C9115-AXI-E-EWC AP JOIN PROCESS ISSUE

[henrielle]

Hello,

I just went back into networking, and I have a project where I should deploy 13 APs in a flat network (for the time being).

As far as I remember in WLC, the WLC holds the images wherein an AP joins, it will download the image from the controller itself.

In EWC I found out that it doesn't keep the image internally and the image location should be pointed in a TFTP server, etc.

Now I added a command in my EWC where to point the joining AP to get the image so it can successfully join to the EWC.

EWC:

wireless profile image-download default
 description "default image download profile"
 image-download-mode tftp
 tftp-image-path /
 tftp-image-server 192.168.0.132

As soon as I plug the AP I wanted to join, it gave me this cli output. (CLI output from the joining AP).

[*05/10/2025 14:45:52.3580] CAPWAP State: Join
[*05/10/2025 14:45:52.3770] shared_setenv PART_BOOTCNT 0 &> /dev/null
[*05/10/2025 14:45:52.9460] OOBImageDnld: OOB Image Download in ap_cap_bitmask(2)
[*05/10/2025 14:45:52.9460] Sending Join request to 192.168.0.174 through port 5258, packet size 1376
[*05/10/2025 14:45:57.6440] OOBImageDnld: OOB Image Download in ap_cap_bitmask(2)
[*05/10/2025 14:45:57.6450] Sending Join request to 192.168.0.174 through port 5258, packet size 1376
[*05/10/2025 14:45:57.6550] Join Response from 192.168.0.174, packet size 1397
[*05/10/2025 14:45:57.6560] AC accepted previous sent request with result code: 0
[*05/10/2025 14:45:58.2250] Received wlcType 2, timer 30
[*05/10/2025 14:45:58.2260] Primary-Replica: role 3, primary server '192.168.0.167', old 0
[*05/10/2025 14:45:58.2340] OOBImageDnld: Do common error handler for OOB image download..
[*05/10/2025 14:45:58.3110] status 'upgrade.sh: Script called with args:[CANCEL]'
[*05/10/2025 14:45:58.3580] do CANCEL, part2 is active part
[*05/10/2025 14:45:58.3790] status 'upgrade.sh: Cleanup tmp files ...'
[*05/10/2025 14:45:58.4030] Primary-Replica: Set to Replica, retry 44, set_tftp_server 0, Primary server 192.168.0.167, Join role Replica, tftp_timeoutCnt 3
[*05/10/2025 14:45:58.4090] Image pre-download request for version 17.15.3.28.
[*05/10/2025 14:45:58.4090] Request to close the file..
[*05/10/2025 14:45:58.4090] wtpOpenImgFile: image file closed, dcb->fd set to -1.
[*05/10/2025 14:45:58.4480] status 'upgrade.sh: Script called with args:[PRECHECK]'
[*05/10/2025 14:45:58.4930] do PRECHECK, part2 is active part
[*05/10/2025 14:45:58.5050] status 'upgrade.sh: Cleanup tmp files ...'
[*05/10/2025 14:45:58.5520] status 'upgrade.sh: /tmp space: OK available 96444, required 50000 '
[*05/10/2025 14:45:58.5530] Memory check success. Memory: (210 required, 1403 current)
[*05/10/2025 14:45:58.5530] Launching image pre-Download task, prev_taskId 0
[*05/10/2025 14:45:58.5530] Launching image pre-Download task success and id(1598524976)
[*05/10/2025 14:45:58.5530] Starting image pre-download task, role=1
[*05/10/2025 14:45:58.5530] Replica Predownload(5): Primary server 192.168.0.167, ver 0.0.0.0, retry 44, timeout 3, image str /tftpboot/ap1g7-pre17.15.3.28.tar
[*05/10/2025 14:45:58.5540] Save Predownload URI tftp://192.168.0.167//tftpboot/ap1g7-pre17.15.3.28.tar
[*05/10/2025 14:45:58.5960] status 'upgrade.sh: Script called with args:[-r 44 -s tftp://192.168.0.167//tftpboot/ap1g7-pre17.15.3.28.tar -n /tftpboot/ap1g7-pre17.15.3.28.tar FLEX_REPLICA_DOWNLOAD]'
[*05/10/2025 14:45:58.6330]
[*05/10/2025 14:45:58.6330] CAPWAP State: Image Data
[*05/10/2025 14:45:58.6390] AP image version 17.9.4.27 backup 17.9.4.27, Controller 17.15.3.28
[*05/10/2025 14:45:58.6390] Version does not match.

Now I am confused why is it pointing to the 192.168.0.167, when I manually configured in the EWC that the tftp is 192.168.0.132 (which is my PC).

Somehow after a while the joining AP suddenly joined the EWC process, and I cannot comprehend what happened exactly.

The 2nd AP became Standby as well.

If anyone can explain I will be very happy.

Thanks in advance.

[Rich R]

Are your APs all the same model? If so then it is possible for them to download the image from the EWC - still best to have TFTP setup/configured though.

You might have another AP also running as EWC controller?

Check your controller config - you might have multiple tags and profiles configured with the other config?
Check the AP is getting the tags and profiles you're expecting - show ap tag summ.
The AP might have that config configured before it even joins (from previous setup) - have you done a factory default reset of the AP?

[henrielle]

Hi!

Yes, all APs are C9115AXI-E and yes, I reset the AP that I made as controller due to some software upgrade mistake. 

The rest of the APs are untouched (12 qty). 1 I used for testing, and it is weird since it joined the controller after some time.

There are no profiles other than the default.

[Rich R]

That config has to come from somewhere! 
Either another EWC controller or saved on that AP.
Use a packet capture to see what the AP is talking to?