| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Frans Hals <fhals7(at)googlemail(dot)com> |
| Cc: | pgsql-general(at)postgresql(dot)org |
| Subject: | Re: Large index operation crashes postgres |
| Date: | 2010-03-24 20:26:52 |
| Message-ID: | 28293.1269462412@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
Frans Hals <fhals7(at)googlemail(dot)com> writes:
> Whenever I try to add the follwing index to the table "placex", one of
> the postmaster processes dies and the server restarts.
Can you provide a stack trace from the crash?
> I try:
> CREATE INDEX idx_placex_sector ON placex USING btree
> (geometry_sector(geometry), rank_address, osm_type, osm_id);
> The table counts around 50.000.000 rows.
> The first 20.000.000 are indexed in 20-30 minutes. Nice!
> Then indexing becomes slow and slower, first taking 100.000 rows in
> ten minutes while further consequently decreasing speed.
> When the job reaches something around row 25.000.000 postgres goes down:
On what exactly do you base these statements about number of rows
processed? CREATE INDEX doesn't provide any such feedback that I'm
aware of.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Paul Ramsey | 2010-03-24 20:46:07 | Re: Large index operation crashes postgres |
| Previous Message | Frans Hals | 2010-03-24 20:15:56 | Large index operation crashes postgres |