So we have an issue
We are trying to upgrade a stack (new hardware same config as previous stack), and I have provisioned all the stack members on the stack master (3 switch stack all 2960X Catalyst 48 port).
Some documents (like this one: https://ipwithease.com/switch-stacking-basic-setup-and-configuration-steps/) say all I have to do is power on each switch in the order that we want the stack member to be.
Others say we have to login and declare the membership for each member of the stack.
The issue we have is the previous configuration has radius installed and used on the console port, and I am unsure if the Radius is coming in via the second or 3rd member of the stack (perhaps its a port channel with 1 port on each stack, which is configured, but anyway I'm unsure), I wasn't the original person to configure the stack settings.
How do I log in to the switch stack and declare the 2nd and 3rd switch membership, if I cannot login as I don't have access via Radius, as its coming in via one of those switch members?
You take all the config from OLD switch and implement in new switch - removing AAA config, that can be implemented later once stack build and working. you can create a Local user account with priv 15 to have console access./ telnet /ssh
regard to stacking.
1. all should be same version of IOS as stack rules (hope you have checked already stacking rules)
2. Make them numbering and stack priority - each switch
3. wr the config all device. (stand alone)
4. Power them off
5. connect stack cables.
6. start with high priority and low priority is suggested method always.
7. Once stack booted check show switch make sure all the switches joined in stack and working as expected.
8. Paste old config to new switch (wotu out AAA config).
9. make a maintenance window to replace old with new switch.
10. cut over with arrangement done.
11. once Stack working in production - apply AAA config (hope you have fall back to Local authentication)
Hope this make sense ?
*** Rate All Helpful Responses ***
- The old switches now are in production, with the same configuration to
be placed on the new switches
BB - that is the reason i have advised to prepare stack offline - with old Live config - so you can replace easily (with out AAA config).
- We are replacing the entire stack
BB - i do understand the requirement, hence that suggestion made.
Accordingly, I have provisioned all 3 of the stack members on the stack
master (the configuration for switch 2 and switch 3 are on the stack master)
BB - i was suggest basic config on the standalone switch was, stack priorioty and switch numbering - this have your control fully.
Since we are replacing the entire stack, why would this not work?
1. Install Stack Master
BB - in this case your whole config of OLD stack can not take it, since only 1 switch available.
2. Connect up the stack members (#2 and #3) via stacking cables (only)
BB - yes once prepare offline you can do this way understand control
Master - Priority 10
Member - can be master if master fails - Priority 09
Member - 08 or 0 ( can not be act as Master any time.
3. Power on the switches in order (first 2, then 3)
4. Let the stackmaster configure both members with
BB - if you adding this device to master single switch, you need change the config manually for the new ports added.
hope this make sense ?
Other note : your approach also works , but i would suggest better control way - so your change over time can be used betterway, not last minute rush.
you can focus on patching rather config issues around.
*** Rate All Helpful Responses ***
Not sure i follow, You have a stack of 3 switches and you want to access the two member switches correct?
First use the designated master switches console port as/when you’ve logged in, you can session on to the member switches.
session x <stack member switch>
No we're replacing the full switch stack with the same configuration, but
So you build your stack first ,specifying the master and member switches and then apply the config from the old stack, if you apply radius authentication in that configuration as long it includes fallback to local access if/when the radius server is unavailable you should be able to gain access the new stack via the master switch console port using local authentication credentials
username <username> privilege 15 secret <password>
aaa new model
aaa authentication login default group radius local
aaa authorization exec default group radius local if-authenticated