Header Only - DO NOT REMOVE - Extreme Networks

DFE Reboots after 497 Days of Uptime, with f/w prior to 5.42.06


Userlevel 3
Article ID: 10911

Products
Matrix N-Series DFE; firmware 5.42.04 and lower

Symptoms
Module resets.
Watchdog "
code:
client 26
" and/or "
code:
client 37
" error message in the Fault Log (5101); for example:
<
code:
1
>
code:
DistServ[1.tDsBrdOk]serverWatchDog.5, client 26 in recv for 8261 tics

<
code:
1
>
code:
DistServ[3.tDsBrdOk]serverWatchDog.1, client 37 in recv for 8677 tics


If the user is running NetSight and has traps enabled they will get notification that the board reset due to 497 days.

Solution/Workaround
For the 5.x line, upgrade to firmware 5.42.06 or higher.
Release notes state, in the '
code:
Firmware Changes and Enhancements
' section:
[code]Network Discovery Services[/code]
[code]Escalation 539131[/code]

code:
A problem in Network Discovery Services (CDP/LLDP/CiscoDP) that could result in a reset when sysUptime wraps back to zero has been corrected. sysUptime is a 32bit centisecond counter that wraps every 497 days.


For the 6.x line, upgrade to firmware 6.01.01.0020 or higher.
This completes the fix of the issue explained in 5719.

Note: All uptime counters of whatever size do eventually have to roll over, then start again at zero. The sysUpTime counter is no exception, doing this every 497 days, 2 hours, and 28 minutes of uptime. It does not trigger a device reset, and is not known to be disruptive to any onboard processes. External processes which monitor system uptime could, however, have odd behavior when the field they are monitoring suddenly has a very small value.

0 replies

Be the first to reply!

Reply