zfs discuss - Jul 2010 - NexentaStor 3.0.3 vs OpenSolaris - Patches more up to date?

I see in NexentaStor''s announcement of Community Edition 3.0.3 they
mention some backported patches in this release.

Aside from their management features / UI what is the core OS difference if we
move to Nexenta from OpenSolaris b134?

These DeDup bugs are my main frustration - if a staff member does a rm * in a
directory with dedup you can take down the whole storage server - all with 1%
cpu load and relatively little disk i/o due to DeDup DDT not fitting in the SSD
+ RAM (l2arc+arc). This is rediculous, something must be single threaded and it
can''t be that difficult to at least allow reads from other files..
Writes perhaps are more complex - But in our case the "other files"
don''t even have DeDup enabled on them and they can''t be read.

It seems like some of these bugs have been fixed but Oracle hasn''t
published a new build - Perhaps we should be updating to newer builds, I
haven''t invested much time in seeking these out but b134 is the latest
"obvious" build I see. Am I just not RTFM enough on finding new
builds?

I hate to move to Nexenta, I would think in the future Oracle will maintain this
better than a third party and don''t want to switch back and forth.

Steve Radich - www.BitShop.com - www.LinkedIn.com/in/SteveRadich
-- 
This message posted from opensolaris.org

On Jul 2, 2010, at 12:53 PM, Steve Radich, BitShop, Inc. wrote:
> I see in NexentaStor''s announcement of Community Edition 3.0.3
they mention some backported patches in this release.
Yes.  These patches are in the code tree, currently at b143 (~18 weeks
newer than b134)
> Aside from their management features / UI what is the core OS difference if
we move to Nexenta from OpenSolaris b134?
You''re not stuck at b134 for ZFS anymore ;-)
> These DeDup bugs are my main frustration - if a staff member does a rm * in
a directory with dedup you can take down the whole storage server - all with 1%
cpu load and relatively little disk i/o due to DeDup DDT not fitting in the SSD
+ RAM (l2arc+arc). This is rediculous, something must be single threaded and it
can''t be that difficult to at least allow reads from other files..
Writes perhaps are more complex - But in our case the "other files"
don''t even have DeDup enabled on them and they can''t be read.
Some are fixed, more are in the upstream development queue.
> It seems like some of these bugs have been fixed but Oracle hasn''t
published a new build - Perhaps we should be updating to newer builds, I
haven''t invested much time in seeking these out but b134 is the latest
"obvious" build I see. Am I just not RTFM enough on finding new
builds?
No, what you see is what you get.  After the CIC there hasn''t been a 
binary release from Oracle, just source releases.  I read this as saying
the community should build their own distros. In a quick look at 
http://www.genunix.org it appears that Nexenta and EON are the only
distro releases since early March.  Rich Lowe has released a b142
tarball, too, but does that qualify as a distro?
> I hate to move to Nexenta, I would think in the future Oracle will maintain
this better than a third party and don''t want to switch back and forth.

I understand, but if actions speak louder than words, then consider joining
the Nexenta core platform community at http://www.nexenta.org  
But don''t forget to stay up to date with ZFS on zfs-discuss :-)
 -- richard

-- 
Richard Elling
richard at nexenta.com   +1-760-896-4422
ZFS and NexentaStor training, Rotterdam, July 13-15, 2010
http://nexenta-rotterdam.eventbrite.com/

