Tips to check for a replacement switch before connecting?

[13jobsp90]

I had a issue when one of our access switches got dead and it replaced with another one  connected to another switch, earlier it was working fine. The moment I connected  the replacement switch, the entire neighboring switches and others started flapping. What might be the reason for this? Is it due to the switch root bridge issues with the new one or something else? I dont see any issues with the replacement switch while configuring it. All are same models 2960 models.

[marce1000]

  - You need to make sure that it does not become the root bridge of the network, and check the configuration accordingly
     before taking the replacement switch in production,

 M.

[13jobsp90]

how do I make sure?

[marce1000]

  - @13jobsp90  You need to look at it from a reverse viewpoint ;
                          meaning that you prepare a (valid) running configuration that for instance   does not interfere
                          with root bridge priorities (accidently becoming the root)  , VTP ,...

   M.

[13jobsp90]

what command should I check for that before moving to production?

[Jens Albrecht]

Hello @13jobsp90,

there are some common best practices that you can follow to check the replacement switch, especially if it was in use before or refurbished:

• Turn it on without any network cables attached and connect via console
• Verify that it does not have a startup-config. If yes, delete it.
• Verify that is does not have a vlan.dat file. If yes, delete it.
• Set it to VTP mode transparent, even if you want to configure VTP later.
• Verify that the software version is the one you want to use in production and upgrade if necessary.
• Reboot the device if you had to delete any files or upgraded the software in the steps above.
• Upload the config from your backup of the faulty device and confirm that it was loaded without any problems/errors.
• If everything looks fine, save the config and move the device to production.

This ensures that the switch has a well-known state when you move it to your production network and avoids bad surprises.
Of course, you need to explicitly configure the STP root bridges in your network. Otherwise the switch could still become root if it has the lowest MAC address in your network.

Some people might call such procedures overcautious but spending a few minutes on checking the device is definitely better than causing a network outage.

HTH!

[Leo Laohoo]

---

@13jobsp90 wrote:
The moment I connected  the replacement switch, the entire neighboring switches and others started flapping.

---

Network storm. 

Someone has connected a hub/switch and bridged multiple ports. 

Turn on BPDU Guard and see what port(s) go into error-disable. Alternatively, count the number of MAC addresses generated by each port.  Any switch port with three or more is a suspect.