KEDA HTTP ADD ON 0.12.1 - Bad Gateway - health check function failed - table has not synced

[FlogFr]

Hello 👋

I am trying to make it work in a DEV environment in order to do cold start on a web service. My environment is on EKS with ALB.

I installed the helm chart kedacore/keda-add-ons-http version 0.12.1 with the following values in the keda namespace :

scaler:
  replicas: 1 # 2 pour la PROD

interceptor:
  replicas:
    min: 1
    max: 2
    waitTimeout: 60s

I have my 3 Pods up and running :

$ kubectl -n keda get pods 
NAME                                                    READY   STATUS    RESTARTS   AGE
keda-add-ons-http-controller-manager-6fbf584b68-9dwtx   1/1     Running   0          5h36m
keda-add-ons-http-external-scaler-75c5c4cd85-svhqq      1/1     Running   0          5h36m
keda-add-ons-http-interceptor-66f9fdbc7-2g5wc           1/1     Running   0          136m
keda-admission-webhooks-9c69b7f8-s7vrh                  1/1     Running   0          19h
keda-operator-7c4b548c45-4259d                          1/1     Running   0          19h
keda-operator-metrics-apiserver-7f59b767c8-rgdbt        1/1     Running   0          19h

I setup an Ingress that targets the svc/keda-add-ons-http-interceptor-proxy inside the keda namespace.

On the rgive-boa namespace (which contains my service), I installed the following HTTPScaledObject :

apiVersion: http.keda.sh/v1alpha1
kind: HTTPScaledObject
metadata:
  name: backoffice
  namespace: rgive-boa
spec:
  hosts:
  - boa.admin.test.rgive.com
  replicas:
    max: 1
    min: 1
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: backoffice
    port: 8000
    service: backoffice
  scaledownPeriod: 300
  scalingMetric:
    requestRate:
      granularity: 1s
      targetValue: 1
      window: 1s

KEDA HTTP ADD ON successfuly reconciled and create the ScaledObject :

$ kubectl -n rgive-boa get scaledobject 
NAME                  SCALETARGETKIND      SCALETARGETNAME   MIN   MAX   READY   ACTIVE   FALLBACK   PAUSED   TRIGGERS        AUTHENTICATIONS   AGE
backoffice            apps/v1.Deployment   backoffice        1     1     True    False    Unknown    False    external-push                     19m

Unfortunately, when I am trying to reach my service :

$ kubectl -n keda port-forward svc/keda-add-ons-http-interceptor-proxy 8080:8080
$ curl -H 'Host: boa.admin.test.rgive.com' -v http://localhost:8080/queue 
* Host localhost:8080 was resolved.
* IPv6: ::1
* IPv4: 127.0.0.1
*   Trying [::1]:8080...
* Connected to localhost (::1) port 8080
* using HTTP/1.x
> GET /queue HTTP/1.1
> Host: boa.admin.test.rgive.com
> User-Agent: curl/8.14.1
> Accept: */*
> 
* Request completely sent off
< HTTP/1.1 502 Bad Gateway
< X-Keda-Http-Cold-Start: false
< Date: Thu, 19 Feb 2026 16:14:34 GMT
< Content-Type: text/plain; charset=utf-8
< Transfer-Encoding: chunked
< 
* Connection #0 to host localhost left intact
Bad Gateway
$ curl -v http://localhost:8080/queue 
* Host localhost:8080 was resolved.
* IPv6: ::1
* IPv4: 127.0.0.1
*   Trying [::1]:8080...
* Connected to localhost (::1) port 8080
* using HTTP/1.x
> GET /queue HTTP/1.1
> Host: localhost:8080
> User-Agent: curl/8.14.1
> Accept: */*
> 
* Request completely sent off
< HTTP/1.1 404 Not Found
< Date: Thu, 19 Feb 2026 16:14:53 GMT
< Content-Type: text/plain; charset=utf-8
< Transfer-Encoding: chunked
< 
* Connection #0 to host localhost left intact
Not Found

From the logs of the interceptor :