On Fri, Jul 2, 2010 at 8:06 PM, Richard Elling <richard at nexenta.com>
wrote:
> On Jul 2, 2010, at 12:53 PM, Steve Radich, BitShop, Inc. wrote:
>
> > I see in NexentaStor''s announcement of Community Edition
3.0.3 they
> mention some backported patches in this release.
>
> Yes.  These patches are in the code tree, currently at b143 (~18 weeks
> newer than b134)
>
> > Aside from their management features / UI what is the core OS
difference
> if we move to Nexenta from OpenSolaris b134?
>
> You''re not stuck at b134 for ZFS anymore ;-)
>
> > These DeDup bugs are my main frustration - if a staff member does a rm
*
> in a directory with dedup you can take down the whole storage server - all
> with 1% cpu load and relatively little disk i/o due to DeDup DDT not
fitting
> in the SSD + RAM (l2arc+arc). This is rediculous, something must be single
> threaded and it can''t be that difficult to at least allow reads
from other
> files.. Writes perhaps are more complex - But in our case the "other
files"
> don''t even have DeDup enabled on them and they can''t be
read.
>
> Some are fixed, more are in the upstream development queue.
>
> > It seems like some of these bugs have been fixed but Oracle
hasn''t
> published a new build - Perhaps we should be updating to newer builds, I
> haven''t invested much time in seeking these out but b134 is the
latest
> "obvious" build I see. Am I just not RTFM enough on finding new
builds?
>
> No, what you see is what you get.  After the CIC there hasn''t been
a
> binary release from Oracle, just source releases.  I read this as saying
> the community should build their own distros. In a quick look at
> http://www.genunix.org it appears that Nexenta and EON are the only
> distro releases since early March.  Rich Lowe has released a b142
> tarball, too, but does that qualify as a distro?
>
> > I hate to move to Nexenta, I would think in the future Oracle will
> maintain this better than a third party and don''t want to switch
back and
> forth.
>
>
> I understand, but if actions speak louder than words, then consider joining
> the Nexenta core platform community at http://www.nexenta.org
> But don''t forget to stay up to date with ZFS on zfs-discuss :-)
>  -- richard
>
> --
> Richard Elling
> richard at nexenta.com   +1-760-896-4422
> ZFS and NexentaStor training, Rotterdam, July 13-15, 2010
> http://nexenta-rotterdam.eventbrite.com/


Given that the most basic of functionality was broken in Nexenta, and not
Opensolaris, and I couldn''t get a single response, I have a hard time
recommending ANYONE go to Nexenta.  It''s great they''re
employing you now,
but the community edition has an extremely long way to go before it comes
close to touching the community that still hangs around here, despite
Oracle''s lack of care and feeding.

http://www.nexenta.org/boards/1/topics/211


--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20100702/8fcecfaf/attachment.html>

On Jul 2, 2010, at 6:48 PM, Tim Cook wrote:
> Given that the most basic of functionality was broken in Nexenta, and not
Opensolaris, and I couldn''t get a single response, I have a hard time
recommending ANYONE go to Nexenta.  It''s great they''re
employing you now, but the community edition has an extremely long way to go
before it comes close to touching the community that still hangs around here,
despite Oracle''s lack of care and feeding.
> 
> http://www.nexenta.org/boards/1/topics/211
I can''t test that, due to lack of equivalent hardware, but did you file
a bug?
The dladm code and nge drivers come from upstream, so look for an equivalent
opensolaris bug,  perhaps something like
http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6913874
 -- richard

-- 
Richard Elling
richard at nexenta.com   +1-760-896-4422
ZFS and NexentaStor training, Rotterdam, July 13-15, 2010
http://nexenta-rotterdam.eventbrite.com/

On Fri, Jul 2, 2010 at 9:25 PM, Richard Elling <richard at nexenta.com>
wrote:
> On Jul 2, 2010, at 6:48 PM, Tim Cook wrote:
> > Given that the most basic of functionality was broken in Nexenta, and
not
> Opensolaris, and I couldn''t get a single response, I have a hard
time
> recommending ANYONE go to Nexenta.  It''s great they''re
employing you now,
> but the community edition has an extremely long way to go before it comes
> close to touching the community that still hangs around here, despite
> Oracle''s lack of care and feeding.
> >
> > http://www.nexenta.org/boards/1/topics/211
>
> I can''t test that, due to lack of equivalent hardware, but did you
file a
> bug?
> The dladm code and nge drivers come from upstream, so look for an
> equivalent
> opensolaris bug,  perhaps something like
> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6913874
>  -- richard
>
>
No, I didn''t file a bug.  I couldn''t get a response to the
issue to even
begin troubleshooting, so I had no desire to file a bug or continue using a
product that was broken out of the box.  Opensolaris worked, so I went back
to it.  I can say with a fair amount of confidence, the same wouldn''t
happen
with Opensolaris proper.  Even if I chose not to continue running down a
problem, I''ve never run into a situation where I didn''t at
least get a
suggestion for troubleshooting tips.

