On 5 August 2011 23:49, Christopher Browne <cbbrowne at afilias.info> wrote:
> That's presumably taking place somewhere inside the Postgres "stack",
> as that description isn't found anywhere in the Slony code base.
>
> If we can get a stack trace from the Postgres logs indicating
> specifically where it was executing when this happened, it might be
> possible to toss in a more descriptive error message, presumably
> somewhere in the subscribe set code.

I ran into this same problem and I've captured some more output:

Sequence of queries (extracted from strace, so not complete):

begin transaction;
select count(*) FROM "config_rep".sl_subscribe where sub_set=1 AND
sub_receiver=2  and sub_...
select set_origin from "config_rep".sl_set where set_id=1
>>> result is SELECT 0, followed by
<stdin>:4: row number 0 is out of range 0..-1
>>> and then segmentation fault

Backtrace looks like this (not sure how much optimisation was used):

#0  0xaf5c2c29 in ____strtol_l_internal (nptr=0x0, endptr=0x0,
base=10, group=0, loc=0xaf6cd360) at strtol_l.c:298
#1  0xaf5c29d8 in strtol (nptr=0x0, endptr=0x0, base=10) at strtol.c:110
#2  0x1747b8b1 in atoi (stmt=0x1749bb68) at /usr/include/stdlib.h:286
#3  slonik_subscribe_set (stmt=0x1749bb68) at slonik.c:3506
#4  0x1747deca in script_exec_stmts (script=<value optimized out>,
hdr=0x1749bb68) at slonik.c:1416
#5  0x1747e281 in script_exec_stmts (script=<value optimized out>,
hdr=0x1749bdd0) at slonik.c:1127
#6  0x1747e372 in script_exec (script=0x1749be18) at slonik.c:1091
#7  0x1747f2a5 in main (argc=1, argv=0xb8f874f4) at slonik.c:156

Hope this helps,
-- 
Martijn van Oosterhout <kleptog at gmail.com> http://svana.org/kleptog/