Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 1 Oct 2011 22:27:18 -0700
From:      Xin LI <delphij@gmail.com>
To:        Jeremy Chadwick <freebsd@jdc.parodius.com>
Cc:        freebsd-fs@freebsd.org, Adrian Chadd <adrian@freebsd.org>, delphij@freebsd.org
Subject:   Re: is TMPFS still highly experimental?
Message-ID:  <CAGMYy3sCCxOiVqeP4PVbvxnpcyoyQZoL%2Bw3nY8STYnpUNfj6JQ@mail.gmail.com>
In-Reply-To: <20111002020231.GA70864@icarus.home.lan>
References:  <CAOfDtXMm9K_fbOmvG2gvXxDfKakkgpPt9MLifqDxa4wCibMExg@mail.gmail.com> <alpine.GSO.1.10.1110011122030.882@multics.mit.edu> <CADLo83-s_3H8PbbxOPPxbe0m10U0U5JW-feB294dFs%2BQ3iTWvg@mail.gmail.com> <CAGMYy3ssi%2BkAuufDTHA1z6u7jRrZwRRkCCkcO94GHNGF9Rku_w@mail.gmail.com> <20111002020231.GA70864@icarus.home.lan>

next in thread | previous in thread | raw e-mail | index | archive | help

On Sat, Oct 1, 2011 at 7:02 PM, Jeremy Chadwick
<freebsd@jdc.parodius.com> wrote:
> On Sat, Oct 01, 2011 at 06:02:39PM -0700, Xin LI wrote:
>> On Sat, Oct 1, 2011 at 8:48 AM, Chris Rees <crees@freebsd.org> wrote:
>> > I've also not heard of anyone using it with zfs successfully- it tends to
>> > shrink rapidly.
>>
>> I'm quite surprised with this assertion.  I use tmpfs on my own system
>> and I never see such problem as long as one have sufficient swap
>> space.
>>
>> Not to say there is no problem --there is no way to say "commit this
>> amount of memory to ZFS" but really I have never hit this exact
>> alleged problem...
>
> Its been reported multiple times by multiple people, and there has been
> no official word on -fs, -stable, or zfs-devel that this specific
> problem has been fixed:
>
> * miyamoto moesasji
>  - 2011/01/01
>  - 8.2-PRERELEASE (thus RELENG_8)
>  - World/kernel build date unknown
>  - http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060850.html
>  - http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060852.html
>  - http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060860.html
>  - http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060861.html
>  - Statement from Ivan Voras that it's a known problem:
>  - http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060867.html
>
> * Atilla Nagy
>  - 2011/01/19
>  - 8.2-PRERELEASE (thus RELENG_8)
>  - World/kernel build date of 2011/01/08
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010496.html
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010497.html
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010498.html
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010499.html
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010501.html
>  - http://lists.freebsd.org/pipermail/freebsd-fs/2011-January/010569.html
>  - Ivan Voras mentions a thread he started on -CURRENT circa 2010/11/21
>    about this problem:
>    http://www.mail-archive.com/freebsd-current@freebsd.org/msg126491.html
>
> * Michael Loftis
>  - 2009/12/08
>  - FreeBSD 8.0-RELEASE (RELENG_8_0)
>  - World/kernel build date of 2009/11/21
>  - Might be a different problem altogether, but tmpfs is explicitly
>    mentioned in the fix/commit text from avg@
>  - http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/141305
>
> Plus Chris Rees' report.  So that makes 3, possibly 4 if you consider
> the PR from Michael Loftis.
>
> Possibly it has something to do with the exact FreeBSD version they're
> using.  Not just version (e.g. 8.2-RELEASE vs. 8.2-STABLE), but also
> kernel/world build date -- because as we've established, ZFS changes are
> happening all the time and there is very little transparency when it
> comes to communicating these changes to the community.
>
> Maybe the problem has been fixed in RELENG_8 sometime after 2011/01/08,
> but we do not know.  ZFS committers would need to help.  Else, folks
> should be reaching out to the first two people I mentioned above and
> asking if they can reproduce the problem on present-day RELENG_8.

The problem Ivan have asserted was not confirmed by anyone who have
swap configured properly.  Gleb have pointed out that it might be
related to a series of integer overflow by the way (he have also fixed
a lot of tmpfs issues by the way).

Cheers,
-- 
Xin LI <delphij@delphij.net> https://www.delphij.net/
FreeBSD - The Power to Serve! Live free or die



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGMYy3sCCxOiVqeP4PVbvxnpcyoyQZoL%2Bw3nY8STYnpUNfj6JQ>