For this weeks meeting agenda...

[Colin Charles]

I'd like to discuss:

FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths
How can we use this further, and get it somewhat Red Hat blessed if we
need to

FedoraUsers - http://fedorausers.org/
This is a site that hasn't actually been born yet, but I'd like the
relevant parties to show up to discuss how we can have a site that makes
sense to one of our largest target markets: USERS

Does anyone have contact information for who's behind fedorausers.org ?
Whois doesn't tell me too much

Triaging guidelines - http://fedoraproject.org/wiki/TriagingGuidelines
How can we market to triagers? How can we make it _absolutely easy_ for
triagers to get on the bandwagon? If I didn't know Bugzilla, how'd I use
it kind of improvements, all user targetted

Fedora LiveCD - http://fedoraproject.org/wiki/Kadischi
We have a LiveCD in the making. LiveCDs as we all know are incredibly
useful as a marketing tool (and a rescue tool, a demo tool, etc.). Let's
find some use cases, maybe more developers to join, and lets give this a
spin if there's techincal prowess among us. I think a lot of inspiration
with regards to marketing livecd's can be honed from
http://live.gnome.org/GnomeLiveCd

Get people on the CMC lists
Greg has created this, I have access to it, so its time to get off my
butt and get cracking

All this in addition to our current schedule. Just thought I'd put this
up here, so we all have some time to prepare. Yes, hefty plate for that
one hour, I don't expect it all to be done then, but hey, I'm giving it
a shot =)
-- 
Colin Charles, http://www.bytebot.net/

[Bryan Smith]

On Sat, 2005-09-10 at 23:37 +1000, Colin Charles wrote:
>  I'd like to discuss:
>  FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths 
You need to _expand_, _technically_ on these points.  Especially the
filesystem comments.  Don't be general, be specific.

E.g., ReiserFS does not support SELinux, ACLs and other extents well,
let alone NFS and other core functionality/compatibility issues are a
"show-stopper" for shipping in Fedora releases.  JFS suffers from much
of the same compatibility.

As far as XFS, be complementary given the fact that it _does_ have a lot
of support -- equaling Ext3 in many areas -- but at this time, most
people only consider the official SGI kernel releases to be of
"production quality" and not the stock kernel implementation.

Those are very sound facts.  If someone would like me to write the
section, I can do so very _point-by-point_ in the focus of what Fedora
(like Red Hat before) must have in a filesystem.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Toshio Kuratomi]

On Sun, 2005-09-11 at 00:03 -0500, Bryan J. Smith wrote:
>  On Sat, 2005-09-10 at 23:37 +1000, Colin Charles wrote:
>  > I'd like to discuss:
>  > FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths
>  
>  You need to _expand_, _technically_ on these points.  Especially the
>  filesystem comments.  Don't be general, be specific.
>  
>  E.g., ReiserFS does not support SELinux, ACLs ... 
Might also be worth mentioning
namesys/Hans Reiser are not working on
adding support (Although some people at SuSE are.)

Last post I saw from the SuSE people was:
  http://lkml.org/lkml/2004/12/7/142
which is in our kernel.  I understand there's been more issues (which
may have been fixed in 2.6.13) reported since.

-Toshio

[Patrick W. Barnes]

Bryan J. Smith wrote:

> On Sat, 2005-09-10 at 23:37 +1000, Colin Charles wrote:
>   
>
> >I'd like to discuss:
> >FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths
> >    
> >
>
> You need to _expand_, _technically_ on these points.  Especially the
> filesystem comments.  Don't be general, be specific.
>
> E.g., ReiserFS does not support SELinux, ACLs and other extents well,
> let alone NFS and other core functionality/compatibility issues are a
> "show-stopper" for shipping in Fedora releases.  JFS suffers from much
> of the same compatibility.
>
> As far as XFS, be complementary given the fact that it _does_ have a lot
> of support -- equaling Ext3 in many areas -- but at this time, most
> people only consider the official SGI kernel releases to be of
> "production quality" and not the stock kernel implementation.
>
> Those are very sound facts.  If someone would like me to write the
> section, I can do so very _point-by-point_ in the focus of what Fedora
> (like Red Hat before) must have in a filesystem.
>
>
>   
>  
I don't know if that is really within the scope of this document.  We
could very easily write entire pages on some of the points (and indeed
we have) and trying to cram it all on this one page would result in a
massive document that goes far beyond what we really want to present
here.  On some of the points that require more elaboration, we can
redirect the users to more information, or write sub-pages of the
FedoraMyths page to hold the more in-depth information.  Please keep in
mind that many people will only be interested in one or two of the
points, and many novice users will be looking at this document.

-- 
Patrick "The N-Man" Barnes
nman64(a)n-man.com

www.n-man.com
--

[Bryan Smith]

On Sun, 2005-09-11 at 01:03 -0500, Patrick Barnes wrote:
>  I don't know if that is really within the scope of this document.
>  We
>  could very easily write entire pages on some of the points (and indeed
>  we have) and trying to cram it all on this one page would result in a
>  massive document that goes far beyond what we really want to present
>  here.  On some of the points that require more elaboration, we can
>  redirect the users to more information, or write sub-pages of the
>  FedoraMyths page to hold the more in-depth information.  Please keep in
>  mind that many people will only be interested in one or two of the
>  points, and many novice users will be looking at this document. 
Okay, I can thin my comments it out into 3-4 statements total.

But I think the current form is technically misleading.  Instead on
focusing on saying what is and isn't experimental, why not say that Red
Hat requires certain functions to work on a filesystem, JFS and ReiserFS
do not, and while XFS does, it's support is still experimental and not
as well tested.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Bryan Smith]

On Sun, 2005-09-11 at 04:54 -0500, Bryan J. Smith wrote:
>  Okay, I can thin my comments it out into 3-4 statements total.
>  But I think the current form is technically misleading.  Instead on
>  focusing on saying what is and isn't experimental, why not say that Red
>  Hat requires certain functions to work on a filesystem, JFS and ReiserFS
>  do not, and while XFS does, it's support is still experimental and not
>  as well tested. 
Dang, I'm half-asleep.  Do-over ...

Say something like ...

Fedora Core, more than most other Linux distributions, requires the
filesystem have extensive kernel/application support.  Ext2/Ext3 have a
long history on all Linux distributions, so kernel and application
support is commonly implemented.  This includes kernel NFS services,
quotas and Extended Attributes (EAs) for things such as Access Control
Lists (ACLs) and Mandatory Access Controls (MACs) like SELinux.

ReiserFS is not working on supporting many of these features, which is a
show-stopper, and has a long history if compatibility issues with
traditional services Fedora Core has been used for.  JFS is still
missing many of these components, and suffers from the same
compatibility history as ReiserFS.  While XFS does have extensive
support, and SGI has produced releases of XFS for prior Red Hat kernels,
Fedora Core has not tested XFS extensive, and there are many known
issues with XFS in the Linux kernel (outside of SGI's control).


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Bryan Smith]

On Sun, 2005-09-11 at 05:01 -0500, Bryan J. Smith wrote:
>  Dang, I'm half-asleep.  Do-over ...
>  Say something like ...
>  
>  Fedora Core, more than most other Linux distributions, requires the
>  filesystem have extensive kernel/application support.  Ext2/Ext3 have a
>  long history on all Linux distributions, so kernel and application
>  support is commonly implemented.  This includes kernel NFS services,
>  quotas and Extended Attributes (EAs) for things such as Access Control
>  Lists (ACLs) and Mandatory Access Controls (MACs) like SELinux.
>  
>  ReiserFS is not working on supporting many of these features, which is a
>  show-stopper, and has a long history if compatibility issues with
>  traditional services Fedora Core has been used for.  JFS is still
>  missing many of these components, and suffers from the same
>  compatibility history as ReiserFS.  While XFS does have extensive
>  support, and SGI has produced releases of XFS for prior Red Hat kernels,
>  Fedora Core has not tested XFS extensive, and there are many known
>  issues with XFS in the Linux kernel (outside of SGI's control). 
Actually, I just re-read that and it's too subjective and, worse yet,
makes statements on behalf of other entities (which we obviously can't
do).

