Repository navigation
bug: tasks main page fails to load new tasks for docker on docker setups #20250
Description
Is there an existing issue for this?
- I have searched the existing issues
Current Behavior
Bug report from a customer --
"Hey Coder team. A question on Coder tasks. Have you tested it with a sysbox/envbox based k8s pod? The thing I am seeing is when I spin up a Coder task, the pod spins up and our outer container starts, which is the container that kicks off the envbox inner container. As soon as the outer container starts, the Coder Tasks status page errors out. The error looks like this. If I go to my workspace view immediately after it fails, I can see the inner container is pulling its image and starting just like a regular workspace would. I see Claude Code starting up and getting ready. Its just that Coder Tasks already thinks there is an error and the task doesn't get executed, it sits showing idle.
See Slack for more info.
Relevant Log Output
Error
No apps found in task
Error: No apps found in task
at https://<coder-host>/assets/TaskPage-8xVlpnGH.js:11:29088
at Object.Df [as useState] (https://<coder-host>/assets/mui-D9D5a0wR.js:22:20303)
at w.useState (https://<coder-host>/assets/emoji-mart-jHVE9nvg.js:9:6319)
at Wn (https://<coder-host>/assets/TaskPage-8xVlpnGH.js:11:29021)
at kc (https://<coder-host>/assets/mui-D9D5a0wR.js:22:17376)
at jm (https://<coder-host>/assets/mui-D9D5a0wR.js:24:44650)
at Nm (https://<coder-host>/assets/mui-D9D5a0wR.js:24:40252)
at E1 (https://<coder-host>/assets/mui-D9D5a0wR.js:24:40180)
at $a (https://<coder-host>/assets/mui-D9D5a0wR.js:24:40033)
at Yc (https://<coder-host>/assets/mui-D9D5a0wR.js:24:36328)Expected Behavior
In a docker-on-docker setup, tasks should startup without issue.
Steps to Reproduce
tbd
Environment
- Host OS:
- Coder version:
Additional Context
No response
Activity
DanielleMaywood commentedon Oct 22, 2025
I've given this a complexity of 3 as I'd need to first re-create the environment (could be a 1 but depends on how easy it is to re-create)
david-fraley commentedon Oct 23, 2025
@DanielleMaywood fyi I messaged the customer again asking for their template/setup scripts and am still waiting to hear back. We can keep this in a holding pattern for the moment.
DanielleMaywood commentedon Oct 30, 2025
I've done a quick look into this with the limited information provided:
The error the customer is running into was removed in release 2.26. They might have misconfigured their template, but upgrading to a newer version should at least resolve this error.
mtojek commentedon Nov 12, 2025
@DanielleMaywood Let's close the issue then 👍