Bug #436: backend closed the channel unexpectedly

Started by PostgreSQL Bugs Listover 24 years ago2 messagesbugs
Jump to latest
#1PostgreSQL Bugs List
pgsql-bugs@postgresql.org

Michael Ablass (ma@netresearch.de) reports a bug with a severity of 1
The lower the number the more severe it is.

Short Description
backend closed the channel unexpectedly

Long Description
Using Postgresql 7.1.3 on Redhat 7.1 and 6.2, currently we have 2 projects running. a few days the databases do work perfectly - then we get error message following. it seems not to be the problem of certain sql queries as sometimes everything works and sometimes it does. we increased the amount of connections but it did not help. next thing: on redhat 6.2 the database seems to work more unstable than on the redhat 7.1 server.

Warning: PostgreSQL query failed: FATAL 1: This connection has been terminated by the administrator. pqReadData() -- backend closed the channel unexpectedly. This probably means the backend terminated abnormally before or while processing the request. in /shares/maxtor1/home/httpd/server/global_classes/db_pgsql.inc on line 69
Database error: Invalid SQL: SELECT * from offen where o_id='1' AND o_offen='J'
PostgreSQL Error: 1 (FATAL 1: This connection has been terminated by the administrator. pqReadData() -- backend closed the channel unexpectedly. This probably means the backend terminated abnormally before or while processing the request. )
Session halted.

Sample Code

No file was uploaded with this report

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: PostgreSQL Bugs List (#1)
Re: Bug #436: backend closed the channel unexpectedly

pgsql-bugs@postgresql.org writes:

Using Postgresql 7.1.3 on Redhat 7.1 and 6.2, currently we have 2
projects running. a few days the databases do work perfectly - then we
get error message following.

Warning: PostgreSQL query failed: FATAL 1: This connection has been
terminated by the administrator.

This message indicates that the backend got a SIGTERM signal from
someplace. If the postmaster got a SIGINT signal it would turn around
and SIGTERM all the backends, but there's no other mechanism in Postgres
that would cause this to happen.

Is the postmaster still alive after the backends shut down, or does it
quit too? What shows up in the postmaster's stderr log? (If you're
sending stderr to /dev/null, now is a good time to stop doing so.)

regards, tom lane