Horton

Re: Handing off SLRU fsyncs to the checkpointer

Thomas Munro <thomas.munro@gmail.com>

View thread View on postgr.es
From: Thomas Munro <thomas.munro@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Jakub Wartak <Jakub.Wartak@tomtom.com>, "alvherre@2ndquadrant.com" <alvherre@2ndquadrant.com>, Robert Haas <robertmhaas@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2020-09-25T07:09:36Z
Lists: pgsql-hackers 
On Fri, Sep 25, 2020 at 12:53 PM Thomas Munro <thomas.munro@gmail.com> wrote:
> Here's a new version.  The final thing I'm contemplating before
> pushing this is whether there may be hidden magical dependencies in
> the order of operations in CheckPointGuts(), which I've changed
> around.  Andres, any comments?

I nagged Andres off-list and he opined that it might be better to
reorder it a bit so that ProcessSyncRequests() comes after almost
everything else, so that if we ever teach more things to offload their
fsync work it'll be in the right order.  I reordered it like that; now
only CheckPointTwoPhase() comes later, based on the comment that
accompanies it.  In any case, we can always reconsider the ordering of
this function in later commits as required.  Pushed like that.

Commits

  1. Remove unused function prototypes.

  2. Defer flushing of SLRU files.

  3. Improve the vacuum error context phase information.

  4. Cache smgrnblocks() results in recovery.

  5. Refactor the fsync queue for wider use.

  6. Increase maximum number of clog buffers.

  7. Make the number of CLOG buffers adaptive, based on shared_buffers.

  8. Replace implementation of pg_log as a relation accessed through the