headers
Alar Pandis | 1 Mar 2009 11:40
Picon
Picon

RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact every 24 hours

Hi and thanks again Randy!
Tried, same happen. Seems to that at weekends it work some 36 hours and
weekdays 25.
More thanks, Alar.

>>> Randy Grein RGrein at tpchd.org Thu Feb 26 17:23:37 GMT 2009 <<<

What jumps out at me is the nonreferenced delay time - 30 minutes is a LONG
time. It's been a while since I've had to deal with a traditional volume,
but try cutting it back to 30 seconds - you can always bump it back up. I
seem to recall some odd things happening with directory cache buffer
exhaustion.

Randy Grein
Sr. Network Engineer

-----Original Message-----
From: Alar Pandis [mailto:Alar.Pandis <at> mtk.ut.ee] 
Sent: Thursday, February 26, 2009 5:48 PM
To: 'Novell LAN Interest Group'
Subject: RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact
every 24 hours

Hi and thanks Randy!

I'd load proxy with cc option on the second day, I think, when this start to
happen ... some two weeks and few days ago. Right now proxy isn't loaded for
week or so. In console log I see that just some bordermanager audit-nlm is
autoloaded.
(Continue reading)

Permalink | Reply | 1 comment |
headers
Peter Van Lone | 2 Mar 2009 18:18
Picon

is it still possible to purchase a support incident for NW5.1?

I've got a customer that called and the server is stuck, can't get to
it due to a looping error.

I've been trying to help him work through it, but at this point we
can't get the networking to load (volumes mount, though) and we can't
re-apply sp6 because we can't run nwconfig/install products because
btrieve is not loading.

GRRR

When I do good, I feel good. When I do bad, I feel bad. That is my religion.

-Abraham Lincoln

http://www.the-brights.net
http://xkcd.com/167
Permalink | Reply | 8 comments |
headers
James Taylor | 2 Mar 2009 18:31

Re: is it still possible to purchase a support incident for NW5.1?

I don't believe you are going to get any help on this from Novell.

However, I do have a suggestion.
Locate an NW5.1sp6 overlay disk and do a down server upgrade.  The procedure is documented and can be easily
found by doing a search on  "down server upgrade" at the Novell site.
If you don't have an sp6 overly, you can use any 5.1 disk and then apply the SP.

-jt

James Taylor
The East Cobb Group, Inc.
678-697-9420
james.taylor <at> eastcobbgroup.com
http://www.eastcobbgroup.com

>>> Peter Van Lone <petervl <at> gmail.com> 3/2/2009 12:18 PM >>> 
I've got a customer that called and the server is stuck, can't get to
it due to a looping error.

I've been trying to help him work through it, but at this point we
can't get the networking to load (volumes mount, though) and we can't
re-apply sp6 because we can't run nwconfig/install products because
btrieve is not loading.

GRRR

When I do good, I feel good. When I do bad, I feel bad. That is my religion.

-Abraham Lincoln
(Continue reading)

Permalink | Reply | 7 comments |
headers
Peter Van Lone | 2 Mar 2009 18:56
Picon

Re: is it still possible to purchase a support incident for NW5.1?

I cannot find an overlay CD -- and a down server upgrade that is
actually a downgrade worries me some. In fact -- I have little
confidence that it will work.

Do you have any idea where I could download the overlay CD quickly?

P

On Mon, Mar 2, 2009 at 11:31 AM, James Taylor
<James.Taylor <at> eastcobbgroup.com> wrote:
> I don't believe you are going to get any help on this from Novell.
>
> However, I do have a suggestion.
> Locate an NW5.1sp6 overlay disk and do a down server upgrade.  The procedure is documented and can be easily
found by doing a search on  "down server upgrade" at the Novell site.
> If you don't have an sp6 overly, you can use any 5.1 disk and then apply the SP.
>
> -jt
>
>
> James Taylor
> The East Cobb Group, Inc.
> 678-697-9420
> james.taylor <at> eastcobbgroup.com
> http://www.eastcobbgroup.com
>
>
>
>
>
(Continue reading)

Permalink | Reply | 6 comments |
headers
Randy Grein | 2 Mar 2009 19:16

RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact every 24 hours

Hmm. I'm out of ideas - unless you're getting a lot of network errors. Can you check the network adapter stats
AND the switch port?

