Hi,
I run into "this" frequently on my Win 7 (64-bit) laptop. I had a sense the
problem might
be related to changing networks and hibernation, but shamefully I haven't made
any
real effort to root cause it.
In my random attempts to deal with it, I found that reboots (usually more than
one) hel
I don't think that yet exists. You could distinguish Cephfs from RBD. The
former would not (currently) seem worth attempting. Perhaps RBD could form a
backing store for partitions. I'm not convinced, but I'd be interested in any
data points.
Matt
- "Steven Presser" wrote:
> If any on
Congratulations, this is wonderful news. I've been following this on the
Cygwin list.
Thanks for the effort.
Matt
- "Jeffrey Altman" wrote:
> As of Cygwin 1.7.19, there is support for OpenAFS 1.7 mount points
> and
> symlinks.
>
> Jeffrey Altman
>
--
Matt Benjamin
The Linux Box
206 S
Hi,
I don't think that's what Troy meant.
At any rate, he -might- have meant he presumed there would be no interest in
standardizing rxk5 unless it turned out to be something that a significant
number of real sites wanted to use.
Matt
- "Gary Buhrmaster" wrote:
> On Tue, Oct 30, 2012 at
Hi,
Just for clarification, and I don't know who to ask at this point, this mail
discusses "tcpoob." What happened to rxtcp?
Matt
- "Andrew Deason" wrote:
> On Wed, 17 Oct 2012 10:45:05 +0200
> Peter van der Reest wrote:
>
> > In fact, today already one question came up:
> > what is th
Hi,
Obviously, Marcus and I thought having such a mechanism was a good idea. When
we started work, the idea of "standardizing" the protocol hadn't been
formalized.
The objections early on amounted somewhat, I feel, to "the great is the enemy
of the good." It has been claimed that rxk5 is "un
Hi,
Well, we used the fuse rather extensively for locking and dirformat testing.
It's experimental, but science experiment might be a little strong.
Matt
- "Russ Allbery" wrote:
> Troy Benjegerdes writes:
>
> > I'm looking to get all the low-hanging fruit with unskilled
> testing.
> >
Hi,
Marcus and I consider rxk5 to be a closed effort. We did make rxk5 partly as a
challenge to the OpenAFS participants to adopt a candidate solution quickly,
since the transport security problem was (and is) pressing. It should have
been seriously considered for integration in 2006-7, since
Tue Sep 20 15:06:07 2011 Assertion failed! file
> /build/buildd-openafs_1.6.0-1-amd64-YBR2T1/openafs-1.6.0/src/afs/UKERNEL/afs_usrops.c,
> line 1284.
>
>
>
> On Tue, Sep 20, 2011 at 02:58:33PM -0400, Matt W. Benjamin wrote:
> > Hi Troy,
> >
> > When I just rece
Hi Troy,
When I just recently used fuse UKERNEL I did the following:
afsd.fuse -memcache -mountdir /fafs -cachedir /vcache -d
and got initial results. I haven't worked much with it yet, but the cm started
and responded to vfsops.
Matt
- "Troy Benjegerdes" wrote:
> Something is broken
he fileserver configuration is the default from the repository:
> /usr/lib/openafs/fileserver -p 23 -busyat 600 -rxpck 400 -s 1200 -l
> 1200 -cb 65535 -b 240 -vc 1200
>
>
>
> On Wed, Aug 31, 2011 at 6:01 AM, Matt W. Benjamin < m...@linuxbox.com
> > wrote:
>
&g
Hi,
I think some more detail on the network/network path would be helpful.
Regards,
Matt
- "Ken Elkabany" wrote:
> Hi,
>
>
> On a network capable of scp-ing files between machines at 60MB/sec, we
> are only able to achieve 2-3MB/sec of throughput when using AFS. We've
> been conducting
Hi,
If you're primarily interested in direct access, you might want to look at
rxosd's vicep-access, which makes AFS volumes directly available to the client
over a cluster file system--Lustre or GPFS, but perhaps you could make
glusterfs go easily, I'm not sure. This actually has demonstrated
Actually, not exactly. And Russ already sent a template with some good
defaults. If you want more than that number of callbacks, increase the value
for -cb.
