On 11/05/06, Mark Linimon [EMAIL PROTECTED] wrote:
On Sun, May 07, 2006 at 08:16:42PM +0100, Chris wrote:
GNATS didnt reply so I will send the info about email address I used
and see if I can find the pr number as well.
Any problems you have with GNATS not responding, please email them to
David Kirchner wrote:
Here's how to reproduce the snapshot deadlock I'm seeing, with 6.1-RC2
cvsup'd as of 5 or 6 hours ago:
[...]
It locks up every time for me, with no further disk activity.
Unfortunately, for some reason, my server console became unaccessable,
so I'm not able to get to the
On Sun, May 07, 2006 at 08:16:42PM +0100, Chris wrote:
GNATS didnt reply so I will send the info about email address I used
and see if I can find the pr number as well.
Any problems you have with GNATS not responding, please email them to
[EMAIL PROTECTED] and we'll try to figure out what's
On Sat, May 06, 2006 at 09:26:51PM +0100, Chris wrote:
1 - it does increase stability if the extra time is spent fixing bugs
and testing the fixes.
This assumes that you can persuade committers to focus on bugfixes instead
of on other things that they consider more fun.
I happen to think
On Sun, 7 May 2006, Dmitry Morozovsky wrote:
DM KK Also, as an FYI, several quota-related snapshot fixes went into CVS
in the
DM KK last 24-48 hours. Once they've settled for a few weeks, and assuming
they
DM KK don't have complex dependencies, they will get merged to RELENG_6.
DM KK
On Sun, May 07, 2006 at 07:26:55PM +0400, Dmitry Morozovsky wrote:
On Sun, 7 May 2006, Dmitry Morozovsky wrote:
DM KK Also, as an FYI, several quota-related snapshot fixes went into CVS
in the
DM KK last 24-48 hours. Once they've settled for a few weeks, and
assuming they
DM KK
On Sun, 7 May 2006, Konstantin Belousov wrote:
KB Alas, it is locked again:
KB
KB [EMAIL PROTECTED]:~# ps axlww | grep snap
KB 032 0 0 -4 0 0 8 snaplk DL??0:04.55
[bufdaemon]
KB 035 0 0 -4 0 0 8 snaplk DL??0:01.11
On Sun, May 07, 2006 at 07:59:33PM +0400, Dmitry Morozovsky wrote:
On Sun, 7 May 2006, Konstantin Belousov wrote:
KB Alas, it is locked again:
KB
KB [EMAIL PROTECTED]:~# ps axlww | grep snap
KB 032 0 0 -4 0 0 8 snaplk DL??0:04.55
[bufdaemon]
KB 0
On Sun, 7 May 2006, Konstantin Belousov wrote:
KB I'm running RELENG_6 with hand-merged sys/ufs/ffs changes from 1 to 7 May.
KB It would be great to show the patchset. Also note that relevant changes are
KB not limited to ufs/ffs. For instance, rev. 1.671 of kern/vfs_subr.c is
KB also
On Sun, May 07, 2006 at 08:29:08PM +0400, Dmitry Morozovsky wrote:
On Sun, 7 May 2006, Konstantin Belousov wrote:
KB I'm running RELENG_6 with hand-merged sys/ufs/ffs changes from 1 to 7
May.
KB It would be great to show the patchset. Also note that relevant changes
are
KB not limited
On Sun, 7 May 2006, Konstantin Belousov wrote:
KB KB I'm running RELENG_6 with hand-merged sys/ufs/ffs changes from 1 to
7 May.
KB
KB KB It would be great to show the patchset. Also note that relevant
changes are
KB KB not limited to ufs/ffs. For instance, rev. 1.671 of kern/vfs_subr.c is
On 07/05/06, Scott Long [EMAIL PROTECTED] wrote:
Chris wrote:
On 05/05/06, Mark Linimon [EMAIL PROTECTED] wrote:
Make Jakubik wrote:
FreeBSD users now demand stability and performance, as opposed to an
influx of new bells and whistles just before the release [...] I fully
understand
Janet Sullivan wrote:
Mike Jakubik wrote:
arrive at the sacrifice of stability. I think FreeBSD should only be
released when known major bugs are worked out. A known broken release
to me and most new users is useless, lets not release simply for the
sake of numbering.