Randy Grein
Sr. Network Engineer
>>> "Alar Pandis" <Alar.Pandis <at> mtk.ut.ee> 3/1/2009 2:40 AM >>>
Hi and thanks again Randy!
Tried, same happen. Seems to that at weekends it work some 36 hours and
weekdays 25.
More thanks, Alar.

>>> Randy Grein RGrein at tpchd.org Thu Feb 26 17:23:37 GMT 2009 <<<

What jumps out at me is the nonreferenced delay time - 30 minutes is a LONG
time. It's been a while since I've had to deal with a traditional volume,
but try cutting it back to 30 seconds - you can always bump it back up. I
seem to recall some odd things happening with directory cache buffer
exhaustion.

Randy Grein
Sr. Network Engineer

-----Original Message-----
From: Alar Pandis [mailto:Alar.Pandis <at> mtk.ut.ee] 
Sent: Thursday, February 26, 2009 5:48 PM
To: 'Novell LAN Interest Group'
Subject: RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact
every 24 hours

Hi and thanks Randy!
(Continue reading)

Permalink | Reply | 0 comments |
headers
jrd | 2 Mar 2009 20:27
Picon
Picon
Favicon

Re: is it still possible to purchase a support incident for NW5.1?

Peter Van Lone wrote:
> I cannot find an overlay CD -- and a down server upgrade that is
> actually a downgrade worries me some. In fact -- I have little
> confidence that it will work.
> 
> Do you have any idea where I could download the overlay CD quickly?
> 
--------------
	http://netlab1.oucs.ox.ac.uk/
Choose Complete File Archives, then allupdates.
Scan down the list and find nw51sp6,7 ,exe files
and nw51sp8 .iso (which is what you want).
	These files were mirrored from Novell's
ftp site until they decided to withdraw the service.
	To increase your confidence, copy your problem
server (disk imaging) to a spare drive and then use
the .iso file to build a test server. If you have a
handy Linux box then a free sector copy tool is the
dd command, such as  dd if=/dev/hda of=mydisk   to
create the image file, and then on a fresh drive do
dd if=mydisk of=/dev/hda. This copies the entire drive,
including partition table, MBR etc. You can copy just
partitions by adding the partition number to the drive
ident,  /dev/hda2  for example.
	Here I use the partition approach, one file
for C: and another for SYS:. Takes about 6 minutes
to do or undo an image file. This may give you an idea
of the down time involved.
	Do save those original sector images in case
all goes pear shaped.
(Continue reading)

Permalink | Reply | 5 comments |
headers
Alar Pandis | 3 Mar 2009 10:34
Picon
Picon

RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact every 24 hours

Hi again and thanks Randy!
I can tell from 3Com switch to BM box (outside line) -- 1671 CRC errors
right now since ... port is up (I changed it from one switch to another to
except problems) 5 days or so. I see there on another ports much worse
statistics, probably ... it is ok. I'm not sure.
But I'd check every switch connection to BM box NIC's and 3Com 4500 tells
(when I log in via telnet) about our WiFi subnet (there is 3Com AP's and all
students with WiFi devices) that "The link partner of Ethernet1 /0/44 may be
bad,sending lots of error packets !". Inside box is for this subnet old 3Com
3c5x9 NIC. For outside line 3c59x and other three is 3c90xc NIC's.
More thanks, Alar.

>>> Randy Grein RGrein at tpchd.org Mon Mar 2 18:16:32 GMT 2009 <<<

Hmm. I'm out of ideas - unless you're getting a lot of network errors. Can
you check the network adapter stats AND the switch port?

Randy Grein
Sr. Network Engineer

-----Original Message-----
From: Alar Pandis [mailto:Alar.Pandis <at> mtk.ut.ee] 
Sent: Sunday, March 01, 2009 12:41 PM
To: 'Novell LAN Interest Group'
Subject: RE: NW 5.1 loose contact every 24 hours -- BM 3.6 looses contact
every 24 hours

Hi and thanks again Randy!
Tried, same happen. Seems to that at weekends it work some 36 hours and
weekdays 25.
(Continue reading)

Permalink | Reply | 803 comments |
headers
Peter Van Lone | 3 Mar 2009 15:56
Picon

Re: is it still possible to purchase a support incident for NW5.1?

