Bug Report: pg_cron.sh references /var/lib/postgresql/data/ directly

[joeshin]

Describe the bug

When PGDATA env var is set to override the default data directory, the container errors on startup:

/usr/local/bin/docker-entrypoint.sh: running /docker-entrypoint-initdb.d/pg_cron.sh
shared_buffers = 512MB
max_connections = 200
max_prepared_transactions = 200
max_locks_per_transaction = 512
max_pred_locks_per_transaction = 256
listen_addresses = '*'
shared_preload_libraries = 'pg_cron'
cron.database_name = 'circle_test'
grep: /var/lib/postgresql/data/postgresql.conf: No such file or directory
/docker-entrypoint-initdb.d/pg_cron.sh: line 18: /var/lib/postgresql/data/postgresql.conf: No such file or directory
cat: /var/lib/postgresql/data/postgresql.conf: No such file or directory
cat: /var/lib/postgresql/data/postgresql.conf: No such file or directory

exited with error (exited with 1)

Tracked down this this line:

cimg-postgres/pg_cron.sh

Lines 21 to 22 in 418064b

	cat /var/lib/postgresql/data/postgresql.conf | grep shared
	cat /var/lib/postgresql/data/postgresql.conf | grep include

To Reproduce
Set PGDATA to any non-default dir.

Expected behavior
Env var is used to avoid errors.

Workarounds
Unsure.

[marboledacci]

Hi @joeshin what is the version used and how are you using the image

[dopry]

@marboledacci this is not the first time this issue has arrived. It's also

• Bug Report: PGDATA: /dev/shm/pgdata/data prevent cimg/postgres:15.4 from starting #146
• Bug Report: Support for user-defined PGDATA broken with pg_cron.sh #113

In #113 you stated a new image had been published, but from testing, the images currently pulled by CI are not working as if this fix were in place... As a long time circleci customer, I would really appreciate it if you guys could get the issues around pg_cron and PGDATA fixed. Right now our CI jobs are taking 25-30% longer to complete. It's really annoying.