switch lockup

[Scott_O'Brien]

hey all,

our 2960 switch locked up this morning and coursed a few headaches, once i found it was the 2960 i rebooted it ( unpluged and repluged it in) as i wanted to get everything back up ASAP so i didnt have time to console into it.

is there any way to see what my have coursed the lockup? ie i cant do a show history as that only goes back to the last reboot, and show logging does the same.

thanks

[Leo Laohoo]

Not unless you've configured a logging server.

[Joe Conklin]

Even with logging you may no be able to determine the cause. However if you enable logging and do not get any logging during the period of the lockup that maybe a good thing at least for troubleshooting. This often indicates that your device is locked up and not responding. Usually one of two things cause this. Electrical issues or hardware problems. If you plug the hardware device into a UPS so you don't experience brown outs or high voltage then you can rull out electrical issues. If you still have the problem that means you have a hardware problem, hopefully you have a smartnet contract because if you don't you're out of luck for any help from cisco.

[Scott_O'Brien]

sadly we have not tech net support on it, whats the best way to set up a logging server?

it happend again :S what could failhardware wise to make the switch still be on but just locked up?

[Dheebshiba Dinakaran]

Hi,

to configure logging sessions:

step 1: configure terminal

step 2: logging buffered [size]  

Desc:Log messages to an internal bufferon the switch. The default buffer size is 4096. The range is 4096 to 2147483647 bytes.

If the switch fails, the log file is lost unless you previously saved it to flash memory. See Step 4.

Note Do not make the buffer size too large because the switch could run out of memory for other tasks. Use the show memory privileged EXEC command to view the free processor memory on the  switch. However, this value is the maximum available, and the buffer  size should not be set to this amount.

Step 3: logging host

Desc: Log messages to a UNIX syslog server host.

For host, specify the name or IP address of the host to be used as the syslog server.

To build a list of syslog servers that receive logging messages, enter this command more than once.

Step 4: logging file flash:filename [max-file-size [min-file-size]] [severity-level-number | type]

Desc: Store log messages in a file in flash memory.

•For filename, enter the log message filename.

•(Optional) For max-file-size, specify the maximum logging file size. The range is 4096 to 2147483647. The default is 4096 bytes.

•(Optional) For min-file-size, specify the minimum logging file size. The range is 1024 to 2147483647. The default is 2048 bytes.

•(Optional) For severity-level-number | type, specify  either the logging severity level or the logging type. The severity  range is 0 to 7. For a list of logging type keywords, see Table 26-3. By default, the log file receives debugging messages and numerically lower levels.

Step 5 : end

Step 6: terminal monitor

Desc:

Log messages to a nonconsole terminal during the current session.

Terminal parameter-setting commands are set locally and do not remain in  effect after the session has ended. You must perform this step for each  session to see the debugging messages.

Step 7: show running-config

To verify the entries

Step 8: wr

Save the config

For your reference, please find the follwoing link:

<http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_25_see/configuration/guide/swlog.html>

Thank you!

Cheers!

DB

[Scott_O'Brien]

thanks for that, problem is i swapped the switch over with a new switch all went well then 3 - 4 hours later the same thing has happned not at work but will need to go in to see if the new switch has locked up.... any way a network fault could couse this? is it was a loop it would affect the whole network not just this one switch.....

this switch is the vtp server and our root bridge if that helps anything.

[Scott_O'Brien]

hey all,

so i come to work really early to day to see if i could see the problem as i wanted to keep it down untill i consoled into it, so happens that the port leading back to the core goes into err-disabled state becouse of loopback, there is no other ports going back to the core and i now have turned of keepalives on the port,

what could be coursing this? any ideas?

thanks