I'll come up with a better set of statements after I get some sleep, as
well as after I find a few, solid mailing list links.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Stuart Ellis]

On Sun, 2005-09-11 at 05:04 -0500, Bryan J. Smith wrote:
>  On Sun, 2005-09-11 at 05:01 -0500, Bryan J. Smith wrote:
>  > Dang, I'm half-asleep.  Do-over ...
>  > Say something like ...
>  > 
>  > Fedora Core, more than most other Linux distributions, requires the
>  > filesystem have extensive kernel/application support.  Ext2/Ext3 have a
>  > long history on all Linux distributions, so kernel and application
>  > support is commonly implemented.  This includes kernel NFS services,
>  > quotas and Extended Attributes (EAs) for things such as Access Control
>  > Lists (ACLs) and Mandatory Access Controls (MACs) like SELinux.
>  > 
>  > ReiserFS is not working on supporting many of these features, which is a
>  > show-stopper, and has a long history if compatibility issues with
>  > traditional services Fedora Core has been used for.  JFS is still
>  > missing many of these components, and suffers from the same
>  > compatibility history as ReiserFS.  While XFS does have extensive
>  > support, and SGI has produced releases of XFS for prior Red Hat kernels,
>  > Fedora Core has not tested XFS extensive, and there are many known
>  > issues with XFS in the Linux kernel (outside of SGI's control).
>  
>  Actually, I just re-read that and it's too subjective and, worse yet,
>  makes statements on behalf of other entities (which we obviously can't
>  do). 
As has already been said, it may not need much detail. The important
points are probably that ReiserFS doesn't yet support key features x, y
and z, and XFS isn't suited or reliable for standard setups.

I tagged that myth on after an IRC discussion about unreasonable user
requests - people semi-regularly claim that Fedora should
support/default to ReiserFS (as SUSE does, I think) because it's
supposedly faster or cleverer or whatever, and that XFS is l33t, so it
should be a standard installation option.

-- 

Stuart Ellis

stuart(a)elsn.org

Fedora Documentation Project: http://fedora.redhat.com/projects/docs/

GPG key ID: 7098ABEA
GPG key fingerprint: 68B0 E291 FB19 C845 E60E  9569 292E E365 7098 ABEA

[Bryan Smith]

On Sun, 2005-09-11 at 16:12 +0100, Stuart Ellis wrote:
>  As has already been said, it may not need much detail. The important
>  points are probably that ReiserFS doesn't yet support key features x, y
>  and z, 
Correct, that _needs_ to be in there.  The problem with so many ReiserFS
advocates is that they've _never_ used it for a production NFS server,
or never used EAs.  And there are recovery issues with the off-line
tools being "out-of-sync" with changes in the filesystem (something that
doesn't happen with Ext3 or XFS which have remained unchanged
structurally for 10+ years).

>  and XFS isn't suited or reliable for standard setups. 
XFS is _very_reliable_, but _only_ in the official SGI releases.  I've
found that XFS in the stock kernel is _not_, and most 3rd party rebuilds
are incomplete and buggy.

I wish Red Hat would support XFS.  It has all the interface/
compatibility, features of Ext3, plus a number more that enterprises
need.  But until Red Hat takes the time to build and test a "complete"
XFS -- I can't trust the stock kernel builds or 3rd parties.

>  I tagged that myth on after an IRC discussion about unreasonable
> user
>  requests - people semi-regularly claim that Fedora should
>  support/default to ReiserFS (as SUSE does, I think) because it's
>  supposedly faster or cleverer or whatever, 
ReiserFS is innovative.  And it utterly _breaks_ standard kernel
interfaces and compatibility as a result.  Even SuSE developers have
told me _not_ to use it for my needs, because their hacks for many
things (from NFS to EAs) are suspect.

>  and that XFS is l33t, so it should be a standard installation option.
Actually, Red Hat needs to realize that Ext3 has scalability issues (I
don't like it above 100GB and I do _not_ trust it above 1TB), and
_lacks_ a lot of user-space features of XFS like xfsdump, xfs_fsr,
etc..., let alone EAs and other meta-data is stored directly in the
inode (which xfsdump then retains).

I still have quite a number of Red Hat Linux 7.3 systems in heavy, heavy
production with SGI's official XFS 1.2.x release, and one major Red Hat
Linux 9 system with SGI's official XFS 1.3.1 release.  But that's the
key issue, they are the _only_ official SGI releases for any distro.

I've tried to integrate SGI's kernel builds from CVS into Red Hat
distros to little avail.  And as I mentioned, the stock kernel releases
are incomplete -- especially the 2.4 backport that is now in the stock
2.4 kernel.  I would _never_ run it, and the stock 2.6 kernel continues
to be suspect.

Which means until Red Hat pro-actively develops and tests XFS in newer
kernel 2.6 Fedora Core releases, I can't trust XFS either.

WHICH MEANS (AND I HOPE RED HAT IS LISTENING ;-), when I need a large,
scalable, high-performance NFS/LAN file server for my Fortune 100
clients -- I deploy Solaris/Opteron now.  Ext3 is _not_ cut it, and it
_never_ will.  I have to agree 100% with Schwartz's comments -- Red Hat
is _ignoring_ a significant segment of the enterprise LAN server market.

I still long for the day of the Ext3 + XFS combination Red Hat
distribution.  Ext3 is better for system and smaller data volumes, XFS
is better for larger (especially large file) volumes.  But that died
once SGI stopped releasing official releases -- the last being 1.3.1 for
Red Hat Linux 9.

It's not about "l33t" -- there are serious enterprise features missing
in Ext3.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Patrick W. Barnes]

Bryan J. Smith wrote:

