You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
On Fedora 41, when I go to interact with a deployed model via a podman-desktop playground I get the error "Inference server is not healthy, current status: unhealthy".
podman-desktop: v1.16.0
podman: 5.3.2
The container in use is podman-desktop-extension-ai-lab-playground-images/ai-lab-playground-chat-cuda:sha256:13e9a2a7103051e9eac1dd2c04f9be92f9f10b44770fe2eb8b55343bb4a989c3
In the pod's log I see.
{
"Start": "2025-02-04T17:50:17.521562596+11:00",
"End": "2025-02-04T17:50:17.667566383+11:00",
"ExitCode": -1,
"Output": "curl: (7) Failed to connect to localhost port 8000: Connection refused\n"
},
I am unsure if the following is the cause of the issue, however, when I inspect the pod via podman-desktop I see that the health check is
Running the first health check CMD-SHELL in the container terminal works as expected, the local API server is listening on 8000 and I get a response, however, running the second ends up with the error curl: (6) Could not resolve host: u003e curl: (3) URL using bad/illegal format or missing URL
Operating system
Fedora 41
Installation Method
from Podman-Desktop extension page
Version
next (development version)
Steps to reproduce
Open podman-desktop
Run an inference server (which pulls the ai-lab-playground-chat-cuda container)
Create a playground
Interact with the LLM in the playground - error recieved: Inference server is not healthy, current status: unhealthy
Relevant log output
Additional context
No response
The text was updated successfully, but these errors were encountered:
grantomation
changed the title
Inference server is not healthy, current status: unhealthy - Linux (Fedora 41)
Can't interact with playground - Inference server is not healthy, current status: unhealthy - Linux (Fedora 41)
Feb 4, 2025
Bug description
Hi,
On Fedora 41, when I go to interact with a deployed model via a podman-desktop playground I get the error "Inference server is not healthy, current status: unhealthy".
podman-desktop: v1.16.0
podman: 5.3.2
The container in use is podman-desktop-extension-ai-lab-playground-images/ai-lab-playground-chat-cuda:sha256:13e9a2a7103051e9eac1dd2c04f9be92f9f10b44770fe2eb8b55343bb4a989c3
In the pod's log I see.
I am unsure if the following is the cause of the issue, however, when I inspect the pod via podman-desktop I see that the health check is
however, when I perform an podman inspect from the command line
podman inspect <POD ID> | grep -C10 Healthcheck
on the running pod I see;"Healthcheck": { "Test": [ "CMD-SHELL", "curl -sSf localhost:8000/docs \u003e /dev/null" ], "Interval": 5000000000, "Retries": 20
Running the first health check CMD-SHELL in the container terminal works as expected, the local API server is listening on 8000 and I get a response, however, running the second ends up with the error
curl: (6) Could not resolve host: u003e curl: (3) URL using bad/illegal format or missing URL
Operating system
Fedora 41
Installation Method
from Podman-Desktop extension page
Version
next (development version)
Steps to reproduce
Inference server is not healthy, current status: unhealthy
Relevant log output
Additional context
No response
The text was updated successfully, but these errors were encountered: