OID/XID allocation (was Re: is PG able to handle a >500

OID/XID allocation (was Re: is PG able to handle a >500

Post by Bruce Momji » Wed, 24 Jan 2001 09:27:22




> > What about pre-fetching of OID's.  Does that still happen for every
> > backend?

> Only ones that actually allocate some OIDs, I think.

> > What about XID's?

> XIDs are wasted on a postmaster restart, but not per-backend, because
> they are cached in shared memory instead of locally.  I've been thinking
> about changing the code so that OIDs are allocated in the same fashion.
> That would mean an extra spinlock grab per OID allocation, but so what?
> We grab several spinlocks per row creation already.  And we could
> increase the number of OIDs allocated per pg_variable file update,
> which would save some time.

Yes, I think the leak of 50 oids for every backend that asks for an OID
is really a waste.  I can see this change buying us a few more years
without oid problems.

--
  Bruce Momjian                        |  http://candle.pha.pa.us

  +  If your life is a hard drive,     |  830 Blythe Avenue
  +  Christ can be your backup.        |  Drexel Hill, Pennsylvania 19026

 
 
 

1. OID/XID allocation (was Re: is PG able to handle a >500 GB Da

Already implemented. But - up to 8192 oids may be lost in the event
of crash (ie without normal database shutdown when last fetched oid
is logged to WAL).

Backends don't grab block of oids anymore.

Too late to object -:))
I had to make this changes for WAL. BTW, pg_variable is not used anymore
but still in schema - have to remove it someday.

Vadim

2. Can't Install BOL Update

3. OID/XID allocation (was Re: is PG able to handle

4. 13814-WA-Seattle Downtown - V-UNIX-Windows NT-C-C++-ORACLE-Java-HTML-CTI-Developer

5. OID/XID allocation (was Re: is PG able to handle a >500 GB Da tabase?)

6. Test

7. OID/XID allocation (was Re: is PG able to handle a >500 GB Database?)

8. Call for objections: deprecate postmaster -o switch?

9. is PG able to handle a >500 GB Database?

10. curval was Re: is PG able to handle a >500 GB Database?

11. is PG able to handle a >500 GB Database?