> On Sun, 2005-09-11 at 16:12 +0100, Stuart Ellis wrote:
>   
>
> >As has already been said, it may not need much detail. The important
> >points are probably that ReiserFS doesn't yet support key features x, y
> >and z,
> >    
> >
>
> Correct, that _needs_ to be in there.  The problem with so many ReiserFS
> advocates is that they've _never_ used it for a production NFS server,
> or never used EAs.  And there are recovery issues with the off-line
> tools being "out-of-sync" with changes in the filesystem (something that
> doesn't happen with Ext3 or XFS which have remained unchanged
> structurally for 10+ years).
>
>   
>
> >and XFS isn't suited or reliable for standard setups.
> >    
> >
>
> XFS is _very_reliable_, but _only_ in the official SGI releases.  I've
> found that XFS in the stock kernel is _not_, and most 3rd party rebuilds
> are incomplete and buggy.
>
> I wish Red Hat would support XFS.  It has all the interface/
> compatibility, features of Ext3, plus a number more that enterprises
> need.  But until Red Hat takes the time to build and test a "complete"
> XFS -- I can't trust the stock kernel builds or 3rd parties.
>
>   
>
> >I tagged that myth on after an IRC discussion about unreasonable user
> >requests - people semi-regularly claim that Fedora should
> >support/default to ReiserFS (as SUSE does, I think) because it's
> >supposedly faster or cleverer or whatever,
> >    
> >
>
> ReiserFS is innovative.  And it utterly _breaks_ standard kernel
> interfaces and compatibility as a result.  Even SuSE developers have
> told me _not_ to use it for my needs, because their hacks for many
> things (from NFS to EAs) are suspect.
>
>   
>
> >and that XFS is l33t, so it should be a standard installation option.
> >    
> >
>
> Actually, Red Hat needs to realize that Ext3 has scalability issues (I
> don't like it above 100GB and I do _not_ trust it above 1TB), and
> _lacks_ a lot of user-space features of XFS like xfsdump, xfs_fsr,
> etc..., let alone EAs and other meta-data is stored directly in the
> inode (which xfsdump then retains).
>
> I still have quite a number of Red Hat Linux 7.3 systems in heavy, heavy
> production with SGI's official XFS 1.2.x release, and one major Red Hat
> Linux 9 system with SGI's official XFS 1.3.1 release.  But that's the
> key issue, they are the _only_ official SGI releases for any distro.
>
> I've tried to integrate SGI's kernel builds from CVS into Red Hat
> distros to little avail.  And as I mentioned, the stock kernel releases
> are incomplete -- especially the 2.4 backport that is now in the stock
> 2.4 kernel.  I would _never_ run it, and the stock 2.6 kernel continues
> to be suspect.
>
> Which means until Red Hat pro-actively develops and tests XFS in newer
> kernel 2.6 Fedora Core releases, I can't trust XFS either.
>
> WHICH MEANS (AND I HOPE RED HAT IS LISTENING ;-), when I need a large,
> scalable, high-performance NFS/LAN file server for my Fortune 100
> clients -- I deploy Solaris/Opteron now.  Ext3 is _not_ cut it, and it
> _never_ will.  I have to agree 100% with Schwartz's comments -- Red Hat
> is _ignoring_ a significant segment of the enterprise LAN server market.
>
> I still long for the day of the Ext3 + XFS combination Red Hat
> distribution.  Ext3 is better for system and smaller data volumes, XFS
> is better for larger (especially large file) volumes.  But that died
> once SGI stopped releasing official releases -- the last being 1.3.1 for
> Red Hat Linux 9.
>
> It's not about "l33t" -- there are serious enterprise features missing
> in Ext3.
>
>
>   
>  
Suggestion:  Why don't you go ahead and write a wiki page about the
assorted filesystems, their strengths and weaknesses, and why some
aren't currently in Fedora.  You could create the page at
http://fedoraproject.org/wiki/FAQ/FileSystems and let me know when it is
ready for review.  Once we have a good version in place, we can write in
just a small statement on the FedoraMyths page and direct the curious to
the new page.  If you would like to follow the progress of filesystem
support and keep that page a living document, that would be great. 
We'll also create a link to the new page directly off of the FAQ.

-- 
Patrick "The N-Man" Barnes
nman64(a)n-man.com

www.n-man.com
--

[Bryan Smith]

On Sun, 2005-09-11 at 15:53 -0500, Patrick Barnes wrote:
>  Suggestion:  Why don't you go ahead and write a wiki page about
> the
>  assorted filesystems, their strengths and weaknesses, and why some
>  aren't currently in Fedora.  You could create the page at
>  http://fedoraproject.org/wiki/FAQ/FileSystems and let me know when it is
>  ready for review.  Once we have a good version in place, we can write in
>  just a small statement on the FedoraMyths page and direct the curious to
>  the new page.  If you would like to follow the progress of filesystem
>  support and keep that page a living document, that would be great. 
>  We'll also create a link to the new page directly off of the FAQ. 
Actually, I've written a _lot_ on XFS since I produced some RPMs back in
early 2001 (just months prior to the official SGI XFS 1.0 release).

I have a larger, technical discussion on Ext3 and XFS is in my blog here
which covers a _lot_ of recent status:  
  http://thebs413.blogspot.com/2005/08/filesystem-fundamentals-and-practice...  

I could probably dig up my posts from 2000, 2001, 2002 on-ward on my
usage of different filesystems.  One legacy one from June 2001 is
here:  
  http://www.geocities.com/thebs413/ITEC_JFS_2001Jun13.pdf  

In a nutshell, I adopted Ext3 in 2000 after SuSE told me that ReiserFS
would not be a feasible filesystem for kernel NFS.  I have a few
colleagues with SuSE that are "frank" on the continuing issues with
ReiserFS for such applications.

As far as JFS, the problem was that it was ported from OS/2, and _not_
the AIX implementation.  So the Linux JFS port was basically *0* ready
for an UNIX-like platform, and the resulting "re-write" occurred.  It
had more to do with IBM honoring their Non-Compete clause with SCO in
their Monterey agreement.  This was prior to IBM breaking and violating
the agreement (which SCO sued over years later), so they were honoring
it at the time the Linux JFS fork came about.

Since then, I've been deploying XFS.  Once the SGI XFS 1.2 release hit,
I put it into serious production.  But the lack of good integration in
the kernel, even 2.6 (although much better than 2.4) has left me _not_
trusting anything except official SGI XFS kernels.  And I've had trouble
integrating anything but the old 1.2.x for RHL7.3 and 1.3.1 for RHL9
releases.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Colin Charles]

On Sun, 2005-09-11 at 15:53 -0500, Patrick Barnes wrote:
>  Suggestion:  Why don't you go ahead and write a wiki page about
> the
>  assorted filesystems, their strengths and weaknesses, and why some
>  aren't currently in Fedora.  You could create the page at
>  http://fedoraproject.org/wiki/FAQ/FileSystems and let me know when it
>  is
>  ready for review.  Once we have a good version in place, we can write
>  in
>  just a small statement on the FedoraMyths page and direct the curious
>  to
>  the new page.  If you would like to follow the progress of filesystem
>  support and keep that page a living document, that would be great. 
>  We'll also create a link to the new page directly off of the FAQ. 
Yes Bryan, if you think the filesystem issue is /that/ important, that
users of Fedora must know it all, then please by all means, create a
sub-page on the wiki for this. We'll be happy to review it, and I'm sure
we can get more technical heads from fedora-devel-list interested

In the meantime, telling the marketing list about filesystem limitations
does not help us in our cause in any way. If you're unhappy about the
XFS file support, maybe you want to escalate working on it, cleaning it
up, and submitting patches in Bugzilla (and about now you're going to
point to eon old bugs open asking for xfs to be updated...) or better
still, the upstream kernel

Best regards
-- 
Colin Charles, http://www.bytebot.net/

[Bryan Smith]

On Mon, 2005-09-12 at 11:54 +1000, Colin Charles wrote:
>  Yes Bryan, if you think the filesystem issue is /that/ important,
> that
>  users of Fedora must know it all, then please by all means, create a
>  sub-page on the wiki for this. We'll be happy to review it, and I'm sure
>  we can get more technical heads from fedora-devel-list interested 
And I said I _would_ do this over the next few days.  I'm working 7 days
a week right now, and will get you it ASAP.  It will be polished and
professional in tone, focus as well as legal-base.

>  In the meantime, telling the marketing list about filesystem
> limitations
>  does not help us in our cause in any way. 
I was just pointing out the fact that Fedora/Red Hat has continually
hurt itself on the filesystems point because it does _not_ directly
address why ReiserFS and JFS are not included, plays games in general it
does _not_ have to -- but even worse -- goes one step further, acting
like all of us long-time integrators and Ext3 proponents don't exist
when we say "it's time to get serious about XFS."

>  If you're unhappy about the XFS file support, 
No.  That's not it at all!  Do _not_ belittle my statements as such
because you are belittling a _lot_ of consultants at Fortune 100
companies when you say such.

What I'm unhappy about the "marketing" aspects of why XFS is _not_
supported.  I would much rather Fedora/Red Hat say, "we can't afford the
personnel to develop and support a 2nd filesystem."  That would be
direct, fair and understanding.

But the comments that Ext3 does everything XFS does, Ext3 does not have
limitations that XFS solves, refer to it as "experimental" and other
comments insult the intelligence of those of us who actually deploy
Fedora/Red Hat solutions.

That's the problem.  Again, if Red Hat says it does not want to support
a 2nd filesystem, that's one thing.  But 99% of the comments I see on
XFS, and why Fedora/Red Hat is not looking at it for the future, are
full of FUD.

Sadly enough, it's in Fedora/Red Hat's own, best interest.  Even Sun has
basically all but said it.

