| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
| Cc: | jagan <jaganrvce(at)yahoo(dot)com>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: WAL, xl_heap_insert and tuple oid mystry |
| Date: | 2011-04-13 00:43:43 |
| Message-ID: | 20423.1302655423@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> writes:
> On 12.04.2011 19:42, jagan wrote:
>> In general, why is OID of a tuple relegated as just another tuple
>> data, when it can replace xl_heaptid as a much more stable tuple
>> identifier.
> Possibly. On the other hand, another common complaint is that PostgreSQL
> writes a lot of WAL, much more than other DBMSs. Adding more
> higher-level information to WAL records would make that issue even
> worse.
Actually, the real problem with that proposal is that it assumes that
the OID (a) exists and (b) is unique, neither of which is guaranteed.
Especially not when you consider that WAL log entries have to correctly
distinguish which MVCC version of a tuple is being modified.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Shigeru Hanada | 2011-04-13 02:23:00 | Re: WAL, xl_heap_insert and tuple oid mystry |
| Previous Message | Tom Lane | 2011-04-12 23:34:45 | Re: WAL, xl_heap_insert and tuple oid mystry |