So I just ran into this problem trying to run and create some offline and new VMs. Why is the disk being read as full and in a read only state when there is 600 gb free on the disk?
Error 500: Disk is in a read only state
- Thread starter Firecracker048
- Start date
Can you please provide us with more information, where you see that the disk read only and full?
Can you also check the dmesg for the mentioned disk `dmesg | grep -i ext` or xfs if you're using XFS filesystem?
Can you also check the dmesg for the mentioned disk `dmesg | grep -i ext` or xfs if you're using XFS filesystem?
So this is the exact error im getting when trying to run or create a new vm:
starting worker failed: unable to create output file '/var/log/pve/tasks/F/UPID:server:000B72AC:166DBD99:66F20C6F:qmstart:100:root@pam:' - Read-only file system (500)
I don't know how it got this way, There is always 1 VM running and the other two have been in a shutdown state for a few months now.
The /dev/sda3 dis volume group is in an LVM usage
starting worker failed: unable to create output file '/var/log/pve/tasks/F/UPID:server:000B72AC:166DBD99:66F20C6F:qmstart:100:root@pam:' - Read-only file system (500)
I don't know how it got this way, There is always 1 VM running and the other two have been in a shutdown state for a few months now.
The /dev/sda3 dis volume group is in an LVM usage
If I try to access the server shell on the browser UI I get this error message:
Connection failed (Error 500: unable to open file '/var/tmp/pve-reserved-ports.tmp.750271' - Read-only file system)
Connection failed (Error 500: unable to open file '/var/tmp/pve-reserved-ports.tmp.750271' - Read-only file system)
Did you check the dmesg?
Can you please provide us with the output of the following commands:
Bash:
lvs
vgs
pvsReadonly is set by kernel when write failed - so you need a new OS disk !
This disk is a brand new disk. It was working perfectly fine and the running vm works fine
Do i reset the server and the error message went away but here's both the error message and the wanted output.
https://imgur.com/gallery/vwFIIXO
https://imgur.com/gallery/vwFIIXO
A "brand new disk" means nothing if you didn't put it thru a burn-in test. Even SSD can have bad chips from the factory. Always at least do a SMART long test - and check the output afterward - before putting any disk into use.
Check /etc/fstab , if you're using ext4 for root FS then the default option is to remount it R/O when it encounters an error.
If this is spinning disk then you should backup all VMs/LXCs, replace it with a new disk (that has undergone a burn-in test) and reinstall, and restore your environment.
If SSD, run a SMART long test on it. Odds are you will have to RMA it and do the same ^^ as above.
Typical minimal burn-in test for spinners is to DD zeros to the entire drive (obv this wipes out all data on it) followed by a SMART long test.
Some people like to do multiple passes (which I don't support - as it just puts unnecessary wear/tear on the drive, and can take DAYS with multi-TB modern drives) but the basic test should at least weed out shipping damage.
https://github.com/kneutron/ansitest/blob/master/SMART/scandisk-bigdrive-2tb+.sh
Check /etc/fstab , if you're using ext4 for root FS then the default option is to remount it R/O when it encounters an error.
If this is spinning disk then you should backup all VMs/LXCs, replace it with a new disk (that has undergone a burn-in test) and reinstall, and restore your environment.
If SSD, run a SMART long test on it. Odds are you will have to RMA it and do the same ^^ as above.
Typical minimal burn-in test for spinners is to DD zeros to the entire drive (obv this wipes out all data on it) followed by a SMART long test.
Some people like to do multiple passes (which I don't support - as it just puts unnecessary wear/tear on the drive, and can take DAYS with multi-TB modern drives) but the basic test should at least weed out shipping damage.
https://github.com/kneutron/ansitest/blob/master/SMART/scandisk-bigdrive-2tb+.sh