Hi,
My production GenieACS was 1.2.8 (needed to keep it because of custom change to the code to accomodate issues in a legacy device model) and I want to update it to 1.2.16.
I also needed new package like axios for external script so I build a new container with everything up to date.
So, I have 1.2.8 and 1.2.16, side by side, on 2 differents container, configure to use the same DB. Everything work fine. I can summon devices from both container.
I shutdown 1.2.8 and move the ip address from 1.2.8 to the 1.2.16 container and I can’t summon devices anymore. I get the following error:
DeviceID.ID_value: No contact from CPE
I move back the IP address to 1.2.8 and both can summon fine again.
I’ve made a network capture from the ACS server and looked at it. Here is what I see
Packet 1 from server: GET /tr069
Packet 2 from the CPE: 401 Authorization Required
Packet 3 from the server: GET /tr069 with Digest auth
Packet 4 from the CPE: HTTP 200 OK
Paquet 5 fromt the CPE: HTTP/XML POST
Server don’t answer back but it indicate on the UI “No contact from CPE”
Debug output:
event: incoming HTTP response
timestamp: 2026-05-29T10:43:05.099Z
remoteAddress: 172.17.102.55
deviceId: oui-prdoductclass-225508J000349
connection: 2026-05-29T10:43:05.099Z
statusCode: 200
headers:
date: Fri May 29 06:43:04 2026
server: tr069 http server
content-length: “0”
connection: close
content-type: text/plain; charset=ISO-8859-1
authentication-info: rspauth=“d06be15f53db9b91490efd235e99e10f”, cnonce=“ced685f8f819e461”, nc=00000001, qop=auth
body: null
I’ve been testing it for days, trying to figure out why with Claude and Gemini help but both clearly don’t have any idea of what going on.
What I have is mind is maybe edit DHCP config so ACS discovery to use a new IP address for the ACS server and hoping that it will create a bootstrap event on the CPE to clear everything because it change of ACS server but that a lot of job (several DHCP server) while I’m not even sure it will work.
Anybody have a idea of why it won’t work ? Any hint would be greatly appreciated.
P.S. Periodic inform don’t seems to generate fault. I see the last periodic informs being updated on 1.2.16 but I have no way to be sure the config is really updated on CPE if needed.