Обсуждение: [HACKERS] extending postgres buffer/query limits

Поиск
Список
Период
Сортировка

[HACKERS] extending postgres buffer/query limits

От
dj@pelf.harvard.edu (Diab Jerius)
Дата:
I have a very wide table, too wide to normally be created (due to
buffer size constraints), which was created via inheritance of two classes.

I'm trying to move the table from 6.0 to 6.1, but pg_dump gives a
Bus Error (under sparc Solaris 2.5.1) after partially printing out the
header.  I presume this is because the line to be printed was too long
and overflowed a buffer.

[NOTE to author of pg_dumpall:  it should check the return of
    pg_dump to ensure that there was no error.]

Can someone point me to where I should start looking in order to relieve
the buffer limitations in postgres?  I can't make the table any narrower.

Thanks,

Diab

- -------------
Diab Jerius                       Harvard-Smithsonian Center for Astrophysics
                                  60 Garden St, MS 70, Cambridge MA 02138 USA
djerius@cfa.harvard.edu           vox: 617 496 7575         fax: 617 495 7356

------------------------------

Re: [HACKERS] extending postgres buffer/query limits

От
Bruce Momjian
Дата:
>
>
> I have a very wide table, too wide to normally be created (due to
> buffer size constraints), which was created via inheritance of two classes.
>
> I'm trying to move the table from 6.0 to 6.1, but pg_dump gives a
> Bus Error (under sparc Solaris 2.5.1) after partially printing out the
> header.  I presume this is because the line to be printed was too long
> and overflowed a buffer.
>
> [NOTE to author of pg_dumpall:  it should check the return of
>     pg_dump to ensure that there was no error.]

Done.

>
> Can someone point me to where I should start looking in order to relieve
> the buffer limitations in postgres?  I can't make the table any narrower.

I find this hard to believe.

- --
Bruce Momjian
maillist@candle.pha.pa.us

------------------------------

Re: [HACKERS] extending postgres buffer/query limits

От
dj@pelf.harvard.edu (Diab Jerius)
Дата:
- ---------  Received message begins Here  ---------

> From: Bruce Momjian <maillist@candle.pha.pa.us>
> Message-Id: <199706201630.MAA03762@candle.pha.pa.us>
> Subject: Re: [HACKERS] extending postgres buffer/query limits
> To: djerius@cfa.harvard.edu
> Date: Fri, 20 Jun 1997 12:30:19 -0400 (EDT)

[deletia]
> >
> > Can someone point me to where I should start looking in order to relieve
> > the buffer limitations in postgres?  I can't make the table any narrower.
>
> I find this hard to believe.
>

Well, I won't quibble over data base design here.  The table is completely
normalized, or whatever the correct terminology is.  There's but one key and
several hundred dependent fields (it's the output of a X-ray line fit,
which does indeed have that many parameters).  Obviously I can split the
table up, dump the bits, and rejoin them, but that's not the point.

The point is that it is possible to create tables which pg_dump can't handle.
This shouldn't happen.  This may be fixed in 6.1, but I can't test that until I
can move my data over, which now promises to be a real pain.


- -------------
Diab Jerius                       Harvard-Smithsonian Center for Astrophysics
                                  60 Garden St, MS 70, Cambridge MA 02138 USA
djerius@cfa.harvard.edu           vox: 617 496 7575         fax: 617 495 7356

------------------------------

End of hackers-digest V1 #393
*****************************