--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20100702/e3d95600/attachment.html>

On  3/07/10 12:25 PM, Richard Elling wrote:
> On Jul 2, 2010, at 6:48 PM, Tim Cook wrote:
>> Given that the most basic of functionality was broken in Nexenta, and
not Opensolaris, and I couldn''t get a single response, I have a hard
time recommending ANYONE go to Nexenta.  It''s great they''re
employing you now, but the community edition has an extremely long way to go
before it comes close to touching the community that still hangs around here,
despite Oracle''s lack of care and feeding.
>>
>> http://www.nexenta.org/boards/1/topics/211
>
> I can''t test that, due to lack of equivalent hardware, but did you
file a bug?
> The dladm code and nge drivers come from upstream, so look for an
equivalent
> opensolaris bug,  perhaps something like
> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6913874
>   -- richard
>

Hi Tim,
does this CR match what you were experiencing?

http://bugs.opensolaris.org/view_bug.do?bug_id=6901419
6901419 dladm create-aggr -u incorrectly rejects some valid ethernet 
addresses

If so - fixed in snv_136.

The only other dladm CR I can see in the push logs for builds
post 134 is

6932656 "dladm set-linkprop -p cpus" can''t take more than 32
CPUs
fixed in 138.


hth,
James
--
Senior Software Engineer, Solaris
Oracle
http://www.jmcp.homeunix.com/blog

On Fri, Jul 2, 2010 at 9:55 PM, James C. McPherson <jmcp at
opensolaris.org>wrote:
> On  3/07/10 12:25 PM, Richard Elling wrote:
>
>> On Jul 2, 2010, at 6:48 PM, Tim Cook wrote:
>>
>>> Given that the most basic of functionality was broken in Nexenta,
and not
>>> Opensolaris, and I couldn''t get a single response, I have
a hard time
>>> recommending ANYONE go to Nexenta.  It''s great
they''re employing you now,
>>> but the community edition has an extremely long way to go before it
comes
>>> close to touching the community that still hangs around here,
despite
>>> Oracle''s lack of care and feeding.
>>>
>>> http://www.nexenta.org/boards/1/topics/211
>>>
>>
>> I can''t test that, due to lack of equivalent hardware, but did
you file a
>> bug?
>> The dladm code and nge drivers come from upstream, so look for an
>> equivalent
>> opensolaris bug,  perhaps something like
>> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6913874
>>  -- richard
>>
>>
>
> Hi Tim,
> does this CR match what you were experiencing?
>
> http://bugs.opensolaris.org/view_bug.do?bug_id=6901419
> 6901419 dladm create-aggr -u incorrectly rejects some valid ethernet
> addresses
>
> If so - fixed in snv_136.
>
> The only other dladm CR I can see in the push logs for builds
> post 134 is
>
> 6932656 "dladm set-linkprop -p cpus" can''t take more
than 32 CPUs
> fixed in 138.
>
>
> hth,
> James
> --
> Senior Software Engineer, Solaris
> Oracle
> http://www.jmcp.homeunix.com/blog

Hi james,

Nope.  I''m not sure what exactly I was hitting.  I''ve never
run into a
problem on any release of Opensolaris.  I believe I''ve tested on 126,
132,
133, and 134 (as well as many iterations of older versions).  The dladm
issue was exclusively on nexenta.