>  maybe you want to escalate working on it, cleaning it
>  up, and submitting patches in Bugzilla (and about now you're going to
>  point to eon old bugs open asking for xfs to be updated...) or better
>  still, the upstream kernel 
You're kidding me, right?
Do you _know_ the effort that would be involved?!

There is already a company, SGI, and their OSS team that is
_more_than_willing_ to work with Red Hat and the Fedora developers.  But
at this point, from what I've seen, it's been a 1-way street.  That's
the problem.

My point has been, and will continue to be, that Red Hat, including
those who work on Fedora Core as their paid job functions, needs to put
people on integration and support of XFS for its _own_ future.  It's not
"XFS v. Ext3" -- it's "XFS to complement Ext3."

99% of the comments I see are "versus" comments.  Hence the continued
problem, hence the actual reversal _away_ from Fedora Core and Red Hat
Enterprise Linux for some applications.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

What I'm unhappy about the "marketing" aspects of why XFS is _not_

> supported.  I would much rather Fedora/Red Hat say, "we can't
> afford the
> personnel to develop and support a 2nd filesystem."  That would be
> direct, fair and understanding.
>   
>  
The official answer for RHEL is here: 
http://kbase.redhat.com/faq/FAQ_80_5737.shtm

> But the comments that Ext3 does everything XFS does, Ext3 does not
> have
> limitations that XFS solves, refer to it as "experimental" and other
> comments insult the intelligence of those of us who actually deploy
> Fedora/Red Hat solutions.
>
> That's the problem.  Again, if Red Hat says it does not want to support
> a 2nd filesystem, that's one thing.  But 99% of the comments I see on
> XFS, and why Fedora/Red Hat is not looking at it for the future, are
> full of FUD.
>  
Again, You should be specific about what you consider FUD in the Fedora 
Myths page. Avoid long rants.  If anyone is willing to step up and 
maintain XFS or any other filesystem for that matter in Fedora, it can 
be send to the fedora development list.

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 09:00 +0530, Rahul Sundaram wrote:
>  The official answer for RHEL is here: 
>  http://kbase.redhat.com/faq/FAQ_80_5737.shtm 
And while it is applicable to ReiserFS and JFS, it has _not_ been
applicable to XFS.  There have been numerous points during the 2.4
kernel where XFS was _better_ than Ext3 in support areas Red Hat sorely
needed -- from quotas to ACLs to (and still today) filesystem backup
(especially on-line).

This answer is oblivious to reality on XFS.  I've read the same comments
again and again -- others have done the same -- and everytime it's FUD.
People making the comments are actually unfamiliar with all the issues
solved by XFS.

Again, that viewpoint _is_ applicable to ReiserFS and JFS.  As a
proponent of Ext3 myself, I _do_ advocate why Red Hat does not support
ReiserFS and JFS.  But that advocacy _falls_flat_ when it comes to XFS.
Anyone who knows the history of XFS' development and release on Linux
knows this.

It's a _total_joke_ in the XFS group when Fedora/Red Hat come back and
say "oh, there's nothing XFS does that Ext3 doesn't do" and there are
all these "exceptions" they end up agreeing to -- such as the lack of so
many "standard" on-line user-space tools in traditional UNIX filesystems
that XFS has (since day 1 on Linux), the varying EA support and history
of quota support, the removal on relying on the LVM2/DM stack to solve
so many things (such as the lack of on-line user-space tools) which
introduce more race conditions, etc...

That's what I'm talking about -- key enterprise features that are
_expected_ in a multi-TB UNIX filesystem.  Which forces us to send our
clients to Solaris, instead of Fedora Core or Red Hat Enterprise Linux.

>  Again, You should be specific about what you consider FUD in the
> Fedora 
>  Myths page. Avoid long rants. 
I've got plenty of specifics, and I've tried to address them here, as
well as in posts before.  But apparently you don't want to read what you
consider a "long rant."  That's just sad, because it is _not_ a "long
rant."

>  If anyone is willing to step up and maintain XFS or any other
> filesystem
>  for that matter in Fedora, it can be send to the fedora development
>  list. 
Try SGI and the XFS team.  Unfortunately, it takes some level of "formal
engagement" from Red Hat/Fedora kernel developers too.  That hasn't
happened from what I've seen over 4 years and, sadly enough, it's in Red
Hat's own, best interest.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Bryan J. Smith wrote:

> On Mon, 2005-09-12 at 09:00 +0530, Rahul Sundaram wrote:
>   
>
> >The official answer for RHEL is here: 
> >http://kbase.redhat.com/faq/FAQ_80_5737.shtm
> >    
> >
>
> And while it is applicable to ReiserFS and JFS, it has _not_ been
> applicable to XFS.  There have been numerous points during the 2.4
> kernel where XFS was _better_ than Ext3 in support areas Red Hat sorely
> needed -- from quotas to ACLs to (and still today) filesystem backup
> (especially on-line).
>  
That FAQ doesnt talk about specific features and hence applies to all 
filesystems

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 09:27 +0530, Rahul Sundaram wrote:
>  That FAQ doesnt talk about specific features and hence applies to all
>
>  filesystems 
Again, that's the blanket problem.
And it's not about "features."

So from the Fedora Myths page, it's about the reality that Fedora Core
ships as a platform that does X, Y and Z.

Because ReiserFS and JFS don't support X, Y and/or Z, they are not
options -- and likely never will be until they do.  The lack of such
support is the "show stopper" for inclusion.

But then that's a problem for FS, because it _does_ do X, Y and Z -- EA
ACL/SELinux, Quotas, NFS, etc...  So there is another reason why it's
not included by default.  E.g., XFS has known issues with 4KiB stacks,
and some NFS support issues have been introduced in kernel 2.6 (that
didn't exist with SGI's kernel 2.4 implementations).

I can be very, very technically accurate, but not "over-assuming" at the
same time -- especially from the legal aspect of claiming things that
might be disputable.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Hi

> doOn Mon, 2005-09-12 at 09:27 +0530, Rahul Sundaram wrote:
>   
>
> >That FAQ doesnt talk about specific features and hence applies to all 
> >filesystems
> >    
> >
>
> Again, that's the blanket problem.
> And it's not about "features."
>   
>  
Its also about resources.  Fedora myths page can simply say this " Red 
Hat does not have the resources require to support other filesystems 
besides ext3 though it is provided as a option during the installation 
with specific parameters. If the community has sufficient interest in 
other filesystems to maintain it,  send in a proposal to the Fedora 
development list with the details."

I would recommend not talking about specific features unless someone is 
willing to keep it updated
regards
Rahul

[Rahul Sundaram]

> >
> >WHICH MEANS (AND I HOPE RED HAT IS LISTENING ;-), when I need a large,
> >scalable, high-performance NFS/LAN file server for my Fortune 100
> >clients -- I deploy Solaris/Opteron now.
> > 

>  Ext3 is _not_ cut it, and it
> _never_ will.  I have to agree 100% with Schwartz's comments -- Red Hat
> is _ignoring_ a significant segment of the enterprise LAN server market.
> \\\\\
>   
>  
Never is a long time. Ext3 is under active development and will continue 
to get more enhancements . If you want to send in customer feedback 
using the appropriate support channels would be a better idea. If you do 
believe that the page can highlight things in a better way that is 
invasive, make a alternative wiki page as a draft and point that to this 
list for dicussions

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 04:13 +0530, Rahul Sundaram wrote:
>  Never is a long time. Ext3 is under active development and will
> continue 
>  to get more enhancements . 
Let me be the _first_ to thank Red Hat and Tweedie for the Ext3
filesystem.  I trust it implicitly.  It started saving my bacon in early
2000, and I have trusted it ever since.  It was a much needed,
evolutionary filesystem from _trusted_ Ext2 which has not changed since
the mid-'90s.  There's nothing like knowing you can always do a full
Ext2 fsck -- that's something I trust.

