Header Only - DO NOT REMOVE - Extreme Networks

Both nodes in a 2 switch stack show Role as Master


Switch 2, previously with Role Backup, is now showing at Role Master. Switch 1 is still actively the Master. All ports on Switch 2 are now showing as Ready. Devices connected to Switch 2 are no longer responding.

8 replies

Userlevel 6
Can you send a "show switch" and "show slot"?
Patrick Voss wrote:

Can you send a "show switch" and "show slot"?

Show switch displays only switch 1. Here's output from show slot:
Slot-1 SJCO-DMZ-X460-48T-2.23 # sh slot
Slots Type Configured State Ports
--------------------------------------------------------------------
Slot-1 X460-48t X460-48t Operational 58
Slot-2 X460-48t X460-48t Failed 58
Slot-3 Empty 0
Patrick Voss wrote:

Can you send a "show switch" and "show slot"?

version 12.5.4.5
Slot-2 information:
State: Failed
Download %: 100
Restart count: 1 (limit 5)
Serial number: 800322-00-07 1151G-82157
Hw Module Type: X460-48t
SW Version: 15.2.3.2
SW Build: v1523b2-patch1-12
Configured Type: X460-48t
Ports available: 58
Recovery Mode: Reset
Node MAC: 02:04:96:83:9E:27
Current State: FAIL
Image Selected: secondary
Image Booted: secondary
Primary ver: 12.6.4.1
Secondary ver: 15.2.3.2
Config Selected: primary.cfg
Patrick Voss wrote:

Can you send a "show switch" and "show slot"?

Sorry, 15.2.3.2 is the active version.
Userlevel 6
Can you send a "show slot 2 detail" and also what version of code is running on this stack? The "show switch" should provide that with the patch.
Userlevel 6
I am sure a reboot of slot 2 will resolve this. As for why there might be something in the logs. Do we know when this happened?

On the other side I would recommend upgrading this stack to the latest of 16.2. In 15.4 some changes were made to the stacking libraries in the code and it prevented these dual master situations.
Patrick Voss wrote:

I am sure a reboot of slot 2 will resolve this. As for why there might be something in the logs. Do we know when this happened?

On the other side I would recommend upgrading this stack to the latest of 16.2. In 15.4 some changes were made to the stacking libraries in the code and it prevented these dual master situations.

Rebooting the slot resolved the issue. It's back online. Once again, thanks for the help.
Log states Backup Node Down. Log entries go back only to 06:00 PST this morning. That's the first indication.
I'll reboot Node 2. Thanks very much for responding.

Reply