Mike, can you add some more details on the issue (problem, setup, reproduction steps), and as to why this has been raised against NFS - since it's mentioned abending in LLDAPSDK.NLM ?  Download of the dumps is going to take a while.

I'll start with the easiest question first:  Why this is raised against NFS?

This is because I cannot find an individual assignment for LLDAPSDK.NLM, nor do I know which "group" this NLM actually belongs to.  If you could please put this in the appropriate group for LLDAPSDK (whatever that is), I would appreciate it.

Details on the problem:

Customer enables SSHD on the server and tries to connect with a client and the server abends.  There is no other type of configuration beyond this that I am aware of.

This is NetWare 6.5 Support Pack 3.

There have been 3 coredumps of this issue:

Coredump #1:  Customer said the server abended while trying to establish SSH connection to the server.  Coredump was analyzed and we found a problem in the Operating System where memory in the processor queues was being accessed and corrupted leading to the abend.  The abend just happened to occur during the SSH connection phase.

A newer server.exe with debug code (to track if the condition was actually hit, and therefore fixed) was delivered to the customer.  Interestingly enough we had another customer with this exact issue, but the abend was not seen when SSH connections were trying to be established.  The workaround for this issue is to disable all processors (except CPU 0).


Coredump #2:  Customer tried the updated server.exe from OS development.  SSH was enabled on the server again and a client connection was attempted to be established.  The server abended in LLDAPSDK.NLM.

Coredump was taken and given to 3 teams (OS, SSH, and LLDAPSDK).  The OS team looked at the image and saw that the code change was executed 3 times (3 possible entry points for corruption were prevented and therefore we believe the abends seen before were prevented).  They also analyzed the coredump for possible problems they may have introduced that led to this abend ... and found none.  SSH developer looked at this and in cooperation with the LLDAPSDK developers they have been theorizing as to why the server abended.  There is no concrete information on this as of yet.

SSH developer sent debug code with additional information recorded to the logger screen to try and narrow in on the problem.  This updated set of NLMs was given to the customer with the purpose of them trying to duplicate the problem again so we could see additional debugging information.


Coredump #3:  With the updated SSH code the customer tried to reproduce the problem again.  The abend happened with the updated code and the messages on the logger screen for the SSH developer were there.

Both coredumps #2 and #3 have been sent to the LLDAPSDK developers via the mirroring directory on prv-coredump.

Thanks Mike - We'll get it across to the right team.

Sunil, not sure if you are the ldap sdk contact - if not can you assign this to the right person ?  Please also verify the product/component are correct for the issue. Thanks, Madhan.

Palani,

Please take care of this defect and update the latest status.

Thanks
rajkumar 

Being working on the issue. But could not replicate the issue at all with all the configuration provided.

This problem was addressed a few years ago.  This bug wasn't ever closed out.