Matt
- "Matt W. Benjamin" wrote:
> (Yes, exactly.) IIRC, by default it's around 64k, but was going to be
>
(Yes, exactly.) IIRC, by default it's around 64k, but was going to be
increased?
Matt
- "Jaap Winius" wrote:
> Quoting "Matt W. Benjamin" :
>
> > ... If you're not already setting a million or so, for rw file
> > service, you might consid
One way you can take deterministic advantage of more memory in today's
fileservers (caveat, if you have the workload) is to increase the number of
callbacks supported. If you're not already setting a million or so, for rw
file service, you might consider increasing that parameter.
Matt
-
Hi,
It would be helpful to better understand the original design intention here.
Not knowing the semantic intent of whatever remote update has taken place, the
cm cannot reliably decide what is a compatible update. Assuming that no remote
change is compatible with any local change contradicts
Hi,
It's useful to document these limitations, but of course, they don't
necessarily rule out rxosd mirroring having utility for some application.
Also, I had the impression rxosd was undergoing development.
Matt
- "Simon Wilkinson" wrote:
> On 11 Apr 2011, at
Hi,
The variant of OpenAFS called rxosd has write mirroring, too, I believe
(currently, this is a fork, has quite a number of differences to OpenAFS).
However, my intuition is that some other system would work better for this use
case. Just speculatively,
1. you might find the mirroring in g
Hi Thomas,
Mandatory enforcement is the defined behavior for byte-range locks on Windows.
Since you are testing on Windows, you are seeing the Windows behavior. The
same lock would be interpreted as advisory on a Unix client.
Matt
- "Thomas Smith" wrote:
> Hi,
>
> Is it possible to op
Or, I would have thought, copied the module to kernel/fs, run depmod, and used
modprobe to load libafs?
Matt
- "Andrew Deason" wrote:
> On Fri, 7 Jan 2011 11:01:08 -0800
> Ted Creedon wrote:
>
> > libafs: Unknown symbol find_exported_dentry
> > libafs: Unknown symbol export_op_default
>
FYI, I'm -actually- and AFS backup lamer and archive volume dumps in a Bacula
repo. Sorry.
Matt
- "Emil Assarsson" wrote:
> Thanks all for the suggestions :-)
>
> I found this on internet
> http://linuxbox.com/drupal/files/downloads/afs_backup_afsbp2005.pdf
> that seems to bring up some
Hi,
I may be misremembering here, but my recollection of the current Amanda and
Bacula solutions is that they are somewhat primitive, e.g., compared with TiBs
or TSM integrations. (I think Russ has talked a not infrequently about further
work on Bacula integration would be desirable.)
Someone
Hi,
Is "write-on-close" was an expectation which could be broken? It is the case
that AFS has strongly expressed that its semantics are (in general)
_sync-on-close_, and it's meaning is that an application which has closed an
AFS file may consider any writes it made to be stable, and visible t
Hi,
Cant switch off is broken, sure. Cache nothing, however, is reasonable, for
some application. I think a better way to specify "one" of { disabled;
1..ULONG_MAX; always} would be desirable.
The expectation is that there might be a need for different selectors or
thresholds, but there has
Hi Phil,
This fs operation gets, or sets, a threshold (relative to file size) for cache
bypass. You need to be using a CM port with bypass support, i.e., Linux right
now. By default, all reads and writes are through the CM's cache. There is no
threshold unless set, so by default cache bypass
Sorry, I wasn't expressing an opinion...
Matt
- "Derrick Brashear" wrote:
> On Sat, Jun 19, 2010 at 9:39 PM, Matt W. Benjamin
> wrote:
> > wow. do we want that? replacement for rt?
> >
> > Matt
> >
> > - "Adam Megacz" wr
wow. do we want that? replacement for rt?
Matt
- "Adam Megacz" wrote:
> FYI
>
> http://code.google.com/p/gerrit/issues/detail?id=124
>
> ___
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/
If the CM omits to flush in this case, that's a plausible result. Viced will
BCB when the lock is released, but if the changed data hasn't been written
back, the cooperating clients lose.
Matt
- "Simon Wilkinson" wrote:
>
> I strongly suspect that the copy on the client is up to date, at
Hi,
Simon is correct. A byte-range locking implementation for OpenAFS is being
funded by Your File System, Inc., under its DOE SBIR Phase II grant. As stated
elsewhere by Jeff, there are (or will be) structures for making completed
available to the community during the course of the work.
