. 1 Quote Link to comment
JorgeB Posted January 1 Share Posted January 1 Jan 1 13:14:20 Tower kernel: BTRFS info (device sdg1): 451 enospc errors during balance Jan 1 13:14:20 Tower kernel: BTRFS info (device sdg1): balance: ended with status: -28 Pool fail to balance to single due to not enough space, move/delete some data than restart the array and post new diags. Quote Link to comment
trurl Posted January 1 Share Posted January 1 5 minutes ago, wobblewoo said: backup the VMs & Dockers docker.img needs no backup since its contents are easily downloaded again by reinstalling your dockers. https://docs.unraid.net/unraid-os/manual/docker-management/#re-create-the-docker-image-file https://docs.unraid.net/unraid-os/manual/docker-management/#re-installing-docker-applications Plugins: 1 Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 @itimpi Thankyou VM Backup says its a BETA so I've just installed Appdata Backup. I added an un exported share 'Appdatabackup' on the main array (turned Security to Private) and have set this as a backup location for now. Can't decide if this is good enough or i should add an Unassigned Device external drive and use that. Then configured the settings to backup to /mnt/user/Appdatabackup/ It says its backing up: /mnt/user/appdata /mnt/cache/appdata Should i add? /mnt/user/domains/ /mnt/user/isos/ Manual backup currently running. Thankyou. Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 23 minutes ago, JorgeB said: Jan 1 13:14:20 Tower kernel: BTRFS info (device sdg1): 451 enospc errors during balance Jan 1 13:14:20 Tower kernel: BTRFS info (device sdg1): balance: ended with status: -28 Pool fail to balance to single due to not enough space, move/delete some data than restart the array and post new diags. @JorgeB sdg1 is a 'SanDisk_SSD_PLUS_480GB (sdg) 480Gb' with 107Gb free, not sure why that error is coming up. Would this be as ive re-plugged the physical devices in the wrong way round after testing leads? The faulty SSD which is currently 'Sandisk_Milpitas_SSD (sdh)' & should be the same 480Gb SSD as above, is only showing as a 16.4KB drive in unassigned devices - if thats relevant. Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 @trurl I may revisit this as currently my cache, a RAID1 of 2x 480Gb SSDs is a bit large for a simple disk writing cache and i currently also use it for VMs & Dockers. I've got 2x 120Gb SSDs (different brands) & a 240Gb SSD spare so may redo the disk cache as 2x 120Gb RAID1 for redundant data copying use until mover moves the data to the main array. That will free up a 2x 480Gb & a 240Gb for VMs/dockers/backup purposes Quote Link to comment
JorgeB Posted January 1 Share Posted January 1 10 minutes ago, wobblewoo said: sdg1 is a 'SanDisk_SSD_PLUS_480GB (sdg) 480Gb' with 107Gb free, But it's fully allocated, you try running a balance, but may still need to free up some space first. Quote Link to comment
wobblewoo Posted January 1 Author Share Posted January 1 1 hour ago, JorgeB said: But it's fully allocated, you try running a balance, but may still need to free up some space first. balance complete, new diags attached. tower-diagnostics-20240101-1553.zip Quote Link to comment
JorgeB Posted January 2 Share Posted January 2 Balanced OK, now just needs to remove the missing device, re-start the array and post new diags. Quote Link to comment
wobblewoo Posted January 2 Author Share Posted January 2 Wanted to point out that i've now physically removed the faulty drive, i dont have the new one to replace it yet. Stopped array, Pool devices says 'Slots 2', the first (Cache) is my 480Gb and the second (Cache 2) says 'unassigned'. If i change 'Slots' to 1 which is how i assume you "remove the missing device" I then can't start the array as it says 'Wrong Pool State. cache - too many missing/wrong devices' Am i doing something wrong, this is what i see? Quote Link to comment
JorgeB Posted January 2 Share Posted January 2 The other device was not removed yet, for now leave the two slots. 3 hours ago, JorgeB said: re-start the array and post new diags. Quote Link to comment
wobblewoo Posted January 2 Author Share Posted January 2 (edited) Done. Array started and Dockers/VMs are present. tower-diagnostics-20240102-1428.zip Edited January 2 by wobblewoo Added attachment Quote Link to comment
Solution JorgeB Posted January 2 Solution Share Posted January 2 Missing device was now removed, you can now change slots to 1 or add a new device to create a new mirror. Quote Link to comment
wobblewoo Posted January 2 Author Share Posted January 2 Stop array Change Pool Devices | Slots: to '1' That seems to be fixed, thank you for bearing with me. 1 Quote Link to comment
Recommended Posts
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.