Re: NOLOGGING option, or ? - Mailing list pgsql-hackers

From Jochem van Dieten
Subject Re: NOLOGGING option, or ?
Date
Msg-id f96a9b8305060107074de65307@mail.gmail.com
Whole thread Raw
In response to Re: NOLOGGING option, or ?  (Greg Stark <gsstark@mit.edu>)
Responses Re: NOLOGGING option, or ?
List pgsql-hackers
On 01 Jun 2005 04:44:24 -0400, Greg Stark wrote:
> Greg Stark writes:
>>
>> For CREATE TABLE AS in the non-PITR case you don't really need to WAL log the
>> records at all. If it fails in the middle you just drop the table. When it
>> completes you do a checkpoint before acknowledging the COMMIT.
>>
>> I think this is already done for CREATE INDEX/REINDEX, also only in the
>> non-PITR case.

Checkpoint or fsync?


> Sorry to followup to my own message, but it occurs to me that COPY could be
> made to automatically do this for the case of an empty destination table too.

Why only on an empty table? What is the problem with bypassing WAL on
any table as long as all files of that table are fsync'ed before
commit?


> Again this sadly only works in the non-PITR case.

Apart from that problem of course :)

Jochem


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Cost of XLogInsert CRC calculations
Next
From: Tom Lane
Date:
Subject: Re: Cost of XLogInsert CRC calculations