Ho
well done
- "Andrew Deason" wrote:
> On Tue, 12 Jan 2010 23:19:09 +
> Simon Wilkinson wrote:
>
> >
> > On 12 Jan 2010, at 23:12, Gary Gatling wrote:
> >
> > > Thanks a lot for all the replys. Its sounds like we don't need to
> > > be restarting afs every week since we are using a mod
Hi,
Interested folks are welcome to try rxk5 in its current form. Feel free to
email me or Marcus if this would be helpful to you.
Regards,
Matt
- "Russ Allbery" wrote:
> Jack Neely writes:
>
> > I'm researching upgrading our aging (1.2) Kerberos 5 infrastructure
> to
> > 1.6. I'm at
Hi,
It appears this script will tend to keep callbacks registered on fid and the
objects in fid, and discourage their eviction from cache, but doesn't force the
CM to read in segments of the objects. It would be trivial to make it do. As
Marc's comment implies though, without discon, I don't
Hi,
I do think that we need to avoid assuming lack of clamor for some feature means
folks don't think it's important. I haven't seen benchmarks on pthreaded
ubik--but from intuition, I intuit we would, at any rate.
Matt
- "Steven Jenkins" wrote:
> On Thu, Dec 17, 2009 at 2:29 AM, Alista
Hi Simon,
I would generally agree with almost all of this, except, probably removing
--enable-fast-restart? Since, as you note, there are sites which actually do
rely on this, are you in fact meaning to say that the option would exist, but
not be mentioned in the usage? I guess I'd also like
Hi,
I don't think Marcus' remark was meant to be disingenuous, but, in any case.
Simon and I discussed some of the issues yesterday via XMPP. We agree there
are lots of things Marcus and I can do, but a lot of things become easier and
more useful to do when upstream is ready start thinking ab
Hi Folks,
As subscribers to the afs3-standardization list know already, a hypothetical
last draft of the Extended Callback Information proposal has been published,
including changes requested at the Edinburgh AFS hackathon which just took
place.
The process we've agreed to follow resembles the
Hi Tom,
I recall mentioning in Ohio, XCB doesn't have a private view of client callback
state. I'm unsure it's a good tradeoff to serialize XCB messages for "reliable
delivery." You would be changing the goal from "consistency is preserved" to
"clients don't need to update their cache" on fai
Hi,
Thank you, elbarto. (You all might want to get on openafs-devel, and/or
freebsd-afs mailing lists, as well).
Regards,
Matt
--
Emmanuel Vadot
elba...@arcadebsd.org
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.or
Hi,
The client supports 7.0, on i386 and amd64 (and a number of prior releases).
There is at least one significant open issue at 7.x and 8-CURRENT, that
hopefully will be resolved fairly soon.
Matt
- Original Message -
From: "Jerry McAllister"
To: "Derrick Brashear"
Cc: "ENEM|Hans M
Even a more self tuning OpenAFS is going to need to be...tunable to better suit
specific environments. It's the nature of complex systems.
- Original Message -
The long term direction is absolutely to make things more self-tuning.
Getting there is going to take a long time. In the mea
n
then, create another marked state where options are subject to move at a finite
date some time in the future. I mean, let's find some flexibility somewhere,
somehow...
Matt
- Original Message -
From: "Derrick Brashear"
To: "Matt W. Benjamin"
Cc: "OpenAF
moved to a config file in a couple of months. We
will all deal.
Regards,
Matt
----- Original Message -
From: "Matt W. Benjamin"
To: "Derrick Brashear"
Cc: "OpenAFS Devel" , "openafs"
, u+openafsdev-t...@chalmers.se
Sent: Saturday, May
- Original Message -
From: "Derrick Brashear"
To: u+openafsdev-t...@chalmers.se
Cc: "OpenAFS Devel" , "openafs"
Sent: Saturday, May 16, 2009 10:07:12 AM GMT -05:00 US/Canada Eastern
Subject: Re: [OpenAFS-devel] Selecting a configuration file format for OpenAFS
Services
On Sat, May 16
44 matches
Mail list logo