Catalyst c9130axi-b crash during upgrade initiated by controller.

[Icame4theza]

Like the title says, I have a C9130AXI-B that boots fine until it joins the controller. Once joined, the controller initiates a reboot to load the backup image to match the controller and the image seems to be corrupted as the WAP crashes. The WAP reboots and does the same cycle all over again and again in a loop.

I am able to boot into u-boot but I'm not sure what image file I should provide via tftp, nothing I've look at as a solution matches what I see in reality. I've looked into every nook of the interwebs and I'm going in the wrong direction, can I get some assistance in resolving this?

[Leo Laohoo]

---

@Icame4theza wrote:

[*12/20/2023 13:00:02.3991] AP image version 17.6.4.56 backup 17.9.3.50, Controller 17.9.3.50
[*12/20/2023 13:00:02.3992] Version does not match.
[*12/20/2023 13:00:02.3992] AP backup image version matches controller. No download, activate backup.
[*12/20/2023 13:00:02.4440] status 'upgrade.sh: Script called with args:[ACTIVATE]'
[*12/20/2023 13:00:02.4815] do ACTIVATE, part1 is active part
[*12/20/2023 13:00:02.5234] status 'upgrade.sh: activate part2, set BOOT to part2'
[*12/20/2023 13:00:02.8217] status 'upgrade.sh: AP primary version after reload: 17.9.3.50'
[*12/20/2023 13:00:02.8507] status 'upgrade.sh: AP backup version after reload: 17.6.4.56'
[*12/20/2023 13:00:03.3701] status 'upgrade.sh: No btldr.bin found'
[*12/20/2023 13:00:03.3799] Backup image activated. Rebooting AP...​

---

This bit is a normal behaviour but the console output prematurely ends without showing us what happens after the AP boots into 17.9.3.

[Rich R]

@Leo Laohoo 

status 'upgrade.sh: No btldr.bin found'

 then

Backup image activated. Rebooting AP...​

so it never tries to boot the backup because it can't find a bootloader there.

@Icame4theza the way I see it you have a few options:
1. Upgrade the AP to the correct version manually by disconnecting it so that it cannot join the WLC then use console CLI to download correct software version from TFTP. That will replace the software in the backup partition and hopefully resolve the issue.

2. Fire up a 9800-CL running 17.6.4 for the AP to join and then upgrade from WLC CLI telling the AP to download the correct image from TFTP.
ap name APname tftp-downgrade 1.2.3.4 ap1g6a-k9w8-tar.153-3.JPN2.tar

3. Completely re-flash the AP to 8.10.130.0 from u-boot using the process at https://www.cisco.com/c/en/us/support/docs/wireless/catalyst-9120axi-access-point/217537-repairing-c9120-c9115-access-points-from.html then upgrade to 8.10.190.0 then 17.9.3.  If you don't do the intermediate version it will fail as per https://www.cisco.com/c/en/us/td/docs/wireless/controller/9800/17-9/release-notes/rn-17-9-9800.html#Cisco_Concept.dita_59a2987f-2633-4630-8c7b-a8e8aecdeaf7 due to incorrect flash partition sizes.

ps.  You should refer to the TAC recommended versions link below - currently:
Cisco recommends 17.9.4 + SMU_CSCwh87343 + APSP (as needed) OR 17.9.4a + APSP(as needed) for all deployments.
Upgrading your WLC to 17.9.4a would also solve your problem because instead of trying to boot from existing backup partition the AP would download and install the new version instead, so you fix your AP and get your software version up to date at the same time with critical security vulnerability resolved.

[Icame4theza]

@Leo Laohoo Thank you for taking a look at this. The reboot behaves just like the previous time, loads the image from part2 and kernel panics. I've attached continued logs below.

[marce1000 · Accepted solution]

  - Also take a note of this one :   https://www.cisco.com/c/en/us/support/docs/wireless/catalyst-9800-series-wireless-controllers/220443-how-to-avoid-boot-loop-due-to-corrupted.html

 M.

[Icame4theza]

@marce1000  That fixed it. Thanks!

[Rich R]

Glad that fixed it.

I should have remembered to mention that one too - I have the link to it in my signature already - thanks for the reminder @marce1000