Обсуждение: Problem with 7.1.3 and template1.bki during first init

Поиск
Список
Период
Сортировка

Problem with 7.1.3 and template1.bki during first init

От
Ulrich Neumann
Дата:
Hi,

i have a problem that when I run the initdb script that after a while the initialisation stops.

I�ve enabled some debug output, I hope that there is someone who can help in this issue:
...
DEBUG:  heap_getnext returning tuple
DEBUG:  heap_getnext([pg_type,nkeys=0],backw=0) called
DEBUG:  heapgettup(pg_type, tid=0xca9c6e16[65536,26], dir=1, ...)
DEBUG:  heapgettup(..., b=0xca9c6e42, nkeys=0, key=0x0
DEBUG:  heapgettup: relation(u)=`pg_type', 00000001
DEBUG:  heap_getnext returning tuple
DEBUG:  heap_getnext([pg_type,nkeys=0],backw=0) called
DEBUG:  heapgettup(pg_type, tid=0xca9c6e16[65536,27], dir=1, ...)
DEBUG:  heapgettup(..., b=0xca9c6e42, nkeys=0, key=0x0
DEBUG:  heapgettup: relation(u)=`pg_type', 00000001
DEBUG:  heap_getnext returning EOS
DEBUG:  heap_getnext([pg_proc,nkeys=1],backw=0) called
DEBUG:  heapgettup(pg_proc, tid=0x0, dir=1, ...)
DEBUG:  heapgettup(..., b=0xca9c6f22, nkeys=1, key=0xca9c6f64
DEBUG:  heapgettup: relation(u)=`pg_proc', 00000001
DEBUG:  heap_getnext returning EOS
ERROR:  fmgr_info: function 117833728: cache lookup failed

The line in the template1.bki file that causes this problem is the last one of this snippet:

..
create bootstrap pg_attribute(attrelid = oid ,attname = name ,atttypid = oid ,attdispersion = float4 ,attlen = int2
,attnum= int2 ,attnelems = int4 ,attcacheoff = int4 ,atttypmod = int4 ,attbyval = bool ,attstorage = char ,attisset =
bool,attalign = char ,attnotnull = bool ,atthasdef = bool)
 
insert OID = 0 ( 1247 typname 19 0 32 1 0 -1 -1 f p f i f f)

thank you

Ulrich Neumann



Re: Problem with 7.1.3 and template1.bki during first init

От
Tom Lane
Дата:
Ulrich Neumann<u_neumann@gne.de> writes:
> ERROR:  fmgr_info: function 117833728: cache lookup failed

This is evidently an attempt to look up a garbage function OID (none
of the real function OIDs that might be encountered during initdb are
more than a couple thousand).

For this to happen during initdb implies that there's something pretty
broken about the server executable you're using.  What platform are you
on, what compiler are you using, what configure options did you select,
etc etc?
        regards, tom lane