Replacing a failed Cisco Meraki Switch in a Stack

On Jan 6th, one of our the core Meraki switches in the switch stack failed; causing an outage on one of our production hyper-v cluster (because we did not have a redundant network LAGG setup for the storage. It was decided at the time to just have a backup port.) once the port was switched from the dead switch to one of the other working switches the problem was quickly remedied. However now the organization was down a switch and we did need to get it set back up So after trying to reset the switch, and trying anything and everything to bring it back I contacted Meraki support.

To contact meraki support you need your customer number which is located at the bottom of your sign in on the top right of the meraki administration site as shown below.

Meraki Administration Console 

After calling support and jumping through the hoops (turning on/off again, etc) they sent out a new switch. After receiving the switch I put it in placed and replaced it using documentation from https://documentation.meraki.com/MS/Stacking/Switch_Stacks

***NOTE THIS IS PRETTY MUCH VERBATIM

 FROM THE MERAKI SITE WITH SOME NOTES ADDED***

Replacing a Stack Member

The following steps will clone the original stack member and remove it from the stack:

• Power off the stack member to be replaced.



Claim the replacement Meraki device

• Claim the new/replacement switch in the inventory. Navigate to Organization > Inventory
• Enter the serial number of the new switch. If replacing multiple members, list all serials
• Click the Claim button

Connect the new switch to the network; the network should have access to the internet for the switch registration.

Once claimed our licence gave an error message, as we were Overprovisioned on our licencing, which is ok so long as you don't leave it to long. I connected the device via port 1 on one of the other switches which allowed the switch to connect to the Meraki Cloud.

I then cloned the switch using the stack clone and replace member

Cloning The Dead Switch

Now we have a fairly simple network setup and I don't recall needing to bind any profiles meraki has listed in the switch stacks documentation it was pretty much clone dead switch settings to new switch.  Once that was done I rebooted the switch and it came back online; I then connected the stack cables which caused a network outage. I then decided to wait as I was unsure why that had occurred, the switch still had access to the meraki cloud via port 1 and when setting up the stack originally, having the stack cables plugged in didn't cause an issue.  So I shut off the replacement switch to work on it later that night.

I came back at 11:30 pm that night plugged in the stack cables and turned the switch on and it went into the stack with no problem.   I removed the old dead switch from the meraki cloud which fixed the licensing notice we were getting.  I then added 2 new laggs to the x240 switches for the 10GIG link for the Hyper-V iscsi storage. So now they are LACP lagged into stacked switches for additional failover protection; just now on 10Gig instead of 1Gig.

Dealing with Meraki

Dealing with Meraki to get the switch replaced was actually a pretty good experience and would not hesitate recommending Meraki to any organization that can afford to use them.  They mailed the new switch out to us right away and worked with us and our timeline to get the switch put into place and the dead switch sent back.  Replacing the switch was fairly simple and they had pretty good instructions despite the minor issues I experienced.

References

https://documentation.meraki.com/MS/Stacking/Switch_Stacks

https://optionkey.blogspot.com/2018/07/configuring-high-availability-for.html