CodeVerge.Net Beta


   Item Entry   Register  Login  
Microsoft News
Asp.Net Forums
IBM Software
Borland Forums
Adobe Forums
Novell Forums




Can Reply:  Yes Members Can Edit: No Online: Yes
Zone: > NEWSGROUP > Novell Forums > novell.support.cluster-services Tags:
Item Type: Date Entered: 6/12/2009 12:04:15 PM Date Modified: Subscribers: 0 Subscribe Alert
Rate It:
NR
XPoints: N/A Replies: 6 Views: 10 Favorited: 0 Favorite
7 Items, 1 Pages 1 |< << Go >> >|
"Monte Meyer" <
NewsGroup User
Abend - At least one of the nodes is Alive in the old master's node partition6/12/2009 12:04:15 PM
Reply

0

Had some network communication issues last night on one of the nodes of a 2
node NW 65 cluster. The node that is abending was the master ip address
resource at the time. The node will continue to abend as long as the master
ip address resource is on the other node. The node will not abend if it is
hosting the master ip address resource.

I get the abend message:

Abend on P00
At least one of the nodes is alive in the old master's node partition. This
node is not in the old master's node partition.

Although the node is running right now it appears to be having intermittent
communication issues with the other cluster node and another standalone NW
server.

It appears both network cards in the problem node are working correctly
now.

Any ideas how to correct this problem?

Thank you,
Monte
utman <utman@no
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/12/2009 4:56:02 PM
Reply

0


This is not an uncommon problem with 2 node clusters. Generally if you
reboot the node that abended it will come up fine. If not I have had
cases where I had to reboot both nodes. The problem is that the two
nodes can't communicate via the LAN for some reason so you get into a he
said she said scenario.


--
utman
------------------------------------------------------------------------
utman's Profile: http://forums.novell.com/member.php?userid=5893
View this thread: http://forums.novell.com/showthread.php?t=376064

"Monte Meyer" <
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/12/2009 9:19:05 PM
Reply

0

Yes, I have tried several reboots but the problem is still there. The LAN
appears to be working fine. However, the console shows the node having the
abend problem losing communication with the "good" node and also with
another standalone NW 6.5 box. The good node shows it losing &
re-establishing communication with the abending node every 20-30 minutes.
The good node is not losing communication with the standalone NW server.

The conolse message shows the date & time communication is lost and
DS-10054.34-271. Could there be a problem with eDir that is causing the
communication problem and thus the abend?



ataubman <ataub
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/15/2009 12:36:02 AM
Reply

0


No, but lack of communication will cause both cluster and eDir problems.
I'm sure you have an underlying comms issue.

What SP on NW 6.5? What exact lan driver, date and version?


--
Andrew C Taubman
Novell Knowledge Partner
http://forums.novell.com/
(Sorry, support is not provided via e-mail)

Opinions expressed above are not
necessarily those of Novell Inc.
------------------------------------------------------------------------
ataubman's Profile: http://forums.novell.com/member.php?userid=34
View this thread: http://forums.novell.com/showthread.php?t=376064

"Monte Meyer" <
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/16/2009 2:15:38 AM
Reply

0

NW 6.5 SP6, attached to an EMC SAN. EMC has never validated their software
beyond SP6, thus the reason it isn't on at least SP7.

LAN driver is Intel CE1000.LAN, version 7.84, March 9, 2005. It appears the
newer drivers for the Intel Pro/1000 server adapter are named E1000.LAN. Is
there a way to replace the driver with the newer one or do new boards have
to be configured?

Monte
ataubman <ataub
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/17/2009 11:26:02 PM
Reply

0


Simply copy then to sys:system, and edit autoexec.ncf or netinfo.cfg,
whichever you use, and change the LOAD line to E1000 where it says
CE1000.

Note I'm not saying this driver is likely the problem, the Intels have
been pretty good. More likely you have a faulty lan card, cable, router,
switch or similar.


--
Andrew C Taubman
Novell Knowledge Partner
http://forums.novell.com/
(Sorry, support is not provided via e-mail)

Opinions expressed above are not
necessarily those of Novell Inc.
------------------------------------------------------------------------
ataubman's Profile: http://forums.novell.com/member.php?userid=34
View this thread: http://forums.novell.com/showthread.php?t=376064

"Monte Meyer" <
NewsGroup User
Re: Abend - At least one of the nodes is Alive in the old master's nodepartition6/18/2009 10:39:18 PM
Reply

0

I do suspect a bad switch port and will do some more investigating to
isolate the problem.

Monte
7 Items, 1 Pages 1 |< << Go >> >|


Free Download:







dmz - acl open for which servers?

oes2 nss volumes have different mount points on the nodes

nw6.5 sp6 cluster device not staying sharable after reboot

xenlive_template

moving iscsi san to separate nics

cluster resource on 2 different subnets?

configuration cluster 2 nodes oes 2 linux

multipath load balancing options for clustering?

move server

ups issues and cluster nodes

nfs exports on clustered volumes oes2 sp1 linux

expanding a pool/volume

unmounting volumes

multiple sbd partition on single hard drive?

how to reset cluster node if local file system is unavailable

sharable services on mixed ncs cluster (linux netware)

replace cluster hardware

cluster pools not visible on some cluster nodes

odd cluster behavior

clustering newbie questions

imanager cluster manager

cluster resource slp info

moving resources from one cluster to another

imanager cluster management fails on linux

no more .load.out files

nss /errorcode=20897

which to do first?

slow cluster resource load on oes/linuxbookmark

managing linux cluster via console one

oes2 cluster

win client does not reconnect to cluster volume

moving from old to new cluster

clustering iprint oes2sp1 linux cluster

oes1 sp2 rug patches - cluster services not loading

cluster email notification

abend - at least one of the nodes is alive in the old master's node partition

gw7 cluster w/oes2 linux on shared ext3 filesystem

anyone tried clustering of openslp on oes2 ?

resource hangs when unloading

cluster resources go comatose on node failure

connecting a 2d node

cluster-<warning>-<12020> / warning 12038

how do i avoid coredump and abend.log prompt?

imanager clustering error

anyone clustering xen vm's as resources?

migrate or offline cluster resource kills ndsd

cluster 2 sites / bcc

clustering intranet

license oes 2 linux - ncs

cluster enable existing dhcp, dns, iprint

   
  Privacy | Contact Us
All Times Are GMT