Uncontrolled shutdown of the PostgreSQL container

kr3ch3t · January 27, 2026, 9:12am

Hi! From time to time I am facing a problem like this:

pg-gvm-1 | 2026-01-27 01:32:00.527 UTC [1] LOG: received smart shutdown request
pg-gvm-1 | 2026-01-27 01:32:01.727 UTC [178231] gvmd@gvmd FATAL: the database system is shutting down
pg-gvm-1 | 2026-01-27 01:32:01.729 UTC [178232] gvmd@gvmd FATAL: the database system is shutting down
pg-gvm-1 | 2026-01-27 01:32:06.737 UTC [178233] gvmd@gvmd FATAL: the database system is shutting down
pg-gvm-1 | Removed stale pid file.
pg-gvm-1 | Error: /usr/lib/postgresql/17/bin/pg_ctl /usr/lib/postgresql/17/bin/pg_ctl start -D /var/lib/postgresql/17/main -l /var/log/postgresql/postgresql-17-main.log -s -o -c config_file=“/etc/postgresql/17/main/postgresql.conf” -k /var/run/postgresql -c listen_addresses=‘’ exited with status 1:
pg-gvm-1 | 2026-01-27 01:32:20.135 UTC [29] LOG: starting PostgreSQL 17.7 (Debian 17.7-0+deb13u1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 14.2.0-19) 14.2.0, 64-bit
pg-gvm-1 | 2026-01-27 01:32:20.138 UTC [29] LOG: listening on IPv4 address “0.0.0.0”, port 5432
pg-gvm-1 | 2026-01-27 01:32:20.138 UTC [29] LOG: listening on IPv6 address “::”, port 5432
pg-gvm-1 | 2026-01-27 01:32:20.187 UTC [29] FATAL: lock file “/var/run/postgresql/.s.PGSQL.5432.lock” already exists
pg-gvm-1 | 2026-01-27 01:32:20.187 UTC [29] HINT: Is another postmaster (PID 1) using socket file “/var/run/postgresql/.s.PGSQL.5432”?
pg-gvm-1 | 2026-01-27 01:32:20.189 UTC [29] LOG: database system is shut down
pg-gvm-1 | pg_ctl: could not start server
pg-gvm-1 | Examine the log output.
pg-gvm-1 | Error: /usr/lib/postgresql/17/bin/pg_ctl /usr/lib/postgresql/17/bin/pg_ctl start -D /var/lib/postgresql/17/main -l /var/log/postgresql/postgresql-17-main.log -s -o -c config_file=“/etc/postgresql/17/main/postgresql.conf” -k /var/run/postgresql -c listen_addresses='’ exited with status 1:
pg-gvm-1 | 2026-01-27 01:32:21.205 UTC [29] LOG: starting PostgreSQL 17.7 (Debian 17.7-0+deb13u1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 14.2.0-19) 14.2.0, 64-bit
pg-gvm-1 | 2026-01-27 01:32:21.205 UTC [29] LOG: listening on IPv4 address “0.0.0.0”, port 5432
pg-gvm-1 | 2026-01-27 01:32:21.206 UTC [29] LOG: listening on IPv6 address “::”, port 5432
pg-gvm-1 | 2026-01-27 01:32:21.237 UTC [29] FATAL: lock file “/var/run/postgresql/.s.PGSQL.5432.lock” already exists
pg-gvm-1 | 2026-01-27 01:32:21.237 UTC [29] HINT: Is another postmaster (PID 1) using socket file “/var/run/postgresql/.s.PGSQL.5432”?
pg-gvm-1 | 2026-01-27 01:32:21.239 UTC [29] LOG: database system is shut down
pg-gvm-1 | pg_ctl: could not start server
pg-gvm-1 | Examine the log output.

…

There were no scans, updates, or any other changes at this time.

P.S. I’d like to describe the problem in as much detail as possible, but that’s really all I have right now.

bricks · January 27, 2026, 9:54am

You can try to delete the psql_socket_vol volume to fix the issue. It gets recreated on startup again automatically.

pat52538 · January 28, 2026, 10:27am

i’ve got the same problem. i can delete file and start the container. but after a reboot i got the same again

serhiime · January 28, 2026, 8:00pm

The same issue after the server rebooted on Ubuntu 25.04 clear fresh install of greenbone

docker logs greenbone-community-edition-pg-gvm-1

Error: /usr/lib/postgresql/17/bin/pg_ctl /usr/lib/postgresql/17/bin/pg_ctl start -D /var/lib/postgresql/17/main -l /var/log/postgresql/postgresql-17-main.log -s -o -c config_file=“/etc/postgresql/17/main/postgresql.conf” -k /var/run/postgresql -c listen_addresses=‘*’ exited with status 1:
2026-01-28 16:22:50.974 UTC [30] LOG: starting PostgreSQL 17.7 (Debian 17.7-0+deb13u1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 14.2.0-19) 14.2.0, 64-bit
2026-01-28 16:22:50.974 UTC [30] LOG: listening on IPv4 address “0.0.0.0”, port 5432
2026-01-28 16:22:50.974 UTC [30] LOG: listening on IPv6 address “::”, port 5432
2026-01-28 16:22:50.976 UTC [30] FATAL: lock file “/var/run/postgresql/.s.PGSQL.5432.lock” already exists
2026-01-28 16:22:50.976 UTC [30] HINT: Is another postmaster (PID 1) using socket file “/var/run/postgresql/.s.PGSQL.5432”?
2026-01-28 16:22:50.981 UTC [30] LOG: database system is shut down
pg_ctl: could not start server
Examine the log output.