*BUT* I have 2 standing rules on my Ext3 deployments:

1.  No Ext3 filesystem is ever larger than 1TB.  I try to keep them
100GB or smaller.  I still use Ext3 for system volumes, and I would
_never_ use another filesystem for /, /tmp, /var and several other
filesystems.

2.  Avoid using Extended Attributes (EAs) on Ext3 for data filesystems,
except for SELinux.  This is more of a backup consideration than
anything else.  There is no "reliable" way to backup/restore EAs from
Ext3.  Sorry, although I personally like Jorg's work, star is not what I
consider "enterprise quality."

Again, Ext3 will _continue_ to be a _good_ filesystem for Red Hat and
its integrators like myself.  But I can_not_ entrust it as a solution
for multi-TB filesystems.  Unfortunately, there's not much I can in
Linux, and that's the problem -- one Red Hat should help solve.

XFS is a no-brainer.

XFS' structure was designed for 64-bit, with extents, with delayed
writes, with EAs and other meta-data in the inodes, with a _full_suite_
of user-space tools from xfsdump to xfs_fsr (defragmentor) to the off-
line xfs_repair tool.

KEY POINT:  This structure has _not_ changed since 1995.

Everything has been built around that structure since 1995.  There are
no new "hacks" to "extend" the filesystem's design.  It was not
left
"incomplete."  And XFS was a _direct_port_ from Irix to Linux, bringing
a _lot_ of capability to the kernel itself.

That included POSIX EA support from day 1, _official_ quota support even
_before_ Ext3, excellent kernel NFS support in the SGI XFS releases for
kernel 2.4 (although this is no longer the case) and many other details.
That is what I trusted from 2001-2004 -- especially the XFS 1.2 release
for Red Hat Linux 7.3 and 1.3.1 release for Red Hat Linux 9.

Unfortunately, there is _no_ other XFS release I can trust.  It's not
the filesystem itself, it's the _lack_ of consideration on several
levels.  XFS and SGI could really, really, _really_ use Red Hat's
assistance on developing XFS as a "standard" filesystem for 2.6 kernel
distributions.  It's not that XFS is "experimental," it's just the
changes that have occurred in the kernel since the filesystem first came
over, and was extremely mission critical.

Including Red Hat changes like 4KiB stacks -- which I _do_ believe is a
good idea.  In fact, I have been extremely complementary and supportive
of Red Hat's decision making versus SuSE, Debian and other
distributions.  9 times out of 10, Red Hat makes the best choice time
and time again.  Ext3 was one of them.  But Ext3 cannot continue to be
extended and still leverage its trusted structure that was _not_
designed for the scale and features that enterprises now need.

The problem is that everytime I bring this up, people treat me like I'm
some ReiserFS enthusiast who has never run large-scale NFS/SMB servers,
or a JFS advocate who doesn't know it's history on Linux.  I _do_ use
Ext3 and I will _continue_ to deploy Ext3 for many filesystems.  But
Ext3 is not "cutting it" for some large data filesystems, and it will
never offer what XFS has -- and more importantly -- has had since the
mid-'90s.

ReiserFS and JFS are basically "no way" on Red Hat, with Red Hat's
services focus, and I explain this to people regularly.  Even SuSE
developers admit what ReiserFS cannot support, and most understand why
Red Hat does not support it.  But XFS quite different in its
compatibility, and it is only because of kernel changes and lack of
distro support why issues are now occurring.

>  If you want to send in customer feedback using the appropriate
> support
>  channels would be a better idea. If you do believe that the page can
>  highlight things in a better way that is invasive, make a alternative
>  wiki page as a draft and point that to this list for dicussions 
I will.  Until then, I invite people to read my blog.

I'm an Ext3 system integrator.  The problem is that the lack of XFS as a
complement to Ext3 is causing me to deploy Solaris/Opteron instead of
Linux/Opteron for LAN servers now.  Again, I'm not so pundit with little
experience deploying mission critical LAN file servers claiming superior
performance out of ReiserFS, or JFS advocate who is oblivious to the
origins of Linux's JFS which has caused it's support issues.

I really don't prefer Sun.  But because of Red Hat's stance, I've gone
from moving from Solaris to Linux only to be moving back towards Solaris
for some services.  Linux is great for web servers, database servers
(using "raw" volumes for Oracle, or smaller MySQL/PostgresSQL
filesystems), grid computing and other areas.  But for large file
servers, the lack of a turnkey Linux distro with XFS (since the older
SGI releases), has pushed me back to UNIX, and Solaris is where it's at
for LAN fileservers -- especially on Opteron.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Patrick W. Barnes]

Bryan J. Smith wrote:

> On Mon, 2005-09-12 at 04:13 +0530, Rahul Sundaram wrote:
>   
>
> >Never is a long time. Ext3 is under active development and will continue 
> >to get more enhancements .
> >    
> >
>
> Let me be the _first_ to thank Red Hat and Tweedie for the Ext3
> filesystem.  I trust it implicitly.  It started saving my bacon in early
> 2000, and I have trusted it ever since.  It was a much needed,
> evolutionary filesystem from _trusted_ Ext2 which has not changed since
> the mid-'90s.  There's nothing like knowing you can always do a full
> Ext2 fsck -- that's something I trust.
>
> *BUT* I have 2 standing rules on my Ext3 deployments:
>
> 1.  No Ext3 filesystem is ever larger than 1TB.  I try to keep them
> 100GB or smaller.  I still use Ext3 for system volumes, and I would
> _never_ use another filesystem for /, /tmp, /var and several other
> filesystems.
>
> 2.  Avoid using Extended Attributes (EAs) on Ext3 for data filesystems,
> except for SELinux.  This is more of a backup consideration than
> anything else.  There is no "reliable" way to backup/restore EAs from
> Ext3.  Sorry, although I personally like Jorg's work, star is not what I
> consider "enterprise quality."
>
> Again, Ext3 will _continue_ to be a _good_ filesystem for Red Hat and
> its integrators like myself.  But I can_not_ entrust it as a solution
> for multi-TB filesystems.  Unfortunately, there's not much I can in
> Linux, and that's the problem -- one Red Hat should help solve.
>
> XFS is a no-brainer.
>
> XFS' structure was designed for 64-bit, with extents, with delayed
> writes, with EAs and other meta-data in the inodes, with a _full_suite_
> of user-space tools from xfsdump to xfs_fsr (defragmentor) to the off-
> line xfs_repair tool.
>
> KEY POINT:  This structure has _not_ changed since 1995.
>
> Everything has been built around that structure since 1995.  There are
> no new "hacks" to "extend" the filesystem's design.  It was not
> left
> "incomplete."  And XFS was a _direct_port_ from Irix to Linux, bringing
> a _lot_ of capability to the kernel itself.
>
> That included POSIX EA support from day 1, _official_ quota support even
> _before_ Ext3, excellent kernel NFS support in the SGI XFS releases for
> kernel 2.4 (although this is no longer the case) and many other details.
> That is what I trusted from 2001-2004 -- especially the XFS 1.2 release
> for Red Hat Linux 7.3 and 1.3.1 release for Red Hat Linux 9.
>
> Unfortunately, there is _no_ other XFS release I can trust.  It's not
> the filesystem itself, it's the _lack_ of consideration on several
> levels.  XFS and SGI could really, really, _really_ use Red Hat's
> assistance on developing XFS as a "standard" filesystem for 2.6 kernel
> distributions.  It's not that XFS is "experimental," it's just the
> changes that have occurred in the kernel since the filesystem first came
> over, and was extremely mission critical.
>
> Including Red Hat changes like 4KiB stacks -- which I _do_ believe is a
> good idea.  In fact, I have been extremely complementary and supportive
> of Red Hat's decision making versus SuSE, Debian and other
> distributions.  9 times out of 10, Red Hat makes the best choice time
> and time again.  Ext3 was one of them.  But Ext3 cannot continue to be
> extended and still leverage its trusted structure that was _not_
> designed for the scale and features that enterprises now need.
>
> The problem is that everytime I bring this up, people treat me like I'm
> some ReiserFS enthusiast who has never run large-scale NFS/SMB servers,
> or a JFS advocate who doesn't know it's history on Linux.  I _do_ use
> Ext3 and I will _continue_ to deploy Ext3 for many filesystems.  But
> Ext3 is not "cutting it" for some large data filesystems, and it will
> never offer what XFS has -- and more importantly -- has had since the
> mid-'90s.
>
> ReiserFS and JFS are basically "no way" on Red Hat, with Red Hat's
> services focus, and I explain this to people regularly.  Even SuSE
> developers admit what ReiserFS cannot support, and most understand why
> Red Hat does not support it.  But XFS quite different in its
> compatibility, and it is only because of kernel changes and lack of
> distro support why issues are now occurring.
>
>   
>
> >If you want to send in customer feedback using the appropriate support
> >channels would be a better idea. If you do believe that the page can
> >highlight things in a better way that is invasive, make a alternative
> >wiki page as a draft and point that to this list for dicussions
> >    
> >
>
> I will.  Until then, I invite people to read my blog.
>
> I'm an Ext3 system integrator.  The problem is that the lack of XFS as a
> complement to Ext3 is causing me to deploy Solaris/Opteron instead of
> Linux/Opteron for LAN servers now.  Again, I'm not so pundit with little
> experience deploying mission critical LAN file servers claiming superior
> performance out of ReiserFS, or JFS advocate who is oblivious to the
> origins of Linux's JFS which has caused it's support issues.
>
> I really don't prefer Sun.  But because of Red Hat's stance, I've gone
> from moving from Solaris to Linux only to be moving back towards Solaris
> for some services.  Linux is great for web servers, database servers
> (using "raw" volumes for Oracle, or smaller MySQL/PostgresSQL
> filesystems), grid computing and other areas.  But for large file
> servers, the lack of a turnkey Linux distro with XFS (since the older
> SGI releases), has pushed me back to UNIX, and Solaris is where it's at
> for LAN fileservers -- especially on Opteron.
>
>
>   
>  
1. You're ranting.
2. You're doing it in the wrong place.
3. If you turn your passions into action, rather than ranting, you might
be pleasantly surprised by the results.