--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20100702/348b1c5a/attachment.html>

I am sorry you feel that way.  I will look at your issue as soon as I am able,
but I should say that it is almost certain that whatever the problem is, it
probably is inherited from OpenSolaris and the build of NCP you were testing was
indeed not the final release so some issues are not entirely surprising.

  - Garrett

Tim Cook <tim at cook.ms> wrote:
>On Fri, Jul 2, 2010 at 8:06 PM, Richard Elling <richard at
nexenta.com> wrote:
>
>> On Jul 2, 2010, at 12:53 PM, Steve Radich, BitShop, Inc. wrote:
>>
>> > I see in NexentaStor''s announcement of Community Edition
3.0.3 they
>> mention some backported patches in this release.
>>
>> Yes.  These patches are in the code tree, currently at b143 (~18 weeks
>> newer than b134)
>>
>> > Aside from their management features / UI what is the core OS
difference
>> if we move to Nexenta from OpenSolaris b134?
>>
>> You''re not stuck at b134 for ZFS anymore ;-)
>>
>> > These DeDup bugs are my main frustration - if a staff member does
a rm *
>> in a directory with dedup you can take down the whole storage server -
all
>> with 1% cpu load and relatively little disk i/o due to DeDup DDT not
fitting
>> in the SSD + RAM (l2arc+arc). This is rediculous, something must be
single
>> threaded and it can''t be that difficult to at least allow
reads from other
>> files.. Writes perhaps are more complex - But in our case the
"other files"
>> don''t even have DeDup enabled on them and they can''t
be read.
>>
>> Some are fixed, more are in the upstream development queue.
>>
>> > It seems like some of these bugs have been fixed but Oracle
hasn''t
>> published a new build - Perhaps we should be updating to newer builds,
I
>> haven''t invested much time in seeking these out but b134 is
the latest
>> "obvious" build I see. Am I just not RTFM enough on finding
new builds?
>>
>> No, what you see is what you get.  After the CIC there hasn''t
been a
>> binary release from Oracle, just source releases.  I read this as
saying
>> the community should build their own distros. In a quick look at
>> http://www.genunix.org it appears that Nexenta and EON are the only
>> distro releases since early March.  Rich Lowe has released a b142
>> tarball, too, but does that qualify as a distro?
>>
>> > I hate to move to Nexenta, I would think in the future Oracle will
>> maintain this better than a third party and don''t want to
switch back and
>> forth.
>>
>>
>> I understand, but if actions speak louder than words, then consider
joining
>> the Nexenta core platform community at http://www.nexenta.org
>> But don''t forget to stay up to date with ZFS on zfs-discuss
:-)
>>  -- richard
>>
>> --
>> Richard Elling
>> richard at nexenta.com   +1-760-896-4422
>> ZFS and NexentaStor training, Rotterdam, July 13-15, 2010
>> http://nexenta-rotterdam.eventbrite.com/
>
>
>
>Given that the most basic of functionality was broken in Nexenta, and not
>Opensolaris, and I couldn''t get a single response, I have a hard
time
>recommending ANYONE go to Nexenta.  It''s great they''re
employing you now,
>but the community edition has an extremely long way to go before it comes
>close to touching the community that still hangs around here, despite
>Oracle''s lack of care and feeding.
>
>http://www.nexenta.org/boards/1/topics/211
>
>
>--Tim
>
>_______________________________________________
>zfs-discuss mailing list
>zfs-discuss at opensolaris.org
>http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

On Sun, Jul 4, 2010 at 12:22 AM, Garrett D''Amore <garrett at
nexenta.com> wrote:
> I am sorry you feel that way. ?I will look at your issue as soon as I am
able, but I should say that it is almost certain that whatever the problem is,
it probably is inherited from OpenSolaris and the build of NCP you were testing
was indeed not the final release so some issues are not entirely surprising.
So would you say that NCP / NexentaStore Community 3.0.3  is good
enough to use today as stand-in replacement for last available build
of Opensolaris when used primarily for storage server?