For me, and many
David Nugent wrote:
Janet Sullivan wrote:
No, I don't use quotas. I doubt the majority of FreeBSD users do.
I don't either, but the majority of FreeBSD users is a shifting
thing, and very hard to make generalisations about. Once quota was
about shell users, now it's more common for mail
On Mon, May 08, 2006 at 01:34:34AM -0400, Mike Jakubik wrote:
David Nugent wrote:
Janet Sullivan wrote:
No, I don't use quotas. I doubt the majority of FreeBSD users do.
I don't either, but the majority of FreeBSD users is a shifting
thing, and very hard to make generalisations about.
On 05/05/06, Mike Tancsa [EMAIL PROTECTED] wrote:
At 02:27 PM 05/05/2006, Mike Jakubik wrote:
This quota/nve problem sure stirred things up, i guess im partly to
blame, but anyway i think that it all boils down to is this; FreeBSD
users now demand stability and performance, as opposed to an
On Sat, 6 May 2006, Chris wrote:
are snapshots a default setting?
In case people want to avoid using snapshots, here are the uses I'm currently
aware of in the base system:
- Background file system check uses snapshots. This can be disabled by using
background_fsck=NO in /etc/rc.conf.
On 05/05/06, Mark Linimon [EMAIL PROTECTED] wrote:
Make Jakubik wrote:
FreeBSD users now demand stability and performance, as opposed to an
influx of new bells and whistles just before the release [...] I fully
understand that this is a volunteer project [...]
I'm sorry, but the former
On Sat, 2006-May-06 21:26:51 +0100, Chris wrote:
1 - it does increase stability if the extra time is spent fixing bugs
and testing the fixes.
It always winds up not working this way. People won't test the early
beta releases and when the final release candidates appear, people
suddenly insist
On Sat, May 06, 2006 at 08:54:34PM +0100, Robert Watson wrote:
On Sat, 6 May 2006, Chris wrote:
are snapshots a default setting?
In case people want to avoid using snapshots, here are the uses I'm
currently aware of in the base system:
- Background file system check uses snapshots.
Robert Watson wrote:
On Sat, 6 May 2006, Chris wrote:
are snapshots a default setting?
In case people want to avoid using snapshots, here are the uses I'm
currently aware of in the base system:
- Background file system check uses snapshots. This can be disabled by
using
Chris wrote:
On 05/05/06, Mark Linimon [EMAIL PROTECTED] wrote:
Make Jakubik wrote:
FreeBSD users now demand stability and performance, as opposed to an
influx of new bells and whistles just before the release [...] I fully
understand that this is a volunteer project [...]
I'm sorry, but
On Thu, May 04, 2006 at 01:54:58AM -0400, Mike Jakubik wrote:
+ Kris Kennaway wrote:
+ On Thu, May 04, 2006 at 12:41:57AM -0400, Mike Jakubik wrote:
+ Then why utilize a known non-functional technology?
+
+
+ Because again, the benefits have been judged by the decision-makers
+ and found to
On Thu, May 04, 2006 at 04:59:33PM -0700, David Kirchner wrote:
Here's how to reproduce the snapshot deadlock I'm seeing, with 6.1-RC2
cvsup'd as of 5 or 6 hours ago:
1) dd if=/dev/zero of=/usr/bigfile bs=1024 seek=209715200 count=0
2) mdconfig -a -t vnode -f /usr/bigfile
3) bsdlabel -w md0
On 5/5/06, Pawel Jakub Dawidek [EMAIL PROTECTED] wrote:
It isn't good to release a software with known, documented bugs, but its
better than shipping an untested software with god-one-knows unknown
bugs.
