On 08.10.2013 13:00, Pavan Deolasee wrote:
> On Tue, Oct 8, 2013 at 3:16 PM, Andres Freund<andres@2ndquadrant.com>wrote:
>
>> It is my impression that there still are several people having pretty
>> fundamental doubts about this approach in general. From what I remember
>> neither Heikki, Simon, Tom nor me were really convinced about this
>> approach.
>>
> IIRC you and Tom were particularly skeptical about the approach. But do you
> see a technical flaw or a show stopper with the approach ? Heikki has
> written pg_rewind which is really very cool. But it fails to handle the
> hint bit updates which are not WAL logged unless of course checksums are
> turned on. We can have a GUC controlled option to turn WAL logging on for
> hint bit updates and then use pg_rewind for the purpose. But I did not see
> any agreement on that either. Performance implication of WAL logging every
> hint bit update could be huge.
Yeah, I definitely think we should work on the pg_rewind approach
instead of this patch. It's a lot more flexible. The performance hit of
WAL-logging hint bit updates is the price you have to pay, but a lot of
people were OK with that to get page checksum, so I think a lot of
people would be OK with it for this purpose too. As long as it's
optional, of course. And anyone using page checksums are already paying
that price.
- Heikki