Your opinions regarding the assorted filesystems really are irrelevant
to this list, as are the technical details behind those reasons.  It is
not that nobody cares, but nobody here is prepared to do anything about
it.  You should offer assistance to the developers of XFS on Linux.  If
upstream support for XFS improves, Red Hat and/or the Fedora Foundation
will be more inclined to support it.  As Rahul and I have suggested, you
are welcome to create an *objective* wiki page, describing the reasons
behind the inclusion/exclusion of different filesystem types.  Anything
else is off-topic on this list, and is likely to be met only with
hostility or frustration.

-- 
Patrick "The N-Man" Barnes
nman64(a)n-man.com

www.n-man.com
--

[Rahul Sundaram]

Bryan J. Smith wrote:

> On Mon, 2005-09-12 at 04:13 +0530, Rahul Sundaram wrote:
>   
>
> >Never is a long time. Ext3 is under active development and will continue 
> >to get more enhancements .
> >    
> >
>
> Let me be the _first_ to thank Red Hat and Tweedie for the Ext3
> filesystem.  I trust it implicitly.  It started saving my bacon in early
> 2000, and I have trusted it ever since.  It was a much needed,
> evolutionary filesystem from _trusted_ Ext2 which has not changed since
> the mid-'90s.  There's nothing like knowing you can always do a full
> Ext2 fsck -- that's something I trust.
>
> *BUT* I have 2 standing rules on my Ext3 deployments:
>
> 1.  No Ext3 filesystem is ever larger than 1TB.  I try to keep them
> 100GB or smaller.  I still use Ext3 for system volumes, and I would
> _never_ use another filesystem for /, /tmp, /var and several other
> filesystems.
>   
>  
While I dont really want to take this further on this list, let me point 
to you the RHEL 4 U1 release notes.

"The ext2 and ext3 filesystems have an internal limit of 8 TB. Devices 
up to this limit have been tested in Red Hat Enterprise Linux 4 Update 1."

> 2.  Avoid using Extended Attributes (EAs) on Ext3 for data
> filesystems,
> except for SELinux.  This is more of a backup consideration than
> anything else.  There is no "reliable" way to backup/restore EAs from
> Ext3.  Sorry, although I personally like Jorg's work, star is not what I
> consider "enterprise quality."
>   
>  
This is not a filesystem limitation

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 06:06 +0530, Rahul Sundaram wrote:
>  While I dont really want to take this further on this list, let me
> point 
>  to you the RHEL 4 U1 release notes.
>  "The ext2 and ext3 filesystems have an internal limit of 8 TB. Devices 
>  up to this limit have been tested in Red Hat Enterprise Linux 4 Update 1." 
And I understand it has even been raised to 17.6TB (16TiB) now.
That's _not_ what I'm talking about.

>  This is not a filesystem limitation 
But lack of a good set of safe, trusted user-space utilities for the
filesystem are.  Red Hat offers a non-filesystem utility, star.  That's
not good for enterprises.  And the more Red Hat tries to distract people
from that, the more they are denying what they cannot offer.

That's _exactly_ the self-defeating marketing I'm talking about.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Hi

> And I understand it has even been raised to 17.6TB (16TiB) now.
> That's _not_ what I'm talking about.
>   
>  
Your personal choices are not a list concern

> >This is not a filesystem limitation
> >    
> >
>
> That's _exactly_ the self-defeating marketing I'm talking about.
>   
>  
We are not talking about RHEL marketing here

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 06:26 +0530, Rahul Sundaram wrote:
>  Your personal choices are not a list concern 
That's fine.  Of course, what you call my "personal choices" is just one
of the _many_ voices.  If Fedora and Red Hat want to ignore these
voices, then they make their own issues.

Again, I'm not some garden variety punk who thinks Fedora should have
this, Fedora should have that.  I'm telling you not only what people who
believe in Fedora Core, like Red Hat Linux before it, are feeling, but
what others are picking up on.

>  We are not talking about RHEL marketing here 
It's all the same.  What doesn't go into Fedora Core rarely makes it
into Red Hat Enterprise Linux.  Frankly, the tunnel vision I see
regularly is what bothers me most.

But the thing I always remember, which is the reason I love Red Hat, is
the developers.  Red Hat is a GPL company first and foremost, and until
that changes, tunnel vision -- let alone the larger issue of marketing
-- really doesn't matter to me.

But I'm still forced to deploy Sun when I'd rather deploy Fedora/Red
Hat.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Hi

>   
>
> >We are not talking about RHEL marketing here
> >    
> >
>
> It's all the same.  What doesn't go into Fedora Core rarely makes it
> into Red Hat Enterprise Linux.  Frankly, the tunnel vision I see
> regularly is what bothers me most.
>  
 Its not tunnel vision. RHEL features and bugfixes are driven by the 
customer requests and support calls in a very real sense and not rants 
in a mailing list especially not ones in Fedora marketing list. If you 
want Red Hat to listen in relationship to RHEL, try 
http://redhat.com/support.

For Fedora, here is how you contribute. Note that the amount of 
oppurtunities  is much more than RHEL

* File bug reports
* File feature requests with good rationale. I would like to have more 
than 8 TB filesystems for such and such scenarios might be better than 
"I want XFS"
* Write code and send in patches
* Write documentation
* Translate UI and docs
* Package in Fedora Extras
* Art work, themes and sound themes
* Advocate
* Last but not the least, use it

There might be a few I have missed


regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 08:51 +0530, Rahul Sundaram wrote:
>  Its not tunnel vision. 
> From the answers I've seen, it is -- it damn is. 
>  RHEL features and bugfixes are driven by the customer requests and
>  support calls in a very real sense 
Because Red Hat is getting web servers and Oracle servers (raw slices)
and grid computing sales.  Red Hat is now _losing_ multi-TB file server
sales because they can_not_ deliver on it with just Ext3 -- especially
given its lack of user-space filesystem support.

