Ok Iam confused with the roll back issue. What is the Edir that is working
Iam hearing Novell rolled back 8.3.6 but I can only see that 8.7.3.4 is
stable but Iam not hearing if 8.7.3.5 is ok. Iam on 8.7.3.5 and Iam having
issue with keeping edir healthy. So is this my issue and or how do you roll
back if I need to. Also I see some said " dsrapir full attened is not the
preferred way of getting ds healthy so what is this about I have not heard
of this before.
Thanks to anyone who can clear this up while I wait for ssupport to get
back to me on the phone.

Hi,

8735 came with NW65sp3 and should be OK
8734 is the highest version available for servers that are not on sp3
8736 has a specific issue that could possibly corrupt your NDS after doing a 
Full Unattended.
http://support.novell.com/cgi-bin/search/searchtid.cgi?/10098331.htm

Peter

> Ok Iam confused with the roll back issue. What is the Edir that is working
> Iam hearing Novell rolled back 8.3.6 but I can only see that 8.7.3.4 is
> stable but Iam not hearing if 8.7.3.5 is ok. Iam on 8.7.3.5 and Iam having
> issue with keeping edir healthy. So is this my issue and or how do you 
> roll
> back if I need to. Also I see some said " dsrapir full attened is not the
> preferred way of getting ds healthy so what is this about I have not heard
> of this before.
> Thanks to anyone who can clear this up while I wait for ssupport to get
> back to me on the phone.
> 

On Tue, 16 Aug 2005 16:38:45 +0000, aldullum wrote:

> Ok Iam confused with the roll back issue. What is the Edir that is working

8.7.3.6 is fine for most customers (all but the largest shouldn't be
affected by the issue that 8.7.3.6 was pulled for).  It was pulled because
there is a slight risk of data corruption should all the right conditions
occur.  Even though the odds are extremely small, it was decided that it
wasn't worth the risk, so the patch was pulled.  I have heard of only two
customers who have actually seen the issue (in their test lab
environments as I understand it - not on production systems), and both of
them have object counts that are fairly high.

8.7.3.5 and earlier do not have the issue that was found in 8.7.3.6; there
was a change in the memory management system that had an unintended side
effect.

SP7 is due out very soon now (at a guess, within 2 weeks, but probably
sooner than that).

> Iam hearing Novell rolled back 8.3.6 but I can only see that 8.7.3.4 is
> stable but Iam not hearing if 8.7.3.5 is ok. Iam on 8.7.3.5 and Iam
> having issue with keeping edir healthy. So is this my issue and or how
> do you roll back if I need to. 

See the announcement in this group for a test to determine if your IR6
servers have the issue on this (I believe Peter vdW linked to the TID in
question).  Chances are that unless you have hundreds of thousands
to millions of objects, you probably are unaffected.

> Also I see some said " dsrapir full
> attened is not the preferred way of getting ds healthy so what is this
> about I have not heard of this before. 

DSREPAIR is a tool to fix the directory; diagnose first, find what issues
are there, and then fix them.  Too many customers get themselves into
trouble by running DSREPAIR without understanding what it is doing or
understanding about what the problem is.  DSREPAIR is not an
administration tool, and it is not the best diagnostic tool.  It is a
"fix" tool, and should only be used that way when it is relatively certain
that it is the right thing to do.

Most people do not understand how (or at what level) DSREPAIR operates -
it operates at the database record level.  Because it operates at such a
low level, if it is used inappropriately, you can cause damage to your
database.

See http://www.novell.com/coolsolutions/feature/15312.html for information
about how to properly use DSREPAIR.

Jim
-- 
 Jim Henderson, CNA6, CDE, CNI, LPIC-1
 eDirectory/Identity Manager Product Specialist, Novell, Inc.
 Homepage at http://hendersj.dyndns.org
 The opinions expressed above are not necessarily those of Novell, Inc.

> See http://www.novell.com/coolsolutions/feature/15312.html for information
> about how to properly use DSREPAIR.
>
Hi Jim,

Good reading material. Bookmarked.
I *always* load dsrepair with option -A  :-)
Do you actually know someone who used it to promote a subordinate to Master 
?

