Multiple Azure Container Apps stuck at "Activating"

Question

Multiple Azure Container Apps stuck at "Activating"

Greig 0

We have a number of our Azure Container Apps with new revisions stuck at "Activating". Today when we attempted to release a new version of an app we saw that the latest revision get stuck at "Activating". The status reads "There are no additional running status details at this time." - the app continues to serve the previous revision.

Upon checking further we noticed we had other container apps that were deployed last week that were also exhibiting the same behaviour. There is no pattern, some apps are .NET, some are NodeJS, completely different codebases and healthchecks.

Both revisions are running. I can use the console in the portal and connect and go into both, so the containers are up and running, it's just like the final step to switch traffic over hasn't happened
Healthchecks aren't failing
Nothing obvious in the container app environment logs
We haven't made any networking or wider Azure changes
Tried stopping app and restarting. In this case both the existing revision AND the new one get stuck at "Activating" - the traffic continues to flow to the oldest one

Based on when we were last able to deploy successfully and the number of apps we now have stuck, I would say something possibly happened around the 11th or 12th June. Was there any updates or deployments to Azure Container Apps around this time?

As it stands now, we are unable to deploy any updates to any of our container apps.

Alex Burlachenko 11,610 Reputation points

2025-06-18T08:56:06.7033333+00:00

hi Greig, thanks for posting this,

if ur container apps are stuck on 'activating', its often related to traffic splitting or revision management. u mentioned healthchecks are fine, so thats good! try forcing a full traffic shift to the new revision. go to the container app, hit 'revisions', pick the new one, and set traffic to 100%. sometimes the backend just needs a nudge )) https://learn.microsoft.com/en-us/azure/container-apps/revisions#traffic-splitting. Pls check if ur container app environment has enough resources. low cpu or memory can cause weird delays. u can peek at the metrics under 'monitoring' in the portal. if its maxed out, scale up a bit. microsoft recommends keeping at least 20% headroom here https://learn.microsoft.com/en-us/azure/container-apps/scale-app.

if u're using custom domains or ssl certs, give em a quick look. misconfigured certs can silently break deployments. this bites people on aws, google cloud, you name it. worth looking into, and one more thing, container registries can act up. if ur images are pulling from a private repo, make sure the credentials didnt expire. happens more often than u'd think )) https://learn.microsoft.com/en-us/azure/container-registry/container-registry-authentication.

U know sometimes the container runtime gets confused. redeploying with a tiny change (like a dummy env var) can push it back to life. works on kubernetes too, not just azure.

microsoft did push some updates mid june, but no known issues match ur case. still, their status page might have clues https://status.azure.com/en-us/status.

hope this helps! let us know if the traffic shift trick does the job...

rgds,

Alex
Greig 5 Reputation points

2025-06-18T11:17:07.46+00:00

Hi Alex,

Thank you for your helpful comments. Unfortunately, I've tried/checked all of these to no avail.

This morning I have found an error in container app logs that might possibly be an issue "unable to fetch metrics from external metrics api: the server is currently unable to handle the request (get s1-upstream_cx_active.external.metrics.k8s.io"

My current thinking is under-the-hood, KEDA/scaling is unable to ascertain that the new replicas are successfully running (there are) so it doesn't re-route the traffic to them.

I will investigate further, and also am waiting for a response from Microsoft via our provider.
Alex Burlachenko 11,610 Reputation points

2025-06-19T08:35:29.67+00:00

well well pls keep me posted.

rgds,

Alex
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-23T08:37:02.15+00:00

Hello Greig,

Could you kindly verify the status of the container app and provide us with an update?
Greig 5 Reputation points

2025-06-23T08:58:06.6+00:00

Hi Praveen,

Still have the issue unfortunately. Our provider is trying to understand the issue and get support from Microsoft. My hunch is that the underlying Container App Environment has been updated by Microsoft and this has broken something with the metrics that KEDA relies upon when scaling to know to divert traffic to the new revision.

It's now been over a week since the problem occurred and we have not been able to deploy new revisions to any of our apps so today we are going to start the work to create a new Container App Environment and start migrating our Container Apps to that to see if that resolves the issue.
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-26T05:02:03.8966667+00:00

hi Greig,

Could you please check the private message and provide the required details for troubleshooting the issue further.

Here is the reference link on how to access & data retention policy of private messages in Microsoft Q&A.
Greig 5 Reputation points

2025-06-26T07:57:54.9333333+00:00

Hi Praveen,

I have no private messages showing on this Q&A thread. Are you able to send it again?
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-26T08:59:51.53+00:00

hi Greig,

Please check now and provide the requested details?
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-27T07:10:45.4266667+00:00

hi Greig,

Following up to see if you have chance to check my previous response and please let us know with an update to assist you further on this.
Greig 5 Reputation points

2025-06-27T08:37:42.9+00:00

Hi Praveen. No matter what I do I don't seem to be able to see any private message. I get the email from you, but when I click the "See private message and reply" button it just brings me back to this page with no private message shown.

