display uptime for each MDS9500 port

Unanswered Question
May 14th, 2006
User Badges:

Does anyone know how to retrieve the last time each port on a mds 9500 series director was issued a shutdown command?

I have the problem listed in the release notes for san-os 1.3(3c) under CSCee89946 where once the uptime for a port reaches 248 days a port link reinitialization fails following a link down state change.

I have run a show port internal info command which displays 3 dates but I cannot determine which of these if any are a recording of the last time the port was shut.

This is an extract:

Lock Info: resource [fc1/1]

type[0] p_gwrap[(nil)]

FREE @ 411108 usecs after Sat Aug 7 22:29:32 2004

type[1] p_gwrap[(nil)]

FREE @ 963835 usecs after Mon Jun 20 00:51:40 2005

type[2] p_gwrap[(nil)]

FREE @ 519076 usecs after Sat Aug 7 22:29:33 2004

I can't upgrade the SAN-OS soon due to internal change management processes, so I need to know which ports will fail soon so I can issue the workaround.

Any help would be appreciated.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
dmcloon Mon, 05/15/2006 - 17:44
User Badges:
  • Bronze, 100 points or more

Perhaps you could look in the accounting log. For example below I shut/no shut fc1/12.

Tue May 16 01:41:15 2006:update:/dev/pts/1_1147743584:admin:Interface fc1/12 state updated to down

Tue May 16 01:41:18 2006:update:/dev/pts/1_1147743584:admin:Interface fc1/12 state updated to up

charlie005 Thu, 05/25/2006 - 20:49
User Badges:


I thought that was going to be the only way. I will just issue a shut/no shut each time I reboot a san attached server until I upgrade the SAN-OS

dmcloon Thu, 05/25/2006 - 21:56
User Badges:
  • Bronze, 100 points or more

Hi Charlie, there is no need to shut/no shut every time you reboot the server. With the 248 day bug, you only need to shut/no shut the port once. As soon as you do that the port will be OK for another 248 days. Hopefully you can upgrade before the next set of 248 days lapses.

charlie005 Thu, 05/25/2006 - 22:12
User Badges:

Thanks dmcloon

Its more a management issue that we have 130 servers connected to the SAN and 4 depts that reboot servers and so it is easier to have a process that everyone follows each time they reboot a server rather than ask them to determine whether it is neccessary or not.




This Discussion



Trending Topics: Storage Networking