No space left on /var/cache/edb

[rvj]

I need assistance to understand the following. This is caused by: emerge --ask --verbose --update --deep --newuse --keep-going @world. All packages emerged successfully until the next to the last one - firefox - with the following error message:

Traceback (most recent call last):
File "/usr/lib/python-exec/python3.12/emerge", line 57, in main
retval = emerge_main()
^^^^^^^^^^^^^
File "/usr/lib/python3.12/site-packages/_emerge/main.py", line 1308, in emerge_main
return run_action(emerge_config)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.12/site-packages/_emerge/actions.py", line 4033, in run_action
retval = action_build(emerge_config, spinner=spinner)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.12/site-packages/_emerge/actions.py", line 660, in action_build
mtimedb.commit()
File "/usr/lib/python3.12/site-packages/portage/util/mtimedb.py", line 127, in commit
self.__write_to_disk(d)
File "/usr/lib/python3.12/site-packages/portage/util/mtimedb.py", line 147, in __write_to_disk
f.close()
File "/usr/lib/python3.12/site-packages/portage/util/__init__.py", line 1526, in close
os.rename(f.name, real_name)
OSError: [Errno 28] No space left on device: b'/var/cache/edb/mtimedb.2915' -> '/var/cache/edb/mtimedb'

The root partition is only 66% full:

Check on inodes:

df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/nvme0n1p6 0 0 0 - /
devtmpfs 2033649 541 2033108 1% /dev
tmpfs 819200 843 818357 1% /run
efivarfs 0 0 0 - /sys/firmware/efi/efivars
shm 2034186 1 2034185 1% /dev/shm
tmpfs 406837 33 406804 1% /run/user/1000

The user files are on a separate partition, but everything else is on the root partition.

Please help suggest next steps. Thanks

[grknight]

What filesystem is this? If it is btrfs, then it may need some maintenance as certain balance and similar commands may be necessary.

[rvj]

Yes, the filesystem is btrfs.

[grknight]

I would suggest reading this from the (archived) Btrfs wiki about this exact problem

[rvj]

Yes, that seems to describe the problem.

btrfs fi df /mnt/gentoo
Data, single: total=68.85GiB, used=43.67GiB
System, DUP: total=8.00MiB, used=16.00KiB
Metadata, DUP: total=2.00GiB, used=1.87GiB
GlobalReserve, single: total=129.86MiB, used=0.00B

shows that metadata is well over 90% used. When I attempt to apply the remedy, I get the following:

btrfs balance start -dusage=5 /mnt/gentoo/
ERROR: error during balancing '/mnt/gentoo/': No space left on device
There may be more info in syslog - try dmesg | tail

dmesg | tail
[ 272.539725] BTRFS info (device nvme0n1p6): using crc32c (crc32c-generic) checksum algorithm
[ 272.539732] BTRFS info (device nvme0n1p6): using free space tree
[ 272.559455] BTRFS info (device nvme0n1p6): enabling ssd optimizations
[ 272.559461] BTRFS info (device nvme0n1p6): auto enabling async discard
[ 549.409104] BTRFS info (device nvme0n1p6): balance: start -dusage=5
[ 549.409171] BTRFS info (device nvme0n1p6): relocating block group 79301705728 flags data
[ 549.435800] BTRFS info (device nvme0n1p6): relocating block group 76266078208 flags data
[ 549.470845] BTRFS info (device nvme0n1p6): relocating block group 62307434496 flags data
[ 549.504834] BTRFS info (device nvme0n1p6): 3 enospc errors during balance
[ 549.504841] BTRFS info (device nvme0n1p6): balance: ended with status: -28

Any recommendations?

[RumpletonBongworth]

SUSE has a fairly detailed knowledge base article concerning this matter.