This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(310.4, 22%, 40%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EHP%3C/text%3E%3C/svg%3E)
I am using AGIC for my cluster. Recently I am observing the behaviour of health probe for each service we deployed is getting added to default probe in app gateway instead of creating a new one.
Initially I thought problem with specific namespace but with other namespace also it is getting added to default probe.
I am using these annotations in helm
appgw.ingress.kubernetes.io/backend-hostname: ""
appgw.ingress.kubernetes.io/use-private-ip: "true"
appgw.ingress.kubernetes.io/health-probe-path: ""
appgw.ingress.kubernetes.io/health-probe-interval: ""
I have tried this with multiple annotations to differentiate with the default probe which are as below.
appgw.ingress.kubernetes.io/backend-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-interval: "45"
appgw.ingress.kubernetes.io/health-probe-path: /api/status
appgw.ingress.kubernetes.io/health-probe-port: "80"
appgw.ingress.kubernetes.io/health-probe-timeout: "45"
appgw.ingress.kubernetes.io/health-probe-unhealthy-threshold: "3"
appgw.ingress.kubernetes.io/use-private-ip: "true"
kubernetes.io/ingress.class: azure/application-gateway
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(230.39999999999998, 86%, 32%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAK%3C/text%3E%3C/svg%3E)
Hi Hari Prasad,
Welcome to the Microsoft Q&A Platform. Thank you for posting your query here.
Based upon your query, It looks like the health probes are being added to the default probe in your App Gateway because the annotations in your Helm chart are not properly set this might cause issue. Specifically, the health-probe-path and health-probe-interval annotations are empty, so AGIC is using the default health probe settings.
To fix this, you need to provide values for these annotations. For example, set health-probe-path to a valid path like /healthz and health-probe-interval to a value like 10. This way, a new health probe will be created for each service with these specific settings.
You should also consider setting the health-probe-protocol annotation to specify whether the probe should use HTTP or HTTPS.
If an answer has been helpful, please consider accepting the answer and "Upvote" to help increase visibility of this question for other members of the Microsoft Q&A community.
The above annotation values been provided with values. I just posted here with empty values, In fact I added below annotations with values. Still I am seeing this issue.
appgw.ingress.kubernetes.io/backend-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-interval: "45"
appgw.ingress.kubernetes.io/health-probe-path: /api/status
appgw.ingress.kubernetes.io/health-probe-port: "80"
appgw.ingress.kubernetes.io/health-probe-timeout: "45"
appgw.ingress.kubernetes.io/health-probe-unhealthy-threshold: "3"
appgw.ingress.kubernetes.io/use-private-ip: "true"
kubernetes.io/ingress.class: azure/application-gateway
While deploying the helm I input the values. Infact I added other annotations as well to differentiate with default probe.
appgw.ingress.kubernetes.io/backend-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-interval: "45"
appgw.ingress.kubernetes.io/health-probe-path: /api/status
appgw.ingress.kubernetes.io/health-probe-port: "80"
appgw.ingress.kubernetes.io/health-probe-timeout: "45"
appgw.ingress.kubernetes.io/health-probe-unhealthy-threshold: "3"
appgw.ingress.kubernetes.io/use-private-ip: "true"
kubernetes.io/ingress.class: azure/application-gateway
Please fins the below annotations with values which I passed for the helm deployment
`appgw.ingress.kubernetes.io/backend-hostname: example.hostname.com`
appgw.ingress.kubernetes.io/health-probe-hostname: example.hostname.com
appgw.ingress.kubernetes.io/health-probe-interval: "45"
appgw.ingress.kubernetes.io/health-probe-path: /api/status
appgw.ingress.kubernetes.io/health-probe-port: "80"
appgw.ingress.kubernetes.io/health-probe-timeout: "45"
appgw.ingress.kubernetes.io/health-probe-unhealthy-threshold: "3"
appgw.ingress.kubernetes.io/use-private-ip: "true"
kubernetes.io/ingress.class: azure/application-gateway
Hi Hari Prasad,
Thanks for getting back to us,
Could you please verify these things, Ensure there are no typos or errors in the annotation keys and values. Confirm that the annotations are compatible with the version of AGIC you are using. Make sure the Application Gateway reflects the changes; sometimes there can be delays in applying updates. Check for any conflicts between annotations that might override your settings. Review the AGIC logs for any errors or warnings that could indicate what’s going wrong. Verify that your Ingress and Service configurations are correctly set up and align with the health probe settings. Ensure the Ingress is in the right namespace and that the ingress.class is set correctly.
If necessary, redeploy the Ingress resources or restart AGIC to apply the changes
If an answer has been helpful, please consider accepting the answer and "Upvote" on my post to help increase visibility of this question for other members of the Microsoft Q&A community.
We observed that issue is specific to one cluster and we could able to deploy that with another cluster with the same templates and it works perfectly there. We are planning to restart the AGIC, Will this affect any existing configuration in the AGIC. If yes, Any idea on what could be the downtime?
Hi Hari Prasad,
If it is working with new cluster there is might issue with old one and if it is working with new one & before taking any action, I will recommend taking a backup of the current configuration before restarting the AGIC. This will ensure that you can restore the configuration in case something goes wrong during the restart process. Additionally, it's a good idea to test the configuration after the restart to ensure that everything is working as expected.
Please follow below safety measures:
Please note that the actual downtime may vary depending on the specific circumstances, and it's always a good idea to plan for some extra time to account for any unexpected issues that may arise.
Tag me in comment, If you have any query, If an answer has been helpful, please consider accepting the answer and "Upvote" on my post to help increase visibility of this question for other members of the Microsoft Q&A community.
Hi Hari Prasad,
Just I wanted to see if you had to chance to review my comment If you found helpful and the issue was resolved. Please click an "upvote" & accept the answer on my post to help increase visibility of this question for other members of the Microsoft Q&A community.
Hi Hari Prasad,
Just I wanted to follow up if you had to chance to review my comment If you found helpful and the issue was resolved. Please click an "upvote" & accept the answer on my post to help increase visibility of this question for other members of the Microsoft Q&A community
Thanks for all the feedback. After trying various solutions, restarting the ingress pod turned out to be the effective option.
Hi Hari Prasad,
Thank you for replying back and providing an update.
If your issue was resolve, Please click an "upvote" & accept the answer on my post to help increase visibility of this question for other members of the Microsoft Q&A community.