There's another reasonable option: the known buggy code could be
disabled by default, and
This quota/nve problem sure stirred things up, i guess im partly to
blame, but anyway i think that it all boils down to is this; FreeBSD
users now demand stability and performance, as opposed to an influx of
new bells and whistles just before the release. FreeBSD already has many
great
On Fri, 5 May 2006, Mike Jakubik wrote:
This quota/nve problem sure stirred things up, i guess im partly to blame,
but anyway i think that it all boils down to is this; FreeBSD users now
demand stability and performance, as opposed to an influx of new bells and
whistles just before the
Mike Jakubik wrote:
This quota/nve problem sure stirred things up, i guess im partly to
blame, but anyway i think that it all boils down to is this; FreeBSD
users now demand stability and performance, as opposed to an influx of
new bells and whistles just before the release. FreeBSD already
One detail of this has to do with version numbering.
The FreeBSD version number says a lot more about the
userland than it does about the kernel per se.
If we were to version the kernel arch, I think it would
look more like this:
'94 1.1.5.1 (Last Net/2) Version 0
Nov '94
At 02:27 PM 05/05/2006, Mike Jakubik wrote:
This quota/nve problem sure stirred things up, i guess im partly to
blame, but anyway i think that it all boils down to is this; FreeBSD
users now demand stability and performance, as opposed to an influx
of new bells and
I think you *are*
Scott Long wrote:
Please contact me privately with a list of issues on the 6.1 TODO list
that are presently affecting you, and I will personally resolve them for
you.
Scott, thanks for the very generous gesture, but i cant ask you
something like this. The problems that are affecting me are
Make Jakubik wrote:
FreeBSD users now demand stability and performance, as opposed to an
influx of new bells and whistles just before the release [...] I fully
understand that this is a volunteer project [...]
I'm sorry, but the former statement proves the latter false.
Let's try to do our
Mike Tancsa wrote:
At 02:27 PM 05/05/2006, Mike Jakubik wrote:
This quota/nve problem sure stirred things up, i guess im partly to
blame, but anyway i think that it all boils down to is this; FreeBSD
users now demand stability and performance, as opposed to an influx
of new bells and
I
Mark Linimon wrote:
Make Jakubik wrote:
FreeBSD users now demand stability and performance, as opposed to an
influx of new bells and whistles just before the release [...] I fully
understand that this is a volunteer project [...]
I'm sorry, but the former statement proves the latter
Guys,
I appreciate the attempts at rational explaination from camp A, and
I appreciate the flood of emotional outpouring from camp B. However:
Mike Jakubik, David Kirchner, and others: You are making a mountain out
of a molehill and exploiting the unprecendented openness of the release
On Fri, 5 May 2006, Paul Allen wrote:
One detail of this has to do with version numbering. The FreeBSD version
number says a lot more about the userland than it does about the kernel per
se.
If we were to version the kernel arch, I think it would look more like this:
'94 1.1.5.1 (Last
Mike Jakubik wrote:
arrive at the sacrifice of stability. I think FreeBSD should only be
released when known major bugs are worked out. A known broken release to
me and most new users is useless, lets not release simply for the sake
of numbering.
For me, and many other quiet users, FreeBSD
On Thu, May 04, 2006 at 01:54:58AM -0400, Mike Jakubik wrote:
Kris Kennaway wrote:
On Thu, May 04, 2006 at 12:41:57AM -0400, Mike Jakubik wrote:
Kris Kennaway wrote:
You missed the part where snapshots have caused deadlocks under
varying conditions since day 1. They have never
On Thu, May 04, 2006 at 10:21:59AM +0400, Dmitriy Kirhlarov wrote:
Hi!
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
However, one could argue that as quotas worked OK in releases prior to
6.0 (and perhaps earlier), that there is a longer-term regression. In
I'm
David Kirchner wrote:
This assumes that 6.1 absolutely must be released
If you don't think that 6.1 must be released then just ignore it and
wait for 6.2. If you like, you can even pretend that 6.1 never existed.
Graham
___
freebsd-stable@freebsd.org
On Wed, 3 May 2006, David Kirchner wrote:
However, one could argue that as quotas worked OK in releases prior to 6.0
(and perhaps earlier), that there is a longer-term regression. In fact, it
seems that enabling snapshots by default appears to have caused a
significant regression for quotas
On Wed, 3 May 2006, David Kirchner wrote:
I wouldn't dare suggest that all bugs must be fixed prior to a new release,
but I will suggest that releases could be delayed to fix critical bugs that,
under a completely stock and default installation: stop the filesystem from
functioning, report
The quantity of significant performance and stability improvements
in 6.1 with respect to 6.0 easily justifies cutting a release: while
it may not correct every problem, it fixes a very large number.
