When LXD 4.5 rolled out it caused my GUI containers to fail, so since then I froze my LXD version on 4.4. Everything worked fine for several months until tonight. When I started the computer the following error appeared:
$ lxc list
Error: Get "http://unix.socket/1.0": dial unix /var/snap/lxd/common/lxd/unix.socket: connect: connection refused
In a related post from April’18 I found the suggestion to have a look at debug output, so here is the error:
$ systemctl stop snap.lxd.daemon
$ lxd --debug --group lxd
DBUG[12-06|00:13:56] Connecting to a local LXD over a Unix socket
DBUG[12-06|00:13:56] Sending request to LXD method=GET url=http://unix.socket/1.0 etag=
INFO[12-06|00:13:56] LXD 4.4 is starting in normal mode path=/var/snap/lxd/common/lxd
INFO[12-06|00:13:56] Kernel uid/gid map:
INFO[12-06|00:13:56] - u 0 0 4294967295
INFO[12-06|00:13:56] - g 0 0 4294967295
INFO[12-06|00:13:56] Configured LXD uid/gid map:
INFO[12-06|00:13:56] - u 0 1000000 1000000000
INFO[12-06|00:13:56] - g 0 1000000 1000000000
INFO[12-06|00:13:56] Kernel features:
INFO[12-06|00:13:56] - closing multiple file descriptros efficiently: no
INFO[12-06|00:13:56] - netnsid-based network retrieval: no
INFO[12-06|00:13:56] - pidfds: no
INFO[12-06|00:13:56] - uevent injection: no
INFO[12-06|00:13:56] - seccomp listener: no
INFO[12-06|00:13:56] - seccomp listener continue syscalls: no
INFO[12-06|00:13:56] - seccomp listener add file descriptors: no
INFO[12-06|00:13:56] - safe native terminal allocation : no
INFO[12-06|00:13:56] - unprivileged file capabilities: yes
INFO[12-06|00:13:56] - cgroup layout: hybrid
WARN[12-06|00:13:56] - Couldn't find the CGroup memory swap accounting, swap limits will be ignored
DBUG[12-06|00:13:56] no such device - Failed to mount shiftfs
INFO[12-06|00:13:56] - shiftfs support: no
INFO[12-06|00:13:56] Initializing local database
DBUG[12-06|00:13:56] Initializing database gateway
DBUG[12-06|00:13:56] Start database node role=voter id=1 address=
EROR[12-06|00:13:57] Failed to start the daemon: Failed to start dqlite server: raft_start(): io: load closed segment 0000000000011160-0000000000011253: found 93 entries (expected 94)
INFO[12-06|00:13:57] Starting shutdown sequence
DBUG[12-06|00:13:57] Not unmounting temporary filesystems (containers are still running)
It looks like there is some issue with the database. I didn’t notice any sign of a disk corruption or any other problems neither during the last shut down nor during the startup. Can someone please help me to fix that and get my containers back running?