Sun is eating that up.  I guess Red Hat has decided it is not worth
doing?  Or is it because the Fedora developments, including those by its
own developers who obliviously think "Ext3 doesn't do anything XFS
does."

That's marketing -- however you slice it -- and a major issue.

>  and not rants in a mailing list 
This isn't a "rant."  This is me, a system integrator at Fortune 100
companies over the last 4 years, telling you not only what I have
experienced, but what others have experienced.  We cannot offer Fedora
Core and Red Hat Enterprise Linux solutions -- and it starts with Fedora
developments.

The fact that you're taking it as a "rant" instead of -- "hmmm, maybe
there is a 4-year 'suggestion' this guy is trying to make" -- goes to
the heart of the matter.  I'm not the only "ranting lunatic" out there
having to "move back" to Solaris for solutions.  

>  especially not ones in Fedora marketing list. 
I honestly don't know where else to turn.  And when I read the rather
"say nothing" portion that just feeds the ReiserFS and JFS proponents
their "Red Hat is unfairly biased towards Ext3" non-sense, I have to
agree with them from the XFS mis-conceptions presented.

>  If you want Red Hat to listen in relationship to RHEL, try 
>  http://redhat.com/support. 
Been there -- a good 4 years.

>  For Fedora, here is how you contribute. Note that the amount of 
>  oppurtunities  is much more than RHEL 
Yes, I know.

>  * File bug reports
>  * File feature requests with good rationale. I would like to have more 
>  than 8 TB filesystems for such and such scenarios might be better than 
>  "I want XFS" 
You didn't hear a word I said.  That is so sad.

>  * Write code and send in patches 
I'll get farther with SGI than I will Red Hat.  I'm asking -- at the
Fedora level -- to try to get Red Hat involved.  I've been trying
various avenues for 4 years now to no avail.

And it's only getting more pertinent as Solaris is now on Opteron.

>  * Write documentation 
Way, way -- WAY AHEAD -- of you there.

>  * Translate UI and docs
>  * Package in Fedora Extras 
I'm actually looking to submit several Fedora Extras packages.

>  * Art work, themes and sound themes
>  * Advocate 
I'm well known as a "Red Hat apologist" (even though I hate the title).
So I find it humorous that you would suggest that to me.  ;->

>  * Last but not the least, use it 
Considering I've been integrating Fedora Core in companies as large as
one major Fortune 20 company in just the last 2 years, and several other
Fortune 100 companies as well -- again, WAY AHEAD OF YOU THERE.

>  There might be a few I have missed 
Again, I'm not some "I want filesystem X" puke.
But I know you'll feel free to take my comments as such.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Greg DeKoenigsberg]

Boys, boys.  You're both pretty.  :)

Bryan, I get your point.  Lemme talk to some smart folks and see what the 
general feeling is around XFS "inside the walls."  I'm kinda curious 
myself now.

--g

_____________________  ____________________________________________
  Greg DeKoenigsberg ] [ the future masters of technology will have
 Community Relations ] [ to be lighthearted and intelligent.  the
             Red Hat ] [ machine easily masters the grim and the 
                     ] [ dumb.  --mcluhan

On Sun, 11 Sep 2005, Bryan J. Smith wrote:

>  On Mon, 2005-09-12 at 08:51 +0530, Rahul Sundaram wrote:
>  > Its not tunnel vision.
>  
>  >From the answers I've seen, it is -- it damn is.
>  
>  > RHEL features and bugfixes are driven by the customer requests and
>  > support calls in a very real sense
>  
>  Because Red Hat is getting web servers and Oracle servers (raw slices)
>  and grid computing sales.  Red Hat is now _losing_ multi-TB file server
>  sales because they can_not_ deliver on it with just Ext3 -- especially
>  given its lack of user-space filesystem support.
>  
>  Sun is eating that up.  I guess Red Hat has decided it is not worth
>  doing?  Or is it because the Fedora developments, including those by its
>  own developers who obliviously think "Ext3 doesn't do anything XFS
>  does."
>  
>  That's marketing -- however you slice it -- and a major issue.
>  
>  > and not rants in a mailing list
>  
>  This isn't a "rant."  This is me, a system integrator at Fortune 100
>  companies over the last 4 years, telling you not only what I have
>  experienced, but what others have experienced.  We cannot offer Fedora
>  Core and Red Hat Enterprise Linux solutions -- and it starts with Fedora
>  developments.
>  
>  The fact that you're taking it as a "rant" instead of -- "hmmm, maybe
>  there is a 4-year 'suggestion' this guy is trying to make" -- goes to
>  the heart of the matter.  I'm not the only "ranting lunatic" out there
>  having to "move back" to Solaris for solutions.  
>  
>  > especially not ones in Fedora marketing list.
>  
>  I honestly don't know where else to turn.  And when I read the rather
>  "say nothing" portion that just feeds the ReiserFS and JFS proponents
>  their "Red Hat is unfairly biased towards Ext3" non-sense, I have to
>  agree with them from the XFS mis-conceptions presented.
>  
>  > If you want Red Hat to listen in relationship to RHEL, try 
>  > http://redhat.com/support.
>  
>  Been there -- a good 4 years.
>  
>  > For Fedora, here is how you contribute. Note that the amount of 
>  > oppurtunities  is much more than RHEL
>  
>  Yes, I know.
>  
>  > * File bug reports
>  > * File feature requests with good rationale. I would like to have more 
>  > than 8 TB filesystems for such and such scenarios might be better than 
>  > "I want XFS"
>  
>  You didn't hear a word I said.  That is so sad.
>  
>  > * Write code and send in patches
>  
>  I'll get farther with SGI than I will Red Hat.  I'm asking -- at the
>  Fedora level -- to try to get Red Hat involved.  I've been trying
>  various avenues for 4 years now to no avail.
>  
>  And it's only getting more pertinent as Solaris is now on Opteron.
>  
>  > * Write documentation
>  
>  Way, way -- WAY AHEAD -- of you there.
>  
>  > * Translate UI and docs
>  > * Package in Fedora Extras
>  
>  I'm actually looking to submit several Fedora Extras packages.
>  
>  > * Art work, themes and sound themes
>  > * Advocate
>  
>  I'm well known as a "Red Hat apologist" (even though I hate the title).
>  So I find it humorous that you would suggest that to me.  ;->
>  
>  > * Last but not the least, use it
>  
>  Considering I've been integrating Fedora Core in companies as large as
>  one major Fortune 20 company in just the last 2 years, and several other
>  Fortune 100 companies as well -- again, WAY AHEAD OF YOU THERE.
>  
>  > There might be a few I have missed
>  
>  Again, I'm not some "I want filesystem X" puke.
>  But I know you'll feel free to take my comments as such.
>  
>  
>  -- 
>  Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
>  ----------------------------------------------------------------------
>  The best things in life are NOT free - which is why life is easiest if
>  you save all the bills until you can share them with the perfect woman
>  
>  --
>  Fedora-marketing-list mailing list
>  Fedora-marketing-list(a)redhat.com
>  http://www.redhat.com/mailman/listinfo/fedora-marketing-list
>

[Bryan Smith]

On Sun, 2005-09-11 at 23:49 -0400, Greg DeKoenigsberg wrote:
>  Bryan, I get your point. 
Thanx.  Remember, I'm an "outsider."  I do a lot of marketing of Fedora
Core, and I've always stuck right by Red Hat's decisions because of all
the trademark mess (no good deed goes unpunished, and that's exactly
what Red Hat's goodwill has always been taken as by so many companies).

