Just updated Pi-Hole, now unusable

Help docker
1

Hello!

I just pulled the latest image and pihole has been 90% unusable since. I get no response from web queries for minutes at a time, then brief periods where I can connect to things.

Here's the logs:

https://tricorder.pi-hole.net/TO660KBd/

I deleted pihole-FTL.db since my search indicated it might be a corrupt database. That worked for a while, an hour or two, before I started getting the same errors again.

The logs just cycle the same problems over and over:

2022-04-12T21:13:30.392743740Z [2022-04-12 16:13:29.142 18542M] Error while trying to close database: database is locked
pihole | 2022-04-12T21:13:30.392856034Z [2022-04-12 16:13:29.295 18542M] Resizing "FTL-queries" from 63995904 to (1146880 * 56) == 64225280 (/dev/shm: 64.5MB used, 67.1MB total, FTL uses 64.5MB)
pihole | 2022-04-12T21:13:30.392921661Z [2022-04-12 16:13:29.295 18542M] WARNING: RAM shortage (/dev/shm) ahead: 96% is used (/dev/shm: 64.5MB used, 67.1MB total, FTL uses 64.5MB)
pihole | 2022-04-12T21:13:30.393046269Z [2022-04-12 16:13:30.296 18542M] add_message(type=7, message=/dev/shm) - SQL error step DELETE: database is locked

and

pihole | 2022-04-12T21:14:32.627352843Z [2022-04-12 16:14:31.204 18607M] Resizing "FTL-queries" from 41746432 to (749568 * 56) == 41975808 (/dev/shm: 42.3MB used, 67.1MB total, FTL uses 42.2MB)

Dozens and dozens of these errors per block, each block repeated every 15 seconds or so. Deleting pihole-FTL.db makes the errors go away again, but only for a few hours.

Help?
Seth

2

It doesn't look like RAM is an actual issue. Here's the readout from 'free -h'

               total        used        free      shared  buff/cache   available
Mem:           3.7Gi       378Mi       1.0Gi        12Mi       2.3Gi       3.2Gi
Swap:           99Mi          0B        99Mi
3

Finally, I had to change the DNS servers back to not-pihole so I could get in and run the debug and post it here. That will probably show in the debug output.