Please provide how to fix this issue

serhiime · January 28, 2026, 8:00pm

when adding this command and restart container socket removed and then againe restart and container is up. but after server reboot - the same

kr3ch3t · January 29, 2026, 3:41am

Perhaps you have an old psql_socket_vol (with a different prefix in the name) among docker volumes (docker volume ls). I deleted them all, and then I raised the containers again (docker compose down, docker volume rm psql_socket_vol, docker compose up -d) everything still seems to be working without errors.

P.S. If the error suddenly appears again, I will write here.

bricks · January 29, 2026, 8:57am

I don’t have a clue what’s happening on a reboot differently form a container restart. Somebody from the community needs to do some investigation. Maybe the containers are not shut down correctly and postgres can’t remove the lock?

Until that there is a workaround by removing the pg docker volume.

bricks · January 29, 2026, 8:59am

It seems to me just restarting docker causes the same issue.

serhiime · January 29, 2026, 10:32am

I’m confirmed too. Here the logs

Please provide any decision how to fix this issue.

kr3ch3t · January 29, 2026, 10:43am

Have you tried removing psql_socket_vol as @bricks wrote and starting the containers again? At least it helped me and it still works.

serhiime · January 29, 2026, 11:55am

Sure. docker-compose down and then delete posgresql-socket volume, run docker-compose up -d. It starts, but, againe after restart docker or reboot the server got the same issue.

kr3ch3t · January 29, 2026, 12:19pm

Do you have another psql_socket_vol (may be with another prefix) in your docker volumes (docker volume ls)?

bricks · January 29, 2026, 12:24pm

As I wrote we have a workaround.

If you have a high demand for a fix, please investigate the issue and let me know about the error cause. In that case I can fix it. Currently investigation of this issue is not on my prio list.

serhiime · January 29, 2026, 12:27pm

No, only one, wich created by default

kr3ch3t · January 29, 2026, 12:49pm

I’ve just had this problem come back on its own. I think you should still consider it serious enough to fix it soon)

As I wrote earlier, there is no additional information other than the logs I provided. The containers have also not been rebooted since the last time I resorted to your temporary solution and reported the completion of the problem.

bricks · January 29, 2026, 1:46pm

I need to know why the postgres lock file is not removed automatically and what we are doing wrong. That needs some investigation. I am not able to spend time on this at the moment.

serhiime · January 30, 2026, 9:19am

I’m using Ubuntu 25.04, Docker version 29.2.0, Docker Compose version v5.0.2.

Maybe problem that 2 containers with psql socket

pg-gvm:
image: registry.community.greenbone.net/community/pg-gvm:stable
restart: on-failure:10
volumes:

psql_data_vol:/var/lib/postgresql
psql_socket_vol:/var/run/postgresql
depends_on:
pg-gvm-migrator:
condition: service_completed_successfully

pg-gvm-migrator:
image: registry.community.greenbone.net/community/pg-gvm-migrator:stable
restart: no
volumes:

psql_data_vol:/var/lib/postgresql
psql_socket_vol:/var/run/postgresql

docker volume ls

but container pg migrator aren’t running

pat52538 · February 2, 2026, 3:59pm

I commented the pg-gvm-migrator out in the yml-file and then pg-gvm-container starts normaly.

bricks · February 3, 2026, 9:39am

This can’t be the solution. If the migration is done, the container actually does nothing on startup.

serhiime · February 3, 2026, 9:47am

Do the same and got the same error Error: /usr/lib/postgresql/17/bin/pg_ctl /usr/lib/postgresql/17/bin/pg_ctl start -D /var/lib/postgresql/17/main -l /var/log/postgresql/postgresql-17-main.log -s -o -c config_file=“/etc/postgresql/17/main/postgresql.conf” -k /var/run/postgresql -c listen_addresses=‘*’ exited with status 1:
2026-02-03 09:15:19.346 UTC [29] LOG: starting PostgreSQL 17.7 (Debian 17.7-0+deb13u1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 14.2.0-19) 14.2.0, 64-bit
2026-02-03 09:15:19.347 UTC [29] LOG: listening on IPv4 address “0.0.0.0”, port 5432
2026-02-03 09:15:19.347 UTC [29] LOG: listening on IPv6 address “::”, port 5432
2026-02-03 09:15:19.349 UTC [29] FATAL: lock file “/var/run/postgresql/.s.PGSQL.5432.lock” already exists
2026-02-03 09:15:19.349 UTC [29] HINT: Is another postmaster (PID 1) using socket file “/var/run/postgresql/.s.PGSQL.5432”?
2026-02-03 09:15:19.354 UTC [29] LOG: database system is shut down

Looks life socket doesn’t normally deleted after container with posgres stops/restarts