Обсуждение: BUG #7676: pgSocketCheck dosen`t return

The following bug has been logged on the website:

Bug reference:      7676
Logged by:          yonchao
Email address:      yongchao.xu@newtouch.cn
PostgreSQL version: 8.4.6
Operating system:   CentOS 5.5=EF=BC=8864bit=EF=BC=89
Description:        =


I have this problem today: one thread of our application hung, and
we found the thread waiting for the result of poll() function. poll()
was called by pgSocketCheck(). So is it a bug of pg? I found the bug
report from
website(http://archives.postgresql.org/pgsql-bugs/2008-02/msg00260.php),
and want to know:

1. Has this problem been solved in the newest version?
2. If not, what shall I do to avoid this problem?

Thanks a lot!

PS: =

gdb info:
(gdb) bt
#0 0x00d31402 in __kernel_vsyscall ()
#1 0x003d3dc3 in poll () from /lib/i686/nosegneg/libc.so.6
#2 0x0013de5d in pqSocketCheck () from /usr/local/pgsql/lib/libpq.so.5
#3 0x0013df8d in pqWaitTimed () from /usr/local/pgsql/lib/libpq.so.5
#4 0x0013e003 in pqWait () from /usr/local/pgsql/lib/libpq.so.5
#5 0x0013d792 in PQgetResult () from /usr/local/pgsql/lib/libpq.so.5


--

yongchao.xu@newtouch.cn writes:
> I have this problem today: one thread of our application hung, and
> we found the thread waiting for the result of poll() function. poll()
> was called by pgSocketCheck(). So is it a bug of pg?

So far as anyone can tell, this trace just indicates that the
application is waiting for a response from the server.  If there's
a bug there, you'd need to provide more or different information to
prove it.  I think more likely you need to be consulting
http://wiki.postgresql.org/wiki/Slow_Query_Questions

            regards, tom lane

On Monday, November 19, 2012 11:31 AM yongchao.xu@newtouch.cn wrote:
> The following bug has been logged on the website:
>=20
> Bug reference:      7676
> Logged by:          yonchao
> Email address:      yongchao.xu@newtouch.cn
> PostgreSQL version: 8.4.6
> Operating system:   CentOS 5.5=EF=BC=8864bit=EF=BC=89
> Description:
>=20
> I have this problem today: one thread of our application hung, and
> we found the thread waiting for the result of poll() function. poll()
> was called by pgSocketCheck(). So is it a bug of pg? I found the bug
> report from
> =
website(http://archives.postgresql.org/pgsql-bugs/2008-02/msg00260.php),
> and want to know:
>=20
> 1. Has this problem been solved in the newest version?
> 2. If not, what shall I do to avoid this problem?
>=20
> Thanks a lot!
>=20
> PS:
> gdb info:
> (gdb) bt
> #0 0x00d31402 in __kernel_vsyscall ()
> #1 0x003d3dc3 in poll () from /lib/i686/nosegneg/libc.so.6
> #2 0x0013de5d in pqSocketCheck () from /usr/local/pgsql/lib/libpq.so.5
> #3 0x0013df8d in pqWaitTimed () from /usr/local/pgsql/lib/libpq.so.5
> #4 0x0013e003 in pqWait () from /usr/local/pgsql/lib/libpq.so.5
> #5 0x0013d792 in PQgetResult () from /usr/local/pgsql/lib/libpq.so.5

I have also observed similar problem in some of tests, but it appears to =
be hung due to network break and it came out after 15-20 minutes.

With Regards,
Amit Kapila.=20