| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com> |
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Phantom Command ID |
| Date: | 2006-09-21 13:52:36 |
| Message-ID: | 29290.1158846756@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Heikki Linnakangas <heikki(at)enterprisedb(dot)com> writes:
> We could also limit the size of the hash table, which takes up most of
> the memory, and only keep the latest phantom cids there. Presumably, if
> current command id is 1000, you're not likely to set cmax to 500 on any
> tuple in that transaction anymore.
The downside of that though is that if you did generate any such, you'd
assign a fresh (duplicate) phantom cid --- so you're bloating the array
in exchange for reducing the hash size.
It is quite easy to have current command counter much greater than the
CID of a still-live command: consider for example an UPDATE that is
firing triggers as it goes, and each trigger executes some queries.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2006-09-21 14:02:44 | Re: 'configure --disable-shared' and 'make check' |
| Previous Message | Simon Riggs | 2006-09-21 12:22:08 | Re: [COMMITTERS] pgsql: Add documentation for new in-core advisory |