ITEM: Q2553L

How to get switch response from every node on the sp2

Question:

How to get switch response from every node on the sp2

Response:


   Env:
        sp2  16 node 
   Desc:
        System went down last night.  He is trying to get it back up.
        He reloaded all nodes.  Estart starts only 14 of the 16 nodes. 
        He has another problem open this morning with display leds not 
        working.  

   Action:
        Paged sp2.

   Next Action:
        Please call customer.

   Test Case:

Response:

I have talked to the customer.  He is having switch problems on two
of his nodes.  We think it's hardware.

Response:

ENV: PSSP 2.1.  Nodes 201 and 203 are having css0 problems.  They lost
power last night.  This seems to maybe have impacted the frame.

DESC: OK, Switch responds for two nodes are red.  We cannot ping the
switch host name from node1 (the Primary node).  It does resolve to
the hostname, but it can't be pinged.  We cannot ping out of the
switch, but the switch can ping itself.  There are no errors in the
errpt or in /var/adm/SPlogs/css about the switch.  We did ifconfig
css0 up on the two nodes, but it didn't work.  This might be a
hardware problem so they are going to call their CE about the switch
adapters in the nodes.  css0 is available on both the nodes.

ACT: The customer is going to be calling their CE out to look at the
css0 adapters.

Next Desc:  The customer is having kerberos problems.  He looked in
his logs and he has an error in
/var/adm/SPlogs/sysman/sp2105.console.log about kinit not being able
to send a request to the server.  The customer also has a lot of
questions about users on the nodes and how they work with kerberos.

NEXT ACTION:  Research and help the customer with kerberos.

Response:

If client calls, please transfer him to Miguel Salgado's office T/L
522-6648.
                           

Response:

Response:

I know Miguel tried to reach him last Friday a couple of times.  I am
trying to get a hold of him now.

Response:

I beeped the customer's beeper.  I will also send him e-mail.

ACT:  The customer called me back from the beep.  I am on the phone
with him now.

Response:

6X428 b194 was created for the switch problem.  I also talked to Dan
F. from hardware support in Poughkeepsie about their setup.

6X430 b194 was created to help with the poe errors they were getting
when trying to run a || job as a user.

NEXT: The customer is still having kerberos problems.  I will try and
talk with the person who takes the pmr about this to see if it needs
to stay here or go to Kingston.

Response:

All kerberos problems were resolved by Kingston.  Closing.



Support Line: How to get switch response from every node on the sp2 ITEM: Q2553L

Dated: April 1996 Category: N/A

This HTML file was generated 99/06/24~13:30:37

Comments or suggestions?
Contact us