Robert N M Watson
Notes from a mostly silent freebsd data center admin:
I am looking
Here's how to reproduce the snapshot deadlock I'm seeing, with 6.1-RC2
cvsup'd as of 5 or 6 hours ago:
1) dd if=/dev/zero of=/usr/bigfile bs=1024 seek=209715200 count=0
2) mdconfig -a -t vnode -f /usr/bigfile
3) bsdlabel -w md0 auto
4) newfs -U md0a
5) fsck -v /dev/md0a # ^C this after a second
On Tue, 2 May 2006, Kris Kennaway wrote:
Ditto, same thing with the recent nve fixes. Why release known broken
code when there are tested patches available? Whats the worst that will
happen? It wont work? Thats already the case...
...
OK, I can't speak to that issue specifically.
Robert Watson wrote:
On Tue, 2 May 2006, Kris Kennaway wrote:
Ditto, same thing with the recent nve fixes. Why release known broken
code when there are tested patches available? Whats the worst that
will
happen? It wont work? Thats already the case...
...
OK, I can't speak to that
On 5/3/06, Robert Watson [EMAIL PROTECTED] wrote:
This means that
they will take a significant amount of time to fix, and that each fix is high
risk, as it is likely to reveal latent bugs. This means that each fix will
require a lot of testing -- months of testing, in fact. So the choice is
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
On 5/3/06, Robert Watson [EMAIL PROTECTED] wrote:
This means that
they will take a significant amount of time to fix, and that each fix is
high
risk, as it is likely to reveal latent bugs. This means that each fix will
require
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
Would there really be harm in putting off a release until these well-
acknowledged bugs are taken care of?
Yes. We tagged the ports tree on April 14th, and that's the last date
of any changes the CDs will ship with. That's a
Kris Kennaway wrote:
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
On 5/3/06, Robert Watson [EMAIL PROTECTED] wrote:
This means that
they will take a significant amount of time to fix, and that each fix is
high
risk, as it is likely to reveal latent bugs. This means that
On 5/3/06, Mark Linimon [EMAIL PROTECTED] wrote:
To summarize: at some point you do, indeed, have to ship something. You
have to choose a point where the least number of users will see regressions
vs. the most number of users will see improvements. (Not everyone uses
quotas.)
FWIW, the
On 5/3/06, Kris Kennaway [EMAIL PROTECTED] wrote:
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
However, one could argue that as quotas worked OK in releases prior to
6.0 (and perhaps earlier), that there is a longer-term regression.
There was a quota regression in 6.0. It
On Wed, May 03, 2006 at 08:05:12PM -0700, David Kirchner wrote:
On 5/3/06, Kris Kennaway [EMAIL PROTECTED] wrote:
On Wed, May 03, 2006 at 06:21:39PM -0700, David Kirchner wrote:
However, one could argue that as quotas worked OK in releases prior to
6.0 (and perhaps earlier), that there is a
On Wed, May 03, 2006 at 08:11:32PM -0700, David Kirchner wrote:
On 5/3/06, Mark Linimon [EMAIL PROTECTED] wrote:
To summarize: at some point you do, indeed, have to ship something. You
have to choose a point where the least number of users will see regressions
vs. the most number of users
Kris Kennaway wrote:
On Wed, May 03, 2006 at 08:11:32PM -0700, David Kirchner wrote:
On 5/3/06, Mark Linimon [EMAIL PROTECTED] wrote:
To summarize: at some point you do, indeed, have to ship something. You
have to choose a point where the least number of users will see regressions
vs.
