Re: Page Scan Mode in Hash Index

On Wed, Sep 20, 2017 at 7:19 AM, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>> Page-at-a-time index vacuum as in _hash_vacuum_one_page doesn't matter
>> because such an operation doesn't allow TIDs to be reused.
>
> Page-at-a-time index vacuum also allows TIDs to be reused but this is
> done only for already marked dead items whereas vacuum can make the
> non-dead entries to be removed. We don't have a problem with
> page-at-a-time vacuum as it won't remove any items which the scan is
> going to mark as dead.

I don't think page-at-a-time index vacuum allows heap TIDs to be
reused. To reuse a heap TID, we have to know that there are no index
entries pointing to it. There's no way for the heap to know that a
page-at-a-time index vacuum has even happened, let alone which TIDs
were affected and that all other indexes have also removed all index
entries for those TIDs.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company