steelrat Posted October 14, 2020 Share Posted October 14, 2020 (edited) After my 500GB (Samsung Evo 860 btrfs) Cache drive started to go twards 90% full, I got a new 1TB Samsung Evo 860 and added it as a second cache drive. Now it seems my Cache is mounted as read-only and I can't start any VMs or Dockers which I had on the Cache. I was advised to attach the logs. Can anyone help? I'm kinda lost here Clarification: - The 1TB Disk was added to the cache (raid1 as I've learned now). - I was happy to see, that the cache seemed to work immideately and resized my vm disk (which is on the cache) and spun it up to start an update which was previously not possible due to missing space *duh* < I'm with stupid... - during the update in the vm froze (obviously due to unraid setting the array to read only to prevent data loss) and I freaked and came here... my learnings so far: - search through the forum before touching your unriad if you plan on doing something you have never done before. Someone else has probably done it already... - the cache is configured as raid 1 per default (had I known before, I wouldn't have just added a 1TB) - don't be so quick to assume, everything will work on it's own So I guess my question now would be more: What data on the cache might still be salvegable? best regards steel primogenitus-diagnostics-20201014-2152.zip Edited October 15, 2020 by steelrat Update to clarify Quote Link to comment
trurl Posted October 15, 2020 Share Posted October 15, 2020 The default btrfs configuration is raid1. With 2 drives, that gives a mirror with total capacity equal to the smaller of the drives. So you have overfilled and corrupted cache since it only has 500G. 1 Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 (edited) Oh shite... Can I actually resolve that or am I royally scr***? best regards and thank you for the help Steel Edited October 15, 2020 by steelrat Should have had a coffee before the initial reply :) Quote Link to comment
Kevek79 Posted October 15, 2020 Share Posted October 15, 2020 4 hours ago, trurl said: The default btrfs configuration is raid1. With 2 drives, that gives a mirror with total capacity equal to the smaller of the drives. So you have overfilled and corrupted cache since it only has 500G. Adding the second drive in raid 1 would give protection and not more space, but if the 500 G drive was 90% full prior to adding the second drive it should not be overfilled just by adding the second drive. @steelrat Did you add any data to the cache after you added the second drive or did the docker and vm services go south directly after reboot. 1 Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 There was data added to the cache, I'm afraid to say... regards Steel Quote Link to comment
Kevek79 Posted October 15, 2020 Share Posted October 15, 2020 Recovering from that state is out of my ballpark. @JorgeB Is there any chance for @steelrat to recover whatever was on the cache? 1 Quote Link to comment
JorgeB Posted October 15, 2020 Share Posted October 15, 2020 Try this: -stop the array -disable docker and VM services -start array, delete/move a couple of GBs from cache then balance to single profile 1 Quote Link to comment
JorgeB Posted October 15, 2020 Share Posted October 15, 2020 BTW, if there's nothing else to delete you can delete the docker image, should be enough and it can easily be recreated later. Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 11 minutes ago, JorgeB said: Try this: -stop the array -disable docker and VM services -start array, delete/move a couple of GBs from cache then balance to single profile Thank you for the help! I should be able to free enough space, will try and update the thread. regards Steel Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 (edited) Quote -start array, delete/move a couple of GBs from cache then balance to single profile @JorgeB Is this the correct command to balance to single profile? btrfs balance start -dconvert=single -mconvert=raid1 /mnt/cache I followed your advice, deleted unnecessary data and should now have enough space to da a balance with the desired outcome. best regards Steel Edited October 15, 2020 by steelrat typo Quote Link to comment
JorgeB Posted October 15, 2020 Share Posted October 15, 2020 You can do it in the GUI, click on cache and scroll down to the balance section. 1 Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 Running - keeping my fingers crossed 🤞 Quote Link to comment
steelrat Posted October 15, 2020 Author Share Posted October 15, 2020 Yay! Everything up and running again 👍 Thank you so much @Kevek79 and @JorgeB as well as @trurl The help was much appreciated! Next Steps for me: - Save Backup Cache to Array. I hope that (CA Backup / Restore Appdata) will do what I need here. - Remove the Cache alltogether - start over with a second 1TB drive in Raid1 Anything I forgot? best regards Steel 2 Quote Link to comment
Kevek79 Posted October 15, 2020 Share Posted October 15, 2020 Good to hear that your Server is back up and running. Please tag your Thread as (SOLVED) if you feel comfortable that the issue is solved. and BTW - Welcome to the Unraid Forum 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.