From: | "Hiroshi Inoue" <Inoue(at)tpf(dot)co(dot)jp> |
---|---|
To: | "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | "pgsql-hackers" <pgsql-hackers(at)postgreSQL(dot)org> |
Subject: | RE: [HACKERS] How to ignore system indexes |
Date: | 2000-01-18 05:48:19 |
Message-ID: | 001d01bf6177$9f6593a0$2801007e@tpf.co.jp |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
> -----Original Message-----
> From: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
>
> "Hiroshi Inoue" <Inoue(at)tpf(dot)co(dot)jp> writes:
> > There are fairly many places using system indexes.
> > Probably I would be able to change them.
> > But is it preferable or possible to force other developers
> > to take ignore_system_indexes mode into account ?
>
> Is it really necessary to touch all those places?
>
> Seems to me that if a person needs to rebuild system indexes,
> he would be firing up a standalone backend and running
> REINDEX --- and darn little else. As long as none of the
> support code required by REINDEX insists on using an index,
> it doesn't matter what the rest of the system requires.
>
OK,I would limit changes only for REINDEX command.
> You might even think about doing the reindex in bootstrap mode,
> though I don't know if that would be easier or harder.
>
Yes,bootstrap mode is a natural selection. Jan has already tried
it and there was a problem of time quliafication. I don't know it is
a big obstacle or not. I prefer standalone postgres because
there's a possibility to call various SQL commands together
with REINDEX command.
Of cource,time qualification is no longer a problem in standalone
postgres.
Regards.
Hiroshi Inoue
Inoue(at)tpf(dot)co(dot)jp
From | Date | Subject | |
---|---|---|---|
Next Message | Tatsuo Ishii | 2000-01-18 05:51:14 | pg_proc.h changed |
Previous Message | Hiroshi Inoue | 2000-01-18 05:48:01 | RE: [HACKERS] How to ignore system indexes |