Header Only - DO NOT REMOVE - Extreme Networks

V2110 Configuration Upgrade failed


Userlevel 2
During EWC V2110 upgrade from 09.21.05.0007 to 10.11.02.0032 I had following info: Aug 31 06:38:47 CEST 2016 V2110 Upgrading From Release 9_21 Aug 31 06:39:41 CEST 2016 Successfully Restored the Internal Configuration V2110:vnsmode:radius:NAC1# apply Exception: 'utf8' codec can't decode bytes in position 1-2: invalid continuation byte. (file=/usr/lib/python2.7/encodings/utf_8.py:16 function=decode) at /var/controller/images/bu_restore/rpm-upgrade-saved-configuration.cli(1) Aug 31 06:39:42 CEST 2016 Configuration Upgrade failed, Aborting the Process! Please contact your technical support I lost all my WLAN and VNS settings I had on my v9.21 controller. Anyone had similar issues when upgrading from 9.21.x to 10.11.x ? I am going to upgrade customer's EWC infrastructure and I don't want to configure all wireless settings from scratch. BEST REGARDS Robert

6 replies

Userlevel 7
I haven't had any migration issues for a long time - you should open a ticket and provide the configuration so the guys could run the same upgrade with your config to see which line in the config forces the problem.

To the lost data I'd only say one word "Backup".
You should always do one before an upgrade and save it a external device.

But even if the upgrade failed the default upgrade option would be to generate a rescue image before the upgrade even start.
Load it back and all data is restored..... if you haven't disabled it which is not recommended.
Ron wrote:

I haven't had any migration issues for a long time - you should open a ticket and provide the configuration so the guys could run the same upgrade with your config to see which line in the config forces the problem.

To the lost data I'd only say one word "Backup".
You should always do one before an upgrade and save it a external device.

But even if the upgrade failed the default upgrade option would be to generate a rescue image before the upgrade even start.
Load it back and all data is restored..... if you haven't disabled it which is not recommended.

+1 on the rescue image. And since it's a virtual controller, take a quick snapshot in VMware first.
Userlevel 2
Ron wrote:

I haven't had any migration issues for a long time - you should open a ticket and provide the configuration so the guys could run the same upgrade with your config to see which line in the config forces the problem.

To the lost data I'd only say one word "Backup".
You should always do one before an upgrade and save it a external device.

But even if the upgrade failed the default upgrade option would be to generate a rescue image before the upgrade even start.
Load it back and all data is restored..... if you haven't disabled it which is not recommended.

Sure, but both: rescue image and/or snapshot takes me back to my older image, which is not the right direction 🙂
Userlevel 2
Ron wrote:

I haven't had any migration issues for a long time - you should open a ticket and provide the configuration so the guys could run the same upgrade with your config to see which line in the config forces the problem.

To the lost data I'd only say one word "Backup".
You should always do one before an upgrade and save it a external device.

But even if the upgrade failed the default upgrade option would be to generate a rescue image before the upgrade even start.
Load it back and all data is restored..... if you haven't disabled it which is not recommended.

One question: does rescue image downgrade reverts back APs fimrware version to match the EWC version ?
Userlevel 7
Ron wrote:

I haven't had any migration issues for a long time - you should open a ticket and provide the configuration so the guys could run the same upgrade with your config to see which line in the config forces the problem.

To the lost data I'd only say one word "Backup".
You should always do one before an upgrade and save it a external device.

But even if the upgrade failed the default upgrade option would be to generate a rescue image before the upgrade even start.
Load it back and all data is restored..... if you haven't disabled it which is not recommended.

No, the rescue image has nothing to do with the AP image BUT during the tunnel setup between the AP/controller the AP will check the version.

If auto upgrade is enabled (default) then the AP will up/downgrade to the same version as the controller is running.
Userlevel 6
I would guess there are some special characters in your config somewhere. try to avoid that and upgrade again...

Reply