On Thu, May 04, 2006 at 12:41:57AM -0400, Mike Jakubik wrote:
Kris Kennaway wrote:
On Wed, May 03, 2006 at 08:11:32PM -0700, David Kirchner wrote:
On 5/3/06, Mark Linimon [EMAIL PROTECTED] wrote:
To summarize: at some point you do, indeed, have to ship something. You
have to choose
Kris Kennaway wrote:
On Thu, May 04, 2006 at 12:41:57AM -0400, Mike Jakubik wrote:
Kris Kennaway wrote:
You missed the part where snapshots have caused deadlocks under
varying conditions since day 1. They have never worked 100% reliably,
and despite our best efforts that will remain
On Tue, May 02, 2006 at 09:18:53PM +0400, Dmitriy Kirhlarov wrote:
Hi, list
I install FreeBSD 6.1-RC1 (from 11 Apr). I use gstripe for /home
volume. Also, I use quota on this volume.
After 6-8 hours without activities I can't read any data from
filesystem -- ls, pwd and any other commands
Hi!
On Tue, May 02, 2006 at 01:22:26PM -0400, Kris Kennaway wrote:
I think it's same problem as in thread fsck_ufs locked in snaplk.
Is this problem fixed in fresh 6.1-PRE?
I think we've reproduced the problem, but it probably won't be fixed
before the release. Sorry, the bug reports
On Tue, May 02, 2006 at 09:44:29PM +0400, Dmitriy Kirhlarov wrote:
Hi!
On Tue, May 02, 2006 at 01:22:26PM -0400, Kris Kennaway wrote:
I think it's same problem as in thread fsck_ufs locked in snaplk.
Is this problem fixed in fresh 6.1-PRE?
I think we've reproduced the problem, but
Dmitriy Kirhlarov wrote:
Hi!
On Tue, May 02, 2006 at 01:22:26PM -0400, Kris Kennaway wrote:
I think it's same problem as in thread fsck_ufs locked in snaplk.
Is this problem fixed in fresh 6.1-PRE?
I think we've reproduced the problem, but it probably won't be fixed
before the
On Tue, May 02, 2006 at 02:03:13PM -0400, Mike Jakubik wrote:
Dmitriy Kirhlarov wrote:
Hi!
On Tue, May 02, 2006 at 01:22:26PM -0400, Kris Kennaway wrote:
I think it's same problem as in thread fsck_ufs locked in snaplk.
Is this problem fixed in fresh 6.1-PRE?
I think we've
Kris Kennaway wrote:
On Tue, May 02, 2006 at 02:03:13PM -0400, Mike Jakubik wrote:
Ditto, same thing with the recent nve fixes. Why release known broken
code when there are tested patches available? Whats the worst that will
happen? It wont work? Thats already the case...
What
Hi!
On Tue, May 02, 2006 at 01:58:59PM -0400, Kris Kennaway wrote:
Well, sorry folks, you should have told me in February. Or if you
only found out about the problem a week ago, you need to recognize
I find it several days ago, when start quota on this server. Another
server with older RC1
On Tue, May 02, 2006 at 02:08:44PM -0400, Mike Jakubik wrote:
Kris Kennaway wrote:
On Tue, May 02, 2006 at 02:03:13PM -0400, Mike Jakubik wrote:
Ditto, same thing with the recent nve fixes. Why release known broken
code when there are tested patches available? Whats the worst that will
On Tue, May 02, 2006 at 10:16:28PM +0400, Dmitriy Kirhlarov wrote:
Hi!
On Tue, May 02, 2006 at 01:58:59PM -0400, Kris Kennaway wrote:
Well, sorry folks, you should have told me in February. Or if you
only found out about the problem a week ago, you need to recognize
I find it several
Quoting Kris Kennaway [EMAIL PROTECTED]:
On February 21 -- that is over 2 months ago -- I sent email to this
list containing a fix for the quota deadlocks that were known at the
time. I got minimal response from users, but it was uniformly
positive. The fix was committed, and the status of
Mike Jakubik wrote:
Dmitriy Kirhlarov wrote:
Hi!
On Tue, May 02, 2006 at 01:22:26PM -0400, Kris Kennaway wrote:
I think it's same problem as in thread fsck_ufs locked in snaplk.
Is this problem fixed in fresh 6.1-PRE?
I think we've reproduced the problem, but it probably won't be
68 matches
Mail list logo