MS225 stack members offline

[Looker44]

moved to MS12.33 all looked well until our switches started going offline. Downgraded to 12.28.1 . Then 2 of 6 switches would not show as online. After numerous reboot attempts (whole stack, just the affected switch, etc) I still have 2 switches offline. Meraki says get out of band logs. However, when i browse to the switch IP and try to login my credentials don't work. I also tried the SN in the user w/no password. same thing. Traffic is being passed through switches

On hold with Meraki for over an hour and same music in the ear. Any help from community?

[cmr]

@Looker44 when you look at the ARP table of the device that is the default gateway for the management VLAN, can you see each switch's MAC address with the right IP address against it?

Have you tried to do a reset on one of the switches so it is forced to get the config again from the cloud (note this will disrupt user traffic and it might take a little while to get it back).

[Looker44]

ARP table is correct. Have patiently rebooted NUMEROUS times 

[cmr]

@Looker44 I'm not suggesting a reboot, but a reset (hold down button behind pinhole on faceplate for 10 seconds) to clear and re-request the config.

You'll want to connect one of the front facing ports to the next hop towards the internet first, so it can get to the cloud when not using the stacking cables as they will initially not be provisioned.  The next hop will also need to supply an IP address via DHCP.

[Looker44]

well at this point thats the best suggestion so far. Will try this weekend

[PhilipDAth]

When this happens you should start by checking the local status page of the switch (es) that are offline.  This will tell you why the switch is offline.

Some causes I have seen; partway through a firmware upgrade (you think because the others are done this one should also be done - you just need to wait longer); stuck doing a firmware upgrade (hold the reset button in for 10s); lost connectivity to the Internet for some reason, perhaps upgrading another switch in the stack has caused this stack member to get cut off from DNS or the Internet (sometimes you might need to just reset the switch, other times you might need to pull it out of the stack, manually connect it to the Internet somehow so it can complete the upgrade, and then put it back into the stack).

The number #1 issue - not waiting long enough.  Thinking because something else has finished it should also have finished.  I have personally been caught by this more times than I would like to admit.