stale portmaster processes

On Sat, 13 Jan 2001, Thomas T. Thai wrote:

Date: Sat, 13 Jan 2001 01:12:29 -0600 (CST)
From: Thomas T. Thai <tom@minnesota.com>
To: PostgreSQL General <pgsql-general@postgresql.org>
Subject: stale portmaster processes

i've noticed that when using php scripts or the "c" cgi search.cgi from
mnogosearch, if i access either script from the web browser and stop it
with ESC in the middle of it trying to access the pgsql db, the associated
postmaster process hangs around forever and won't disappear:

19276 p1 I 0:00.11 postmaster: web mnwork [local] idle (postgres)
19920 p1 I 0:00.08 postmaster: web mnwork [local] idle (postgres)
19958 p1 I 0:00.10 postmaster: web mnwork [local] idle (postgres)

i hope this isn't normal.

i've not yet seen any stale idling postmaster processes with the latest
cvs source. the above was from mid december cvs source where i saw the
stale postmasters (with --enable-debug).

Thomas T. Thai writes:

i've noticed that when using php scripts or the "c" cgi search.cgi from
mnogosearch, if i access either script from the web browser and stop it
with ESC in the middle of it trying to access the pgsql db, the associated
postmaster process hangs around forever and won't disappear:

19276 p1 I 0:00.11 postmaster: web mnwork [local] idle (postgres)
19920 p1 I 0:00.08 postmaster: web mnwork [local] idle (postgres)
19958 p1 I 0:00.10 postmaster: web mnwork [local] idle (postgres)

These look like backend processes left around as part of the persistent
connection thing that PHP does. On the next connection they will get
reused, or so is the theory.

--
Peter Eisentraut peter_e@gmx.net http://yi.org/peter-e/

On Sun, 14 Jan 2001, Peter Eisentraut wrote:

Thomas T. Thai writes:

i've noticed that when using php scripts or the "c" cgi search.cgi from
mnogosearch, if i access either script from the web browser and stop it
with ESC in the middle of it trying to access the pgsql db, the associated
postmaster process hangs around forever and won't disappear:

19276 p1 I 0:00.11 postmaster: web mnwork [local] idle (postgres)
19920 p1 I 0:00.08 postmaster: web mnwork [local] idle (postgres)
19958 p1 I 0:00.10 postmaster: web mnwork [local] idle (postgres)

These look like backend processes left around as part of the persistent
connection thing that PHP does. On the next connection they will get
reused, or so is the theory.

most of the time they didn't get reused. some were from a C based cgi
program too. i tried leaving them there for days and they never
disappeared (no further access from php scripts or cgi)