-- 
Fajar

Compared to b134?  Yes!  We have fixed many bugs that still exist in 134.

"Fajar A. Nugraha" <fajar at fajar.net> wrote:
>On Sun, Jul 4, 2010 at 12:22 AM, Garrett D''Amore <garrett at
nexenta.com> wrote:
>> I am sorry you feel that way. ?I will look at your issue as soon as I
am able, but I should say that it is almost certain that whatever the problem
is, it probably is inherited from OpenSolaris and the build of NCP you were
testing was indeed not the final release so some issues are not entirely
surprising.
>
>So would you say that NCP / NexentaStore Community 3.0.3  is good
>enough to use today as stand-in replacement for last available build
>of Opensolaris when used primarily for storage server?
>
>-- 
>Fajar
>

----- Original Message -----
> Compared to b134? Yes! We have fixed many bugs that still exist in
> 134.
Where can I find a list of these?

Vennlige hilsener / Best regards

roy
--
Roy Sigurd Karlsbakk
(+47) 97542685
roy at karlsbakk.net
http://blogg.karlsbakk.net/
--
I all pedagogikk er det essensielt at pensum presenteres intelligibelt. Det er
et element?rt imperativ for alle pedagoger ? unng? eksessiv anvendelse av
idiomer med fremmed opprinnelse. I de fleste tilfeller eksisterer adekvate og
relevante synonymer p? norsk.

> Where can I find a list of these?
This leads to the more generic question of: where are *any* release notes?

I saw on Genunix that Community Edition 3.0.3 was replaced by 3.0.3-1. What
changed? I  went to nexenta.org and looked around. But it wasn''t
immediately obvious where to find release notes. Also, as Tim Cook noted, the
Nexenta forums aren''t exactly "lively".

For a simple, easily understood and easily navigated web site, you
can''t beat www.openbsd.org. Both Sun/Oracle and Nexenta could learn a
lot from it. And I can also follow very clean, simple instructions for running
the "stable" OpenBSD branch (which is mostly security fixes).
-- 
This message posted from opensolaris.org

In 3.0.3+ new option would list appliance changelog going forward:

nmc$ show version -c

On 07/04/2010 05:58 PM, Bohdan Tashchuk wrote:
>> Where can I find a list of these?
>
> This leads to the more generic question of: where are *any* release notes?
>
> I saw on Genunix that Community Edition 3.0.3 was replaced by 3.0.3-1. What
changed? I  went to nexenta.org and looked around. But it wasn''t
immediately obvious where to find release notes. Also, as Tim Cook noted, the
Nexenta forums aren''t exactly "lively".
>
> For a simple, easily understood and easily navigated web site, you
can''t beat www.openbsd.org. Both Sun/Oracle and Nexenta could learn a
lot from it. And I can also follow very clean, simple instructions for running
the "stable" OpenBSD branch (which is mostly security fixes).

Release Notes information:

If there are new features, each release is added to
http://www.nexenta.com/corp/documentation/release-notes-support.

If just bug fixes, then the Changelog listing is updated:
http://www.nexenta.com/corp/documentation/nexentastor-changelog

Regards,
Spandana
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20100706/9d052b27/attachment.html>

On Tue, Jul 6, 2010 at 4:06 PM, Spandana Goli <spandana at nexenta.com>
wrote:
> Release Notes information:
> If there are new features, each release is added to
> http://www.nexenta.com/corp/documentation/release-notes-support.
>
> If just bug fixes, then the Changelog listing is updated:
> http://www.nexenta.com/corp/documentation/nexentastor-changelog
Is there a bug tracker were one can objectively list all the bugs
(with details) that went into a release ?

"Many bug fixes" is a bit too general.

-- 
Giovanni Tirloni
gtirloni at sysdroid.com