Hi,
In my Gitlab CI/CD Pipeline, my ephemeral containers in a Gitlab CI/CD build pipeline do indeed reach via curl -v $UNITY_ACCELERATOR/api/agent-health where $UNITY_ACCELERATOR endpoint is “unity-accelerator:10080”.
It’s curl-able and ping-able but despite have -EnableCacheServer and specifying the endpoint the editor straight up ignores it.
Also, do I bother with accelerator in the first place?
Thank you
Hello, can you explain a bit more what you mean by “my ephemeral contains indeed can reach via curl”
Do you know what IP address the Accelerator is at?
What is the value of $UNITY_ACCELERATOR and what is the IP and PORT you are specifying in the Editor for the connection to the Accelerator
1 Like
Hi,
When I access the terminal of any container, ephemeral or not, of the same network, $UNITY_ACCELERATOR can be reached. The value of that env is just the endpoint host name “unity-accelerator” with port 10080. It’s also reachable via builds.felsan.net:10080 from an external machine within my home network. Pinging and curling either unity-accelerator:10080 or builds.felsan.net:10080 succeeds. But because the job containers are in the same network unity-accelerator:10080 is sufficient.
Accelerator is listening on 0.0.0.0:10080
Hope that clears it up.
I am unsure if I’ve set up accelerator for insecure access properly.
Thank you
Do I need a separate file, like an accelerator.env?
Can you try as a test specifying the IP address explicitly rather than the host name?
If you open an Editor outside a container, is it able to access the Accelerator in a container?