Re: pgsql 7.2.3 crash - Mailing list pgsql-hackers

From Laurette Cisneros
Subject Re: pgsql 7.2.3 crash
Date
Msg-id Pine.LNX.4.44.0210091309580.14627-100000@visor.corp.nextbus.com
Whole thread Raw
In response to Re: pgsql 7.2.3 crash  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: pgsql 7.2.3 crash  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Wed, 9 Oct 2002, Tom Lane wrote:

> Laurette Cisneros <laurette@nextbus.com> writes:
> > A core file was found in <datadir>/base/326602604
> > and a backtrace shows:
> > (gdb) bt
> > #0  DeferredTriggerSaveEvent (relinfo=0x83335f0, event=0, oldtup=0x0,
> >     newtup=0x8348150) at trigger.c:2056
> 
> Hm.  Line 2056 is this:
> 
>     for (i = 0; i < ntriggers; i++)
>     {
>         Trigger    *trigger = &trigdesc->triggers[tgindx[i]];
> 
> ->        new_event->dte_item[i].dti_tgoid = trigger->tgoid;
> 
> It seems there must be something wrong with the trigdesc data structure
> for that table, but what?  Can you poke around in the corefile with gdb
> print commands and determine what's wrong with the trigdesc?

Here's my poking/printing around with gdb:
(gdb) print trigger
$1 = (Trigger *) 0x1272c9a0
(gdb) print *trigger
Cannot access memory at address 0x1272c9a0
(gdb) print trigger->tgoid
Cannot access memory at address 0x1272c9a0
(gdb) print trigdesc
$2 = (TriggerDesc *) 0x4c86b2d8
(gdb) print &trigdesc
$3 = (TriggerDesc **) 0xbfffea18 
(gdb) print *trigdesc 
$4 = {n_before_statement = {5378, 22310, 37184, 2085}, n_before_row =
{51128, 19585,   62320, 19589}, n_after_row = {45784, 19590, 52748, 2084},
n_after_statement = {   0, 0, 0, 0}, tg_before_statement = {0x4c86b2d8, 0x8259910, 0x0, 0x0}, tg_before_row = {0xa0,
0x350000,0x1f, 0x4006}, tg_after_row =
 
{0x8242920,   0x4c860cb0, 0x4c86eec0, 0x0}, tg_after_statement = {0x0, 0x0, 0x0,
0x0}, triggers = 0x4c86e7a0, numtriggers = 8}

> 
> > I see triggers referenced here and it should be
> > noted that for one of the tables the triggers were first disabled (update
> > pg_class) and re-enabled after the inserts are done (or it takes
> > forever).
> 
> Did that happen while this backend was running?

Yes.  I had run this perl program about 4-5 times in a row (which includes
the sequence, disable triggers, insert rows, enable triggers) and then it
crashed on one of the runs.

Thanks,

-- 
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
------------------------------
It's 10 o'clock...
Do you know where your bus is?



pgsql-hackers by date:

Previous
From: Rod Taylor
Date:
Subject: Re: [GENERAL] Point in Time Recovery WAS: Hot Backup
Next
From: "Magnus Naeslund(f)"
Date:
Subject: Damn slow query