[Help]: WARN [client-schedule-recognition-task-0] [] c.t.s.o.c.d.s.t(): fail to recognize clients - every 30 seconds

[michnixweiss]

Controller Version

6.2.10.17

Describe Your Issue or Question

Every 30 sec I got this warn in the logs:

WARN [client-schedule-recognition-task-0] [] c.t.s.o.c.d.s.t(): fail to recognize clients

Does anyone know whats causing this - is it related to the container, omada, my config, a device or a client?

Expected Behavior

no warn messages

Steps to Reproduce

just start the container (Im running the container as a podman user quadlet)

How You're Launching the Container

[Unit]
Description=TP-Link Omada Controller

[Container]
ContainerName=omadaSDN
EnvironmentFile=/path/to/omadaSDN/env.omada
Image=docker.io/mbentley/omada-controller:6.2-openj9-arm64

HealthCmd=/healthcheck.sh
HealthInterval=60s
HealthOnFailure=kill
HealthRetries=5
HealthStartPeriod=5m
HealthTimeout=30s

PodmanArgs=--memory 4000000000
PodmanArgs=--cpu-shares 1024
PodmanArgs=--stop-timeout=60
PublishPort=8088:8088
PublishPort=8043:8043
PublishPort=8843:8843
PublishPort=19810:19810/udp
PublishPort=27001:27001/udp
PublishPort=29810:29810/udp
PublishPort=29811-29817:29811-29817
Volume=/path/to/omadaSDN/d.omada/data:/opt/tplink/EAPController/data
Volume=/path/to/omadaSDN/d.omada/logs:/opt/tplink/EAPController/logs

[Service]
Restart=always
TimeoutStopSec=60
SuccessExitStatus=0 143

[Install]
WantedBy=podman-ctr.target

Container Logs

WARN [client-schedule-recognition-task-0] [] c.t.s.o.c.d.s.t(): fail to recognize clients: [omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0, omadacId: d48bfc4bc792af06bb4e20ef62e3d8f1, siteId: 6a13fbbe1db8a937a2aa3d9f, originRecognizeCount: 0], response: OperationResponse(errorCode=-7103, msg=Device does not have an ID., result=null, auditLogReturn=null, auditLogReturnList=null)

MongoDB Logs

No response

Additional Context

No response

[mbentley]

I know that I have not seen this behavior. I would suggest filing an issue on the TP-Link community forums or contacting TP-Link's support team as they're more likely to be able to identify the root of the problem. That being said, I ran the message through Claude to get a guess at what it could be:

What's happening:
client-schedule-recognition-task-0 is a background scheduled job in the controller that periodically tries to run "client recognition" — this is the feature that fingerprints connected clients (guessing device type/OS/vendor, sometimes labeled "Client Identification" in the UI) so they show up with nice icons/names instead of just a MAC address.
It batches up a list of clients (identified by omadacId — your controller's ID — and siteId) and sends them off to be recognized, likely to TP-Link's cloud recognition service.
originRecognizeCount: 0 just means these clients haven't been successfully recognized yet (recognition attempt count is 0).
The response comes back as errorCode=-7103, msg=Device does not have an ID, meaning the recognition service rejected the batch because it couldn't associate the request with a valid device ID — essentially the call failed/was malformed on the server side rather than something wrong with your specific clients.

Practical takeaway:
This is almost always benign. It just means the client-recognition feature (which fills in device name/type icons in the client list) isn't successfully phoning home right now — could be a temporary TP-Link cloud-side hiccup, a controller not properly linked to a TP-Link cloud account, or an API/version mismatch between your controller build and what the cloud recognition endpoint expects. It doesn't affect actual network function — clients still connect, get DHCP, pass traffic, etc. The only symptom you'd notice is unidentified clients showing generic icons/names instead of "iPhone," "Windows PC," etc.

[michnixweiss]

Thanks for the explanations. Just to be clear, do I need cloud access for this to work properly?
If so, that explains it, because I'm only using a local account.
So it's just a feature to "make the client list look nicer." Is there a way to disable it?

[michnixweiss]

Its a feature, which needs among others cloud access.
I deactivated it at Global View > Settings > History Data Retention > Client Recognition

https://support.omadanetworks.com/us/document/109041/#_Toc211537466