Troubleshooting guide

What Information Should You Collect?
Advanced System Diagnostics and Troubleshooting Guide 105
What Information Should You Collect?
• Mandatory technical information:
— Before reboot, use the show tech CLI command to collect system information.
— System log (switch log—volatile, NVRAM; external log—complete remote SYSLOG file for the
day of the event)
— Topology diagram
• Background information:
— Any recent hardware or configuration changes
— Detailed description of all the known symptoms
— Was there a service impact?
— Define the scope of the service impact (For example: Only one user? All users on one
BlackDiamond system? All users in the IDC? Multiple switches? Random or pattern? Which
modules? Which ports? Etc.
— Trend: Recurrent event? Frequency? Etc.
— If the problem was resolved, what steps did you take to diagnose and resolve the problem?
• Optional information (upon request from Extreme Networks TAC personnel)
— System dump (CPU memory dump)
• Additional CLI commands for information include:
— show diagnostics
— show log (do the complete log, rather than just part of the log)
— show configuration
— show switch
— show version
Analyzing Data
• Check the log and the output from the show diags command to determine whether a hardware
fault was indicated.
• If a hardware fault is indicated:
— Determine whether a module was taken off line. If “yes,” it should be obvious which module
must be replaced.
— If not, refer to the “Diagnostic Troubleshooting” section (below).
• If the module can be identified clearly from the diagnostic messages, and the outage was
service-affecting, replace the faulty module.
• If the module cannot be identified clearly from the diagnostic messages, contact TAC for further
troubleshooting.