Re: Write Ahead Logging for Hash Indexes

On Thu, Mar 9, 2017 at 12:25 AM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Wed, Mar 8, 2017 at 7:45 AM, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>> Okay, I can try, but note that currently there is no test related to
>> "snapshot too old" for any other indexes.
>
> Wow, that's surprising. It seems the snapshot_too_old test only
> checks that this works for a table that has no indexes. Have you,
> anyway, tested it manually?
>

Yes, I have tested in manually. I think we need to ensure that the
modified tuple falls on the same page as old tuple to make the test
work. The slight difficulty with the index is to ensure the modified
tuple to be inserted into same page as old tuple, this is more true
with hash indexes. Also, for heap, I think it relies on hot pruning
stuff and for index we need to perform manual vacuum. Basically, if
we want we can write a test for index, but not sure if it is worth the
pain to write for hash index when the test for btree is not there.

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com