ed the index build script such that for each table it builds one
index by itself, then builds the remaining indexes in parallel. This
appears to be stable. I made several runs with no errors. I've got some
more testing to do, then I'll try my big run.
Thanks
Wes
problem, too? In
this specific case, I guess I could do 1,2 and 1,3 parallel builds (for the
3 index/4 index tables).
Right now I'm running with rather restricted hardware (1GB memory, two
2-disk RAID 0's and a single disk). If the pilot proves what I think is
possible, and I can get rea
ding two indexes simultaneously, what would happen if I
tried to lock pg_class in the shorter index build transaction? Besides
seeming like a bad idea...
Wes
---(end of broadcast)---
TIP 5: don't forget to increase your free space map settings
On 4/25/06 9:58 AM, "Tom Lane" <[EMAIL PROTECTED]> wrote:
> I can't think of any effect in the PG code that would make CREATE INDEX
> commands tend to converge rather than diverge, so maybe there is some
> external effect here. Wes, is your machine a dual processo
ing
> that exhibits the behavior you see.
My initial attempts last night at duplicating it with a small result set
were not successful. I'll see what I can do.
On 4/25/06 3:25 AM, "Zeugswetter Andreas DCP SD" <[EMAIL PROTECTED]>
wrote:
> Wes, you could most
d it back by Friday.
> I kinda doubt it's *that* platform specific though.
You never know... CPU speed, hard drive speed (slow), etc.
Wes
---(end of broadcast)---
TIP 3: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faq
I usually get at least one occurrence.
Where do you want me to send it to?
Of course, since your hardware is different, it may not show up since it
appears to be a timing thing.. I'm on a PB G4 1Ghz.
Wes
---(end of broadcast)---
TIP 6: explain analyze is your friend
7; that there are no other open connections when the
index builds are running. I tried somewhat bigger test tables (a few rows
to a little over a hundred thousand). I can duplicate the error at will.
Wes
---(end of broadcast)---
TIP 3: Have you
fork
create index1
create index2
exit
fork
create index3
create index4
exit
wait for termination
exit
As my test tables are small and indexing happens almost instantly, I put a
'sleep' after one of the forks so that the two don't complete at the sa
