Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 3 Apr 2014 11:02:38 -0400
From:      John Baldwin <jhb@freebsd.org>
To:        Dmitry Sivachenko <trtrmitya@gmail.com>
Cc:        freebsd-hackers@freebsd.org, Trond =?utf-8?q?Endrest=C3=B8l?= <Trond.Endrestol@fagskolen.gjovik.no>
Subject:   Re: madvise() vs posix_fadvise()
Message-ID:  <201404031102.38598.jhb@freebsd.org>
In-Reply-To: <0AF273E6-CD43-417C-A00C-5B7445090D5B@gmail.com>
References:  <D6BD48AF-9522-495D-8D54-37854E53C272@gmail.com> <201403271141.41487.jhb@freebsd.org> <0AF273E6-CD43-417C-A00C-5B7445090D5B@gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thursday, April 03, 2014 7:29:03 am Dmitry Sivachenko wrote:
>=20
> On 27 =D0=BC=D0=B0=D1=80=D1=82=D0=B0 2014 =D0=B3., at 19:41, John Baldwin=
 <jhb@FreeBSD.org> wrote:
> >>=20
> >> I know about mlock(2), it is a bit overkill.
> >> Can someone please explain the difference between madvise(MADV_WILLNEE=
D) and=20
> > posix_fadvise(POSIX_FADV_WILLNEED)?
> >=20
> > Right now FADV_WILLNEED is a nop.  (I have some patches to implement it=
 for
> > UFS.)  I can't recall off the top of my head if MADV_WILLNEED is also a=
 nop.
> > However, if both are fully implemented they should be similar in terms =
of
> > requesting async read-ahead.  MADV_WILLNEED might also conceivably
> > pre-create PTEs while FADV_WILLNEED can be used on a file that isn't
> > mapped but is accessed via read(2).
> >=20
>=20
>=20
> Hello and thanks for your reply.
>=20
> Right now I am facing the following problem (stable/10):
> There is a (home-grown) webserver which mmap's a large amount of data fil=
es (total size is a bit below of RAM, say ~90GB of files with 128GB of RAM).
> Server writes access.log (several gigabytes per day).
>=20
> Some of mmaped data files are used frequently, some are used rarely. On s=
tartup, server walks through all of these data files so it's content is rea=
d=20
from disk.
>=20
> After some time of running, I see that rarely used data files are purged =
from RAM (access to them leads to long-running disk reads) in favour of dis=
k=20
cache
> (at 0:00, when I rotate and gzip log file I see Inactive memory goes down=
 to the value of log file size).
>=20
> Is there any way to tell VM system not to push mmap'ed regions out of RAM=
 in favour of disk caches?

Use POSIX_FADV_NOREUSE with fadvise() for the log files.  They are a perfect
use case for this flag.  This will tell the VM system to throw the log data
(move it to cache) after it writes the file.

=2D-=20
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201404031102.38598.jhb>