2 answers

Your answer

Alex Burlachenko 11,610 Reputation points

2025-06-18T08:56:06.7033333+00:00

hi Greig, thanks for posting this,

if ur container apps are stuck on 'activating', its often related to traffic splitting or revision management. u mentioned healthchecks are fine, so thats good! try forcing a full traffic shift to the new revision. go to the container app, hit 'revisions', pick the new one, and set traffic to 100%. sometimes the backend just needs a nudge )) https://learn.microsoft.com/en-us/azure/container-apps/revisions#traffic-splitting. Pls check if ur container app environment has enough resources. low cpu or memory can cause weird delays. u can peek at the metrics under 'monitoring' in the portal. if its maxed out, scale up a bit. microsoft recommends keeping at least 20% headroom here https://learn.microsoft.com/en-us/azure/container-apps/scale-app.

if u're using custom domains or ssl certs, give em a quick look. misconfigured certs can silently break deployments. this bites people on aws, google cloud, you name it. worth looking into, and one more thing, container registries can act up. if ur images are pulling from a private repo, make sure the credentials didnt expire. happens more often than u'd think )) https://learn.microsoft.com/en-us/azure/container-registry/container-registry-authentication.

U know sometimes the container runtime gets confused. redeploying with a tiny change (like a dummy env var) can push it back to life. works on kubernetes too, not just azure.

microsoft did push some updates mid june, but no known issues match ur case. still, their status page might have clues https://status.azure.com/en-us/status.

hope this helps! let us know if the traffic shift trick does the job...

rgds,

Alex
Greig 5 Reputation points

2025-06-18T11:17:07.46+00:00

Hi Alex,

Thank you for your helpful comments. Unfortunately, I've tried/checked all of these to no avail.

This morning I have found an error in container app logs that might possibly be an issue "unable to fetch metrics from external metrics api: the server is currently unable to handle the request (get s1-upstream_cx_active.external.metrics.k8s.io"

My current thinking is under-the-hood, KEDA/scaling is unable to ascertain that the new replicas are successfully running (there are) so it doesn't re-route the traffic to them.

I will investigate further, and also am waiting for a response from Microsoft via our provider.
Alex Burlachenko 11,610 Reputation points

2025-06-19T08:35:29.67+00:00

well well pls keep me posted.

rgds,

Alex
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-23T08:37:02.15+00:00

Hello Greig,

Could you kindly verify the status of the container app and provide us with an update?
Greig 5 Reputation points

2025-06-23T08:58:06.6+00:00

Hi Praveen,

Still have the issue unfortunately. Our provider is trying to understand the issue and get support from Microsoft. My hunch is that the underlying Container App Environment has been updated by Microsoft and this has broken something with the metrics that KEDA relies upon when scaling to know to divert traffic to the new revision.

It's now been over a week since the problem occurred and we have not been able to deploy new revisions to any of our apps so today we are going to start the work to create a new Container App Environment and start migrating our Container Apps to that to see if that resolves the issue.
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-26T05:02:03.8966667+00:00

hi Greig,

Could you please check the private message and provide the required details for troubleshooting the issue further.

Here is the reference link on how to access & data retention policy of private messages in Microsoft Q&A.
Greig 5 Reputation points

2025-06-26T07:57:54.9333333+00:00

Hi Praveen,

I have no private messages showing on this Q&A thread. Are you able to send it again?
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-26T08:59:51.53+00:00

hi Greig,

Please check now and provide the requested details?
Praveen Kumar Gudipudi 1,875 Reputation points Moderator

2025-06-27T07:10:45.4266667+00:00

hi Greig,

Following up to see if you have chance to check my previous response and please let us know with an update to assist you further on this.
Greig 5 Reputation points

2025-06-27T08:37:42.9+00:00

Hi Praveen. No matter what I do I don't seem to be able to see any private message. I get the email from you, but when I click the "See private message and reply" button it just brings me back to this page with no private message shown.

Answer 1

@Greig

Thanks for the update! Glad to hear the issue is now resolved, and we really appreciate you sharing the root cause and fix — this will certainly help others facing the same problem.

Issue: Multiple Azure Container Apps were stuck in the “Activating” state when deploying new revisions. While containers were running and console access worked, traffic continued to route to the old revision. Restarting the apps didn’t help — both new and old revisions would remain in "Activating", with no additional details in logs or health checks. The issue began around June 11–12, 2025.

Resolution: Removing the targetLabel property from the Bicep deployment files resolved the issue. After the change, new revisions deployed successfully and activated as expected.

Hope this helps others encountering similar deployment issues with Azure Container Apps!

Answer 2

Greig 5

Our issue is now resolved. In our Bicep files we were using "targetLabel", but we running in single revision mode. This worked no problems for 6 months, but there must have been some kind of update to Azure Container Apps on 12th June that then stopped this from working. Removing the "targetLabel" from the Bicep file resolved the issue.

Share via

Multiple Azure Container Apps stuck at "Activating"

2 answers

Your answer