>  Lemme talk to some smart folks and see what the general feeling is
>  around XFS "inside the walls." 
I'm sure the big issue is the adoption of 4KiB stacks, something I agree
very much with Red Hat on, but I know SGI has had issues.

>  I'm kinda curious myself now. 
Again, I'll put together something for the wiki by Tuesday, possibly
tomorrow night.  I just started with a new company this week -- and
we're deep into hurricane deployments of our technology for
communications.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Hi

> I honestly don't know where else to turn.  And when I read the
> rather
> "say nothing" portion that just feeds the ReiserFS and JFS proponents
> their "Red Hat is unfairly biased towards Ext3" non-sense, I have to
> agree with them from the XFS mis-conceptions presented.
>   
>  
If you are willing to take up the initiative to get other filesystem 
supported then send in a mail to fedora-devel with your detailed 
proposal. Red Hat is certainly biased towards ext3 since Red Hat has the 
necessary expertise to support it and enhance it as required as opposed 
to other filesystems. The only difference in relationship with Fedora is 
that the community can participate and take over maintenance other 
filesystems and things that Red Hat isnt able to allocate resources 
towards. I dont see Fedora myths page talking about feature comparisons 
for precisely this reason

regards
Rahul

[Bryan Smith]

On Mon, 2005-09-12 at 09:20 +0530, Rahul Sundaram wrote:
>  If you are willing to take up the initiative to get other filesystem
>
>  supported then send in a mail to fedora-devel with your detailed 
>  proposal. 
I'll discuss this over on the XFS list.  I need to get a better feel on
what they believe would be required involvement on the Fedora end.  But
I know it's going to bring up some old history and/or bad tastes.

>  Red Hat is certainly biased towards ext3 since Red Hat has the 
>  necessary expertise to support it and enhance it as required as opposed 
>  to other filesystems. 
And I'm telling you until Red Hat puts forth the interest to see XFS as
"the complement" to Ext3 it really needs, then it matters really little.
Which is why I've been trying to see Red Hat get serious about its
future by putting forth the effort to integrate XFS formally.

Especially when there were 2-4 years ago, as there are even some still
now, various things that XFS supports that Ext3 does not.  And no, I'm
not talking about "speed" or other "performance" non-sense.  I'm
talking
about enterprise features -- like on-line user-space tools, etc...

>  The only difference in relationship with Fedora is 
>  that the community can participate and take over maintenance other 
>  filesystems and things that Red Hat isnt able to allocate resources 
>  towards. 
Then _that's_ what needs to be said -- that Red Hat isn't able to
allocate resources towards a 2nd filesystem.  I'd more than accept that.

But anyone who knows anything about the histories of both Ext3 and XFS
on Linux -- especially those of us who have deployed both over the years
-- do _not_ accept the oblivious "XFS doesn't do anything Ext3 can" non-
sense.

It's an argument that _is_ valid against ReiserFS and JFS for Fedora/Red
Hat's focus that also gets, quite inaccurately, applied against XFS.

>  I dont see Fedora myths page talking about feature comparisons 
>  for precisely this reason 
It's not about "features."

It's about things like lack of EA support which makes it impossible for
ReiserFS and JFS to support SELinux, various history of kernel NFS,
quote and other "standard" kernel interface/support issues with ReiserFS
and JFS, etc...  Things that are _core_ to Fedora Core/Red Hat releases
that are "show stoppers."

But things that are _not_ true in the case of XFS.


-- 
Bryan J. Smith     b.j.smith(a)ieee.org     http://thebs413.blogspot.com
----------------------------------------------------------------------
The best things in life are NOT free - which is why life is easiest if
you save all the bills until you can share them with the perfect woman

[Rahul Sundaram]

Hi

> Then _that's_ what needs to be said -- that Red Hat isn't able
> to
> allocate resources towards a 2nd filesystem.  I'd more than accept that.
>   
>  
As far as I know no formal statements that ext3 supports all features 
that other filesystems do have been made by Red Hat. What has been 
repeatedly pointed out is the question of resources vs gain and requests 
in specific features that customers want based on their scenarios. If 
you do send in such a list to Red Hat support or partners , due 
consideration will given to those. For Fedora, get a detailed realistic 
plan to the fedora development list. Rest of the discussion doesnt 
really matter

regards
Rahul

[Rahul Sundaram]

Bryan J. Smith wrote:

> On Mon, 2005-09-12 at 06:26 +0530, Rahul Sundaram wrote:
>   
>
> >Your personal choices are not a list concern
> >    
> >
>
> That's fine.  Of course, what you call my "personal choices" is just one
> of the _many_ voices.  If Fedora and Red Hat want to ignore these
> voices, then they make their own issues.
>
> Again, I'm not some garden variety punk who thinks Fedora should have
> this, Fedora should have that.  I'm telling you not only what people who
> believe in Fedora Core, like Red Hat Linux before it, are feeling, but
> what others are picking up on.
>   
>  
You have mentioned the 1 TB limit on how you would setup your systems. 
However since this isnt  a technical limitation of the filesystem it 
doesnt make a case for other filesystems. It merely states your opinion 
on what you consider the limit which is entirely subjective..

regards
Rahul

[Christopher Aillon]

On 09/10/2005 09:37 AM, Colin Charles wrote:

> I'd like to discuss:
>
> FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths
> How can we use this further, and get it somewhat Red Hat blessed if we
> need to
>  
Drive-by feedback:
I see a whole bunch of negative items in bold, and then the word FACT in 
bold right underneath it.  Someone quickly skimming will get the wrong 
idea.  Highlight the positive as a contra to the negative.  Don't 
highlight the negative assuming people will read the rebuttal.

[Patrick W. Barnes]

Christopher Aillon wrote:

>  On 09/10/2005 09:37 AM, Colin Charles wrote:
>
> > I'd like to discuss:
> >
> > FedoraMyths - http://fedoraproject.org/wiki/FedoraMyths
> > How can we use this further, and get it somewhat Red Hat blessed if we
> > need to
> >
>  Drive-by feedback:
>  I see a whole bunch of negative items in bold, and then the word FACT
>  in bold right underneath it.  Someone quickly skimming will get the
>  wrong idea.  Highlight the positive as a contra to the negative. 
>  Don't highlight the negative assuming people will read the rebuttal.
>  
The current appearance of the document is largely dictated by the
perspective and the form of the wiki.  The 'myths' are the focus of the
document.  Dispelling them with the facts is its purpose.  It is not
meant to be read by drive-by reviewers.  People who view the page are
looking for reasons.  That said, I will try to go back and improve the
contrast to better highlight the facts.

-- 
Patrick "The N-Man" Barnes
nman64(a)n-man.com

www.n-man.com
--

[Christopher Aillon]

On 09/11/2005 02:07 AM, Patrick Barnes wrote:

> The current appearance of the document is largely dictated by the
> perspective and the form of the wiki.  The 'myths' are the focus of the
> document.  Dispelling them with the facts is its purpose.  It is not
> meant to be read by drive-by reviewers.  People who view the page are
> looking for reasons.  That said, I will try to go back and improve the
> contrast to better highlight the facts.
>  
The point is that your document is capable of being read drive-by, 
thanks to the nature of the internet (and e.g. Google).  The people who 
are susceptible to myth and most need to read that document are also the 
most likely to be confused with the wrong message if they happen to come 
across it the same way I did.

Disregarding that, there is still something even AFTER i read the 
document through that makes the human (at least this one) focus on the 
bullet points.  Those being negative really hurts.  Let's focus on the 
positive.  Perhaps something like:

* <b>Q: Is Fedora Core unstable and unreliable?</b>
* <b>A: No.  In fact, many businesses rely on Fedora Core for day-to-day 
operation and, in some cases, critical infrastructure.</b>  The 
misconception that Fedora is unstable is driven by two things:

Basically, give the reader the bottom line information up front, and 
then if they want to read the full response, they can.  This helps 
ensure the information we provide will be construed in a manner 
desirable to us.