when Interact with the Server, All connection attempts failed

[YinSonglin1997]

I start an App and Interact with the Server, when I print "hello", it reported the error shown in the picture below:

Then I start a script that can create an agent and interact with the inference server, it has the same problem httpx.ConnectError: All connection attempts failed.

What is the cause of this error? Could it have something to do with my server address? My server address is 10.0.1.227.

[MontaEllis]

Just disable ipv6

[YinSonglin1997]

Thank your for your reply, it works.
And I have another question, how do you add API Keys for Tools, I don't know exactly how to add, can you provide some specific steps?

[MontaEllis]

Actually, I don't know. I'm still trying to figure it out. I discovered that app/main.py can't process