Strange how opinion on this changed over time. I remember when studying 
directory engineer the advice was: Always do a Full Unattended between each 
fix....

Peter 

On Tue, 16 Aug 2005 18:34:54 +0000, Peter vdW wrote:

> I *always* load dsrepair with option -A  :-) Do you actually know someone
> who used it to promote a subordinate to Master ?

I've had to, actually - in a production environment even.  We did it as
part of a tree split operation, forgot to put a replica of the part of the
tree that was "leaving" (pun not intended) on our real servers, and we
couldn't merge the partition back into the parent.  Since we didn't care
about the subordinate objects (and it wouldn't have mattered if we did
because the sale was done and network disconnected), that was the only way
to clean it up.

> Strange how opinion on this changed over time. I remember when studying
> directory engineer the advice was: Always do a Full Unattended between
> each fix....

I remember those types of recommendations; also things like run an
unattended full (<shudder>) once a month or so.  The thinking has really
changed to using it for fixing only, especially since there are much
better diagnostic tools (iMonitor, dstrace is better documented, etc.)
that the old "run repair and see what it fixes" tactic is no longer
necessary.  Also with the sizes of DIBs these days, running repair except
when it's necessary is impractical.  I know some customers who have DIBs
so big that when they lock the DIB for a repair that requires it, they can
expect the server to be unavailable for many hours.

Jim
-- 
 Jim Henderson, CNA6, CDE, CNI, LPIC-1
 eDirectory/Identity Manager Product Specialist, Novell, Inc.
 Homepage at http://hendersj.dyndns.org
 The opinions expressed above are not necessarily those of Novell, Inc.

Hi Jim,

> I've had to, actually - in a production environment even.  We did it as
> part of a tree split operation, forgot to put a replica of the part of the
> tree that was "leaving" (pun not intended) on our real servers, and we
> couldn't merge the partition back into the parent.  Since we didn't care
> about the subordinate objects (and it wouldn't have mattered if we did
> because the sale was done and network disconnected), that was the only way
> to clean it up.

In the words of the famous dutch philosopher Johan Cruijff:
Every disadvantage has an advantage  :-)

> necessary.  Also with the sizes of DIBs these days, running repair except
> when it's necessary is impractical.  I know some customers who have DIBs
> so big that when they lock the DIB for a repair that requires it, they can
> expect the server to be unavailable for many hours.
>
Tell me about it! Our Nw51 tree dib was 500 MB (90% zfd)  and took 1 1/2 
hours to repair. I used to schedule it at 6 am sunday morning and still got 
complaints from users using Netstorage. And they call *me* a nerd.....

Peter 

On Tue, 16 Aug 2005 18:34:54 GMT, "Peter vdW"
<p.c.vanderwoudeREMOVE@THISsaxion.nl> wrote:

>Do you actually know someone who used it to promote a subordinate to Master 
>?

I've done this, but only in a lab environment. It is the only way to recover
when all M and R/W replicas of a partition have been destroyed and only the
SubRef remains. You can promote it to Master, merge it back in to its parent,
and delete the objects and the container that used to be the partition root.


---------------------------------------------------------------------------
 David Gersic                                            dgersic_@_niu.edu

 I'm tired of receiving rubbish in my mailbox, so the E-mail address is
 munged to foil the junkmail bots. Humans will figure it out on their own.

> I've done this, but only in a lab environment. It is the only way to 
> recover
> when all M and R/W replicas of a partition have been destroyed and only 
> the
> SubRef remains. You can promote it to Master, merge it back in to its 
> parent,
> and delete the objects and the container that used to be the partition 
> root.

David,

Filed and bookmarked...
Like I said to Jim:
 "Every disadvantage has its advantage"    :-)
http://www.answers.com/topic/johan-cruijff

Peter 

On Tue, 16 Aug 2005 20:36:01 +0000, Peter vdW wrote:

> In the words of the famous dutch philosopher Johan Cruijff: Every
> disadvantage has an advantage  :-)

<g>

> Tell me about it! Our Nw51 tree dib was 500 MB (90% zfd)  and took 1 1/2
> hours to repair. I used to schedule it at 6 am sunday morning and still
> got complaints from users using Netstorage. And they call *me* a nerd.....

