Re: pageinspect and hash indexes

On Sat, Mar 18, 2017 at 1:42 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Ashutosh Sharma <ashu(dot)coek88(at)gmail(dot)com> writes:
>> Basically, when we started working on WAL for hash index, we found
>> that WAL routine 'XLogReadBufferExtended' does not expect a page to be
>> completely zero page else it returns Invalid Buffer. To fix this, we
>> started initializing freed overflow page or new bucket pages using
>> _hash_pageinit() which basically initialises page header portion but
>> not it's special area where page type information is present. That's
>> why you are seeing an ERROR saying 'page is not a hash page'. Actually
>> pageinspect module needs to handle this type of page. Currently it is
>> just handling zero pages but not an empty pages. I will submit a patch
>> for this.
>
> That seems like entirely the wrong approach. You should make the special
> space valid, instead, so that tools like pg_filedump can make sense of
> the page.
>

We were not aware that external tools like pg_filedump are dependent
on special space of index, but after looking at the code of
pg_filedump, I agree with you that we need to initialize the special
space in this case (free overflow page). I think we can mark such a
page type as LH_UNUSED_PAGE and then initialize the other fields of
special space. Nonetheless, I think we still need modifications in
hashfuncs.c so that it can understand this type of page.

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