I have a setup with dual redundant Cisco catalyst 4506-E using the WS-X45-SUP8L-E supervisor engine and two 48-port switches below it.
- My task is to access and modify the switch configuration.
- Access shall be accomplished locally only, through console port using serial cable and user/pass combo.
The notion that "thou shalt not use ssh / telnet / SNMP to fix this problem" is broken. Use what you can to fix it. Yes, console logging at a lower baud rate relative to a faster stream of incoming console syslog messages is a pain; however you can't change what's already happened.
If nobody set up SNMP / telnet / ssh authentication methods and you can't change the environment which is generating so many logs, the right answer to take to your management is: "Sorry but this service was not set up in a maintainable manner. We have to take an outage to fix it."
Then fix what you know is wrong:
- Common practice is to disable console logging completely with
no logging console. I use no logging console guaranteed, but that's a personal preference.
- Set up SNMP RO / SNMP RW communities with the right ACLs if not already in place
- Set up SSH access
- It sounds like there is a lack of confidence in the redundancy provided by this service. During your outage window, demonstrate expected outage times when you drop power on the active supervisor (do this at least 3 to 5 times so you know how repeatable it is). If the redundancy performance is not what you want, sometimes you can change timers or the protocol to get what you want.
Our community is here to help if you have questions about how to improve failover performance, or configuring the items above.