Reminds me of a Dilbert cartoon, the final panel of which is Dilbert
saying to Wally "There's a fine line between dedication and insanity.", to
which Wally replies "Which are we?".

(Let's continue this non-support discussion over in novell.community.chat
- OK?)

Jim

-- 
 Jim Henderson, CNA6, CDE, CNI, LPIC-1
 eDirectory/Identity Manager Product Specialist, Novell, Inc.
 Homepage at http://hendersj.dyndns.org
 The opinions expressed above are not necessarily those of Novell, Inc.

> (Let's continue this non-support discussion over in novell.community.chat
> - OK?)

Agreed.
Subject Dutch Yogi Bear  :-)

Peter 

--____QVBFUMQLUVWONYAZJZYR____
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

Jim,

If I have no issues with eDirectory 8.7.3.6, I shouldn't have any problems =
updating to 7 when it arrives, correct?

Brian


>>> Jim Henderson<Jim.Henderson@SysOps.NSC> 8/16/2005 12:46:17 PM >>>
On Tue, 16 Aug 2005 16:38:45 +0000, aldullum wrote:

> Ok Iam confused with the roll back issue. What is the Edir that is =
working

8.7.3.6 is fine for most customers (all but the largest shouldn't be
affected by the issue that 8.7.3.6 was pulled for).  It was pulled because
there is a slight risk of data corruption should all the right conditions
occur.  Even though the odds are extremely small, it was decided that it
wasn't worth the risk, so the patch was pulled.  I have heard of only two
customers who have actually seen the issue (in their test lab
environments as I understand it - not on production systems), and both of
them have object counts that are fairly high.

8.7.3.5 and earlier do not have the issue that was found in 8.7.3.6; there
was a change in the memory management system that had an unintended side
effect.

SP7 is due out very soon now (at a guess, within 2 weeks, but probably
sooner than that).

> Iam hearing Novell rolled back 8.3.6 but I can only see that 8.7.3.4 is
> stable but Iam not hearing if 8.7.3.5 is ok. Iam on 8.7.3.5 and Iam
> having issue with keeping edir healthy. So is this my issue and or how
> do you roll back if I need to.=20

See the announcement in this group for a test to determine if your IR6
servers have the issue on this (I believe Peter vdW linked to the TID in
question).  Chances are that unless you have hundreds of thousands
to millions of objects, you probably are unaffected.

> Also I see some said " dsrapir full
> attened is not the preferred way of getting ds healthy so what is this
> about I have not heard of this before.=20

DSREPAIR is a tool to fix the directory; diagnose first, find what issues
are there, and then fix them.  Too many customers get themselves into
trouble by running DSREPAIR without understanding what it is doing or
understanding about what the problem is.  DSREPAIR is not an
administration tool, and it is not the best diagnostic tool.  It is a
"fix" tool, and should only be used that way when it is relatively certain
that it is the right thing to do.

Most people do not understand how (or at what level) DSREPAIR operates -
it operates at the database record level.  Because it operates at such a
low level, if it is used inappropriately, you can cause damage to your
database.

See http://www.novell.com/coolsolutions/feature/15312.html for information
about how to properly use DSREPAIR.

Jim
--=20
Jim Henderson, CNA6, CDE, CNI, LPIC-1
eDirectory/Identity Manager Product Specialist, Novell, Inc.
Homepage at http://hendersj.dyndns.org
The opinions expressed above are not necessarily those of Novell, Inc.

