Extreme 3802 Installation Guide

Page of 394
 
Meeting Site Requirements
Extreme Networks Consolidated "i" and "e" Series Hardware Installation Guide
37
NOTE
Note that automatic shutdown is based solely on the power supply temperature sensor and is not based 
on the temperature sensor of the MSM. The power supply temperature sensor readings are not visible 
through the ExtremeWare command set. This protection is designed to prevent thermal runaway 
conditions, and does not ensure the protection of the system modules. Internal temperatures above 
55° C (+/- 5° C) can produce a decline in the long-term reliability of the system overall.
The system will recover automatically once the temperature has dropped to 80° C as tracked by the 
independent temperature sensor within the power supply. The temperature sensor remains active when 
the system is in a shutdown state.
System Alarms in an Overheating Condition.  
Extreme supports the following SNMP alarms to 
report overheating conditions:
• Over Temperature Alarm:
 This trap sends notification of an overheating condition, but takes no 
action. The alarm threshold is fixed at 55° C; however, you can control the alarm by turning it ON or 
OFF.
• Overheat:
 This is a trap sent before the system is taken offline. It indicates that the onboard 
temperature sensor has reported an overheating condition. The system will shut down until the unit 
has cooled sufficiently. A cold start trap will be issued when the unit has come back online.
You can use the ExtremeWare command set to capture temperature logs on an hourly basis. However, 
the traps on the system are not configurable.
You can use these traps to trigger an SNMP management platform to react with an external program to 
perform programmed actions to remedy the situation. For example, you could shut down power by 
using a remote power management device available from a third party. The actions performed depend 
on both the capability of the management platform and any third-party device that can be set to control 
the flow of power to the switch.
Monitoring Overheating Symptoms.  
In general, any system reporting temperatures above 40° C 
should be closely monitored. You should take the following actions:
Use the ExtremeWare command set to enable temperature logging and monitor for any rises or 
decline in temperature in the system logs.
Ensure proper environment conditions as described earlier in this section.
Environmental conditions with especially poor airflow can severely impact system performance. 
Numerous racks to the left and right of the system being monitored can result in the system being 
cooled by the exhaust of another system or influenced by the heat emitted from the adjoining 
systems. In addition, enclosed racks can prevent proper airflow. Poor environment placement of this 
type can typically result in as much as a 10° C increase to the ambient temperature (that is, the 
temperature of the environment around the switch) being fed to the system.
Check airflow into and out of the system, ensuring appropriate spacing for proper air circulation.
Verify that any additional equipment installation has not blocked airflow.
Check fans to ensure that there has not been a slowdown or failure of fan operation.
Ensure that proper blank faceplates are used for all open slots.
Openings in a chassis can result in air eddies in the system, resulting in regions of the system 
operating at higher temperatures.