Re: server won't shutdown

I've been seeing the same problem intermittently over the past 2 years
among ~30 production clusters. Today I saw it happen with 3-4 of the
7.2.3 clusters on the same host (PostgreSQL 7.2.3 on
i686-pc-linux-gnu, compiled by GCC 2.96 with 2.4.18-4bigmem kernel).
Each cluster had a number of long-idle connections, no current
activity. Will try to gather a little more data...

Ed

On Monday February 17 2003 3:56, Laurette Cisneros wrote:
> Ok, my production server had this problem (again). So, here's the
> info. you requested. There's a lot of postgres processes still
> left running:
>
> postmaster (proc id 1457):
> (gdb) print Shutdown
> $1 = 2
>
> Other backends (postgres) still hanging around:
> postgres 1458 1457 0 Feb06 ? 00:22:41 postgres: stats
> buffer process InterruptPending: $1 = 0
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 0
> ImmediateInterruptOK = $4 = 0
> InterruptHoldoffCount = $5 = 0
> CritSectionCount = $6 = 0
>
> postgres 1459 1458 2 Feb06 ? 05:31:38 postgres: stats
> collector process InterruptPending: $1 = 0
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 0
> ImmediateInterruptOK = $4 = 0
> InterruptHoldoffCount = $5 = 0
> CritSectionCount = $6 = 0
>
> postgres 1468 1457 0 Feb06 ? 00:04:57 postgres: nbadmin
> db1 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK = $4 = 0
> InterruptHoldoffCount = $5 = 0
> CritSectionCount = $6 = 0
>
> postgres 1461 1457 0 Feb06 ? 00:02:06 postgres: nbadmin
> db2 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 1478 1457 0 Feb06 ? 00:04:13 postgres: nbadmin
> db3 111.123.124.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 9818 1457 0 Feb06 ? 00:00:01 postgres: nbadmin
> db4 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 15677 1457 0 Feb06 ? 00:00:04 postgres: nbadmin
> db5 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 30207 1457 0 Feb09 ? 00:00:00 postgres: nbadmin
> db6 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 7316 1457 0 Feb11 ? 00:00:01 postgres: nbadmin
> db7 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 10440 1457 0 Feb12 ? 00:00:00 postgres: nbadmin
> db8 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 10441 1457 0 Feb12 ? 00:02:15 postgres: nbadmin
> db9 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 27355 1457 0 Feb13 ? 00:00:00 postgres: nbadmin
> db10 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 28491 1457 0 Feb13 ? 00:00:03 postgres: nbadmin
> db11 111.123.123.253 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 30021 1457 0 Feb14 ? 00:00:01 postgres: nbadmin
> db12 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 10663 1457 0 Feb14 ? 00:00:00 postgres: nbadmin
> db13 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> postgres 10665 1457 0 Feb14 ? 00:00:00 postgres: nbadmin
> db13 111.123.123.244 idle InterruptPending: $1 = 65537
> proc_exit_inprogress: $2 = 0
> ProcDiePending: $3 = 1
> ImmediateInterruptOK: $4 = 0
> InterruptHoldoffCount: $5 = 0
> CritSectionCount: $6 = 0
>
> Thank again,
>
> Laurette
>
> On Thu, 13 Feb 2003, Tom Lane wrote:
> > Laurette Cisneros <laurette(at)nextbus(dot)com> writes:
> > > Ok, it reproduced today on my development server.
> >
> > Boy, it sure looks like everything's just sitting idle waiting
> > for client input, and hasn't noticed your shutdown request at
> > all. You sure you used "-m fast"? Could you attach to the
> > postmaster and see what its Shutdown variable contains, and to
> > the backends and see what the variables manipulated by die()
> > contain? (InterruptPending, proc_exit_inprogress,
> > ProcDiePending, ImmediateInterruptOK, InterruptHoldoffCount,
> > CritSectionCount)
> >
> > regards, tom lane