--____QVBFUMQLUVWONYAZJZYR____
Content-Type: multipart/related; boundary="____VKXVBOPKZIUKSPPVJFSY____"


--____VKXVBOPKZIUKSPPVJFSY____
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Diso-8859-1"=
>
<META content=3D"MSHTML 6.00.2900.2722" name=3DGENERATOR></HEAD>
<BODY style=3D"MARGIN-TOP: 2px; FONT: 8pt Tahoma; MARGIN-LEFT: 2px">
<DIV><FONT size=3D2>Jim,</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>If I have no issues with eDirectory 8.7.3.6, I =
shouldn't have=20
any problems updating to 7 when it arrives, correct?</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>Brian</FONT></DIV>
<DIV><BR><BR>&gt;&gt;&gt; Jim Henderson&lt;Jim.Henderson@SysOps.NSC&gt;=20
8/16/2005 12:46:17 PM &gt;&gt;&gt;<BR>On Tue, 16 Aug 2005 16:38:45 =
+0000,=20
aldullum wrote:<BR><BR>&gt; Ok Iam confused with the roll back issue. What =
is=20
the Edir that is working<BR><BR>8.7.3.6 is fine for most customers (all =
but the=20
largest shouldn't be<BR>affected by the issue that 8.7.3.6 was pulled=20
for).&nbsp; It was pulled because<BR>there is a slight risk of data =
corruption=20
should all the right conditions<BR>occur.&nbsp; Even though the odds =
are=20
extremely small, it was decided that it<BR>wasn't worth the risk, so the =
patch=20
was pulled.&nbsp; I have heard of only two<BR>customers who have actually =
seen=20
the issue (in their test lab<BR>environments as I understand it - not =
on=20
production systems), and both of<BR>them have object counts that are =
fairly=20
high.<BR><BR>8.7.3.5 and earlier do not have the issue that was found =
in=20
8.7.3.6; there<BR>was a change in the memory management system that had =
an=20
unintended side<BR>effect.<BR><BR>SP7 is due out very soon now (at a =
guess,=20
within 2 weeks, but probably<BR>sooner than that).<BR><BR>&gt; Iam =
hearing=20
Novell rolled back 8.3.6 but I can only see that 8.7.3.4 is<BR>&gt; stable =
but=20
Iam not hearing if 8.7.3.5 is ok. Iam on 8.7.3.5 and Iam<BR>&gt; having =
issue=20
with keeping edir healthy. So is this my issue and or how<BR>&gt; do you =
roll=20
back if I need to. <BR><BR>See the announcement in this group for a test =
to=20
determine if your IR6<BR>servers have the issue on this (I believe Peter =
vdW=20
linked to the TID in<BR>question).&nbsp; Chances are that unless you =
have=20
hundreds of thousands<BR>to millions of objects, you probably are=20
unaffected.<BR><BR>&gt; Also I see some said " dsrapir full<BR>&gt; =
attened is=20
not the preferred way of getting ds healthy so what is this<BR>&gt; about =
I have=20
not heard of this before. <BR><BR>DSREPAIR is a tool to fix the directory;=
=20
diagnose first, find what issues<BR>are there, and then fix them.&nbsp; =
Too many=20
customers get themselves into<BR>trouble by running DSREPAIR without=20
understanding what it is doing or<BR>understanding about what the =
problem=20
is.&nbsp; DSREPAIR is not an<BR>administration tool, and it is not the =
best=20
diagnostic tool.&nbsp; It is a<BR>"fix" tool, and should only be used that =
way=20
when it is relatively certain<BR>that it is the right thing to do.<BR><BR>M=
ost=20
people do not understand how (or at what level) DSREPAIR operates =
-<BR>it=20
operates at the database record level.&nbsp; Because it operates at =
such=20
a<BR>low level, if it is used inappropriately, you can cause damage to=20
your<BR>database.<BR><BR>See <A=20
href=3D"http://www.novell.com/coolsolutions/feature/15312.html">http://www.=
novell.com/coolsolutions/feature/15312.html</A>=20
for information<BR>about how to properly use DSREPAIR.<BR><BR>Jim<BR>-- =
<BR>Jim=20
Henderson, CNA6, CDE, CNI, LPIC-1<BR>eDirectory/Identity Manager Product=20=

Specialist, Novell, Inc.<BR>Homepage at <A=20
href=3D"http://hendersj.dyndns.org/">http://hendersj.dyndns.org</A><BR>The=
=20
opinions expressed above are not necessarily those of Novell,=20
Inc.<BR><BR><BR><BR></DIV></BODY></HTML>

--____VKXVBOPKZIUKSPPVJFSY____--

--____QVBFUMQLUVWONYAZJZYR____--

On Thu, 25 Aug 2005 19:19:45 +0000, Brian Gibson wrote:

> Jim,
> 
> If I have no issues with eDirectory 8.7.3.6, I shouldn't have any problems
> updating to 7 when it arrives, correct?