$ kubectl -n keda logs -f keda-add-ons-http-interceptor-66f9fdbc7-467lv
2026-02-19T16:16:54Z	INFO	setup	starting interceptor	{"timeoutConfig": {"Connect":500000000,"KeepAlive":1000000000,"ResponseHeader":500000000,"WorkloadReplicas":60000000000,"ForceHTTP2":false,"MaxIdleConns":100,"MaxIdleConnsPerHost":20,"IdleConnTimeout":90000000000,"TLSHandshakeTimeout":10000000000,"ExpectContinueTimeout":1000000000}, "servingConfig": {"CurrentNamespace":"keda","WatchNamespace":"","ProxyPort":8080,"AdminPort":9090,"ConfigMapCacheRsyncPeriod":3600000000000,"DeploymentCachePollIntervalMS":250,"EndpointsCachePollIntervalMS":250,"ProxyTLSEnabled":false,"TLSCertPath":"/certs/tls.crt","TLSKeyPath":"/certs/tls.key","TLSCertStorePaths":"","TLSSkipVerify":false,"TLSPort":8443,"ProfilingAddr":"","EnableColdStartHeader":true,"LogRequests":false}, "metricsConfig": {"OtelPrometheusExporterEnabled":true,"OtelPrometheusExporterPort":2223,"OtelHTTPExporterEnabled":false}}
I0219 16:16:54.789447       1 shared_informer.go:667] "Warning: resync period is too small. Changing it to the minimum allowed value" resyncPeriod="250ms" minimumResyncPeriod="1s"
2026-02-19T16:16:54Z	INFO	setup	Interceptor starting
2026-02-19T16:16:54Z	INFO	Interceptor Version: 0.12.1
2026-02-19T16:16:54Z	INFO	Interceptor Commit: c03d9e20b4754ccd684af32c34d06c56a8a92e95
2026-02-19T16:16:54Z	INFO	Go Version: go1.25.5
2026-02-19T16:16:54Z	INFO	Go OS/Arch: linux/amd64
2026-02-19T16:16:54Z	INFO	setup	starting the routing table
2026-02-19T16:16:54Z	INFO	setup	starting the endpoints cache
2026-02-19T16:16:54Z	INFO	setup	starting the admin server	{"port": 9090}
2026-02-19T16:16:54Z	INFO	runAdminServer.pkg.queue.AddCountsRoute	adding queue counts route	{"path": "/queue"}
2026-02-19T16:16:54Z	INFO	runAdminServer	admin server starting	{"address": "0.0.0.0:9090"}
2026-02-19T16:16:54Z	INFO	starting the prometheus metrics server	{"port": 2223, "path": "/metrics"}
2026-02-19T16:16:54Z	INFO	setup	starting the proxy server with TLS disabled	{"port": 8080}
2026-02-19T16:16:54Z	INFO	proxy server starting	{"address": "0.0.0.0:8080"}
2026-02-19T16:16:54Z	ERROR	Probe	health check function failed	{"error": "table has not synced"}
github.com/kedacore/http-add-on/interceptor/handler.(*Probe).check
	github.com/kedacore/http-add-on/interceptor/handler/probe.go:63
github.com/kedacore/http-add-on/interceptor/handler.(*Probe).Start
	github.com/kedacore/http-add-on/interceptor/handler/probe.go:44

I am getting a Bad Gateway from the interceptor. And I see only 1 error log from the interceptor : "table has not synced".

What am I missing ? I tried to follow strictly the documentation from https://kedacore.github.io/http-add-on/ .

Thank you

[FlogFr]

I found the problem. The Pod needs to expose the port with named http.

[linkvt]

Hi @FlogFr , the error log was not really an error, it came from the health check failing at startup. We fixed this a few days ago and released it as part of 0.12.2 .
I'm not aware how it could be related to port naming so that change might have been unrelated.

[FlogFr]

Actually I had 2 errors. 1 was the one you fixed, thank you.
And the other one was that the SVC was not found because the port needs to be named "http" isn't it ?

[linkvt]

Which port and service do you mean?

[FlogFr]

The SVC you want to proxy with KEDA HTTP ADD ON + the container.
I didn't have the name "http" on both, and it didn't work.
After I added "http" on both : it works.