I'm noticing an issue with certain RPC after upgrading to 5.6. I'm using "db_postgres".
If I run ul.flush like this:
kamcmd ul.flush
# no output
None of the in-memory registrations are flushed to the DB and in the log I see a bunch of lines like this:
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: <core> [db.c:481]: db_use_table(): invalid connection parameter
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: usrloc [ucontact.c:1142]: db_update_ucontact_ruid(): sql use_table failed
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: usrloc [urecord.c:433]: wb_timer(): updating contact in db failed (aor: fake-1@aor.com)
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: <core> [db.c:481]: db_use_table(): invalid connection parameter
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: usrloc [ucontact.c:1142]: db_update_ucontact_ruid(): sql use_table failed
May 12 20:39:16 <dummy-hostname> /usr/sbin/kamailio[9006]: ERROR: usrloc [urecord.c:433]: wb_timer(): updating contact in db failed (aor: fake-2@aor.com)
I see the same thing with ul.db_users:
kamcmd ul.db_users location
# returns the following output
error: 500 - Failed to use table
# in the log
ERROR: <core> [db.c:481]: db_use_table(): invalid connection parameter
Seems to happen consistently on 5.6, using psql as the db engine. I tested on 5.4 and it's not happening.
It's easier to test with kamcmd ul.db_users location
because you need registrations that haven't been flushed to the DB for the error to manifest.
Distributor ID: Debian
Description: Debian GNU/Linux 10 (buster)
Release: 10
Codename: buster
Kernel:
4.19.0-21-cloud-amd64
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.