As far as I know, that is correct.  I've done the upgrade to a beta build
from IR6 myself on Linux a few times, and it's worked fine.

Jim

-- 
 Jim Henderson, CNA6, CDE, CNI, LPIC-1
 eDirectory/Identity Manager Product Specialist, Novell, Inc.
 Homepage at http://hendersj.dyndns.org
 The opinions expressed above are not necessarily those of Novell, Inc.

--____NQZQZIQEPAYEKJTERHVW____
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

I noticed that the final BETA version of eDirectory 8.8 is out.  Can I go =
to that directly without purchasing additional licensing?

Brian

>>> Jim Henderson<Jim.Henderson@SysOps.NSC> 8/25/2005 9:29 PM >>>

On Thu, 25 Aug 2005 19:19:45 +0000, Brian Gibson wrote:

> Jim,
>=20
> If I have no issues with eDirectory 8.7.3.6, I shouldn't have any =
problems
> updating to 7 when it arrives, correct?

As far as I know, that is correct.  I've done the upgrade to a beta build
from IR6 myself on Linux a few times, and it's worked fine.

Jim

--=20
Jim Henderson, CNA6, CDE, CNI, LPIC-1
eDirectory/Identity Manager Product Specialist, Novell, Inc.
Homepage at http://hendersj.dyndns.org
The opinions expressed above are not necessarily those of Novell, Inc.

--____NQZQZIQEPAYEKJTERHVW____
Content-Type: multipart/related; boundary="____EWRCVRALNSKYXARJDMYV____"


--____EWRCVRALNSKYXARJDMYV____
Content-Type: text/html; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Diso-8859-1"=
>
<META content=3D"MSHTML 6.00.2900.2722" name=3DGENERATOR></HEAD>
<BODY style=3D"MARGIN: 4px 4px 1px; FONT: 10pt Tahoma">
<DIV>I noticed that the final BETA version of eDirectory 8.8 is out.&nbsp; =
Can I go to that directly without purchasing additional licensing?</DIV>
<DIV>&nbsp;</DIV>
<DIV>Brian<BR><BR>&gt;&gt;&gt; Jim Henderson&lt;Jim.Henderson@SysOps.NSC&gt=
; 8/25/2005 9:29 PM &gt;&gt;&gt;<BR></DIV>
<DIV style=3D"COLOR: #000000">On Thu, 25 Aug 2005 19:19:45 +0000, Brian =
Gibson wrote:<BR><BR>&gt; Jim,<BR>&gt; <BR>&gt; If I have no issues with =
eDirectory 8.7.3.6, I shouldn't have any problems<BR>&gt; updating to 7 =
when it arrives, correct?<BR><BR>As far as I know, that is correct.&nbsp; =
I've done the upgrade to a beta build<BR>from IR6 myself on Linux a few =
times, and it's worked fine.<BR><BR>Jim<BR><BR>-- <BR>Jim Henderson, CNA6, =
CDE, CNI, LPIC-1<BR>eDirectory/Identity Manager Product Specialist, =
Novell, Inc.<BR>Homepage at <A href=3D"http://hendersj.dyndns.org">http://h=
endersj.dyndns.org</A><BR>The opinions expressed above are not necessarily =
those of Novell, Inc.<BR><BR><BR><BR></DIV></BODY></HTML>

--____EWRCVRALNSKYXARJDMYV____--

--____NQZQZIQEPAYEKJTERHVW____--

eDir on NetWare never ever requires any licensing of its own.
-- 
  Andrew C Taubman
  Novell Support Forums Volunteer SysOp
  http://support.novell.com/forums
  (Sorry, support is not provided via e-mail)

  Opinions expressed above are not
  necessarily those of Novell Inc.