We first settled on an attempt to do a down-server "upgrade" to a
non-patched NW 5.0 version (because that is the only installation
media of NW 5.0 that we could locate).

However, the CD hung at the very earliest stages of this process ...
possibly because there were many newer versions of files on the server
system volume. At any rate, we then decided to attempt a down-server
upgrade by using a NW 5.1 CD ... we downloaded the CD (which took
about 3 hours) and then began the installation process. All appeared
to be going well -- it looked as though we would have a functioning
server however after a long process of updating the server's local NDS
database, the install failed due to the system volume running out of
space.

I knew it was going to be tight before starting the install, but I had
no way of deleting any files, as none of the tools (including netbasic
shell) worked in the server's current shape.

I  did discover during the ultimately failed upgrade that there were a
fair number of files on the server's sys volume that were corrupted.
My earlier attempts to load missing modules by hand from the C: drive
was fated to fail, given the amount of corruption.

Oh, well ... let's trust that we can get SOMETHING installed and that
the tape restore works.

This mornings update: Not good news. I could use some help with the
restore DS from tape process, if anyone has experience there. Read
below for explanations:
(Continue reading)

Permalink | Reply | 4 comments |
headers
Joe Doupnik | 3 Mar 2009 16:36
Picon
Picon
Favicon

Re: is it still possible to purchase a support incident for NW5.1?

Peter Van Lone wrote:
> We first settled on an attempt to do a down-server "upgrade" to a
> non-patched NW 5.0 version (because that is the only installation
> media of NW 5.0 that we could locate).
> 
> However, the CD hung at the very earliest stages of this process ...
> possibly because there were many newer versions of files on the server
> system volume. At any rate, we then decided to attempt a down-server
> upgrade by using a NW 5.1 CD ... we downloaded the CD (which took
> about 3 hours) and then began the installation process. All appeared
> to be going well -- it looked as though we would have a functioning
> server however after a long process of updating the server's local NDS
> database, the install failed due to the system volume running out of
> space.
> 
> I knew it was going to be tight before starting the install, but I had
> no way of deleting any files, as none of the tools (including netbasic
> shell) worked in the server's current shape.
> 
> I  did discover during the ultimately failed upgrade that there were a
> fair number of files on the server's sys volume that were corrupted.
> My earlier attempts to load missing modules by hand from the C: drive
> was fated to fail, given the amount of corruption.
> 
> Oh, well ... let's trust that we can get SOMETHING installed and that
> the tape restore works.
> 
> This mornings update: Not good news. I could use some help with the
> restore DS from tape process, if anyone has experience there. Read
> below for explanations:
(Continue reading)

Permalink | Reply | 3 comments |
headers
Peter Van Lone | 3 Mar 2009 16:55
Picon

Re: is it still possible to purchase a support incident for NW5.1?

On Tue, Mar 3, 2009 at 9:36 AM, Joe Doupnik <joe.doupnik <at> oucs.ox.ac.uk> wrote:
> -------------
>        Might I be so brave as to ask whether you made sector copies of
> the disks BEFORE "improving" them? I gather that this step was skipped
> and bridges were burnt.

of course it was skipped -- there is no hardware available to copy it to ..

>        So, do it now, on all the affected drives. Then get more fresh
> drives, build a NW 5.1 SP8 server with plenty of extra space. Do not let
> users approach it yet. Take a bad drive and mount it in the new box, and
> let old SYS: be renamed to avoid volume name conflicts. If you get this
> far then carefully copy the entire contents of old _NETWARE and /etc/nici,
> which are of course eDir files and the server-specific keys. Copy anything
> else of value to a safe spot (not on the just built drive).
>        Now, on the new box, copy the eDir and NICI files to a holding area.
> Load toolbox, but do not login with credentials (important, /nl
> credential-free).
> Unload DS.nlm. Use toolbox to copy _Netware to another safe spot, and the
> old
> _Netware in its place, ditto /etc/nici files. Try reloading DS. If this
> works
> your server is up with the old eDir, though the IP info is likely not
> correct
> so fix that as well.
>        Should things work to this point you can now fire up your regular
> backup/restore utility and try restoring other important files. You can
> carry
> on with this tactic for each broken server. eDir first, then user files with
> trustee information.
(Continue reading)

Permalink | Reply | 2 comments |

Gmane