From: | Stephen Frost <sfrost(at)snowman(dot)net> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Robert Haas <robertmhaas(at)gmail(dot)com>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Jesper Pedersen <jesper(dot)pedersen(at)redhat(dot)com>, Jeff Janes <jeff(dot)janes(at)gmail(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Write Ahead Logging for Hash Indexes |
Date: | 2017-03-15 15:06:41 |
Message-ID: | 20170315150640.GU9812@tamriel.snowman.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Tom,
* Tom Lane (tgl(at)sss(dot)pgh(dot)pa(dot)us) wrote:
> Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> > That theory seems inconsistent with how mdextend() works. My
> > understanding is that we zero-fill the new blocks before populating
> > them with actual data precisely to avoid running out of disk space due
> > to deferred allocation at the OS level. If we don't care about
> > failures due to deferred allocation at the OS level, we can rip that
> > logic out and improve the performance of relation extension
> > considerably.
>
> See my reply to Stephen. The fact that this fails to guarantee no
> ENOSPC on COW filesystems doesn't mean that it's not worth doing on
> other filesystems. We're reducing the risk, not eliminating it,
> but reducing risk is still a worthwhile activity.
Considering how much work we end up doing to extend a relation and how
we know that's been a hotspot, I'm not entirely sure I agree that
avoiding the relativly infrequent out-of-disk-space concern when
extending the relation (instead of letting it happen when we go to
actually write data into the page) really is a good trade-off to make.
Thanks!
Stephen
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2017-03-15 15:11:23 | Re: scram and \password |
Previous Message | Rafia Sabih | 2017-03-15 15:05:30 | Re: Enabling parallelism for queries coming from SQL or other PL functions |