Re: BUG #1015: Got a signal 11 while trying to create a temp table

Ok, just now I understand that 'ulimit' is a bashbuiltin :P. So I fixed that
;)
I reproduced the crash and got a good core-file (with the symbols, see
below). The only thing is: it's 390Mb :( .. even after gzip it's big: 250Mb.
Would that be a problem for you? (I have the bandwith if you want)

(gdb) bt
#0 0x0819b684 in LogicalTapeWrite ()
#1 0x0819da13 in ApplySortFunction ()
#2 0x0819cf10 in tuplesort_getdatum ()
#3 0x080ea656 in ExecSort ()
#4 0x080e07ae in ExecProcNode ()
#5 0x080eaf6f in ExecGroup ()
#6 0x080e07ef in ExecProcNode ()
#7 0x080e5302 in ExecAgg ()
#8 0x080e07bb in ExecProcNode ()
#9 0x080df3a5 in ExecutorEnd ()
#10 0x080de8ac in ExecutorRun ()
#11 0x0813c4bd in ProcessQuery ()
#12 0x0813a898 in pg_exec_query_string ()
#13 0x0813b93f in PostgresMain ()
#14 0x0811f37b in ClosePostmasterPorts ()
#15 0x0811ee03 in ClosePostmasterPorts ()
#16 0x0811dc36 in PostmasterMain ()
#17 0x0811d3a6 in PostmasterMain ()
#18 0x080f5113 in main ()
#19 0x42015704 in __libc_start_main () from /lib/tls/libc.so.6
(gdb)

These settings are used:
/proc/sys/kernel/shmmax = 500000000
shared_buffers = 33000
sort_mem = 98304
fsync = false

The explain's of both query's:

stats=# explain select * from blocks;
QUERY PLAN
---------------------------------------------------------------------------
Seq Scan on blocks (cost=100000000.00..100000020.00 rows=1000 width=130)
(1 row)

stats=# explain select hostid, sum(amount) from blocks group by hostid;
QUERY PLAN
----------------------------------------------------------------------------
---------------
Aggregate (cost=100000069.83..100000077.33 rows=100 width=8)
-> Group (cost=100000069.83..100000074.83 rows=1000 width=8)
-> Sort (cost=100000069.83..100000072.33 rows=1000 width=8)
Sort Key: hostid
-> Seq Scan on blocks (cost=100000000.00..100000020.00
rows=1000 width=8)
(5 rows)

stats=#

I hope this gives you some more info,

Aarjan

----- Original Message -----
From: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "aarjan langereis" <a(dot)j(dot)langereis(at)chello(dot)nl>
Cc: <pgsql-bugs(at)postgresql(dot)org>
Sent: Sunday, December 21, 2003 6:13 PM
Subject: Re: [BUGS] BUG #1015: Got a signal 11 while trying to create a temp
table

> "aarjan langereis" <a(dot)j(dot)langereis(at)chello(dot)nl> writes:
> > I don't see that "ulimit -c 0".. nowhere in the script. Where do I have
to
> > put the "ulimit -c unlimited"?
>
> I'd suggest adding it to the pg_ctl script.
>
> > Select * from blocks; gave me the whole table (I didn't look at all
records,
> > but got a result in psql)
> > select hostid, sum(amount) from blocks group by hostid; crashed (3,2M
> > records used)
>
> Hm. What do you have sort_mem set to? Does the EXPLAIN output show
> different plans for these queries?
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: the planner will ignore your desire to choose an index scan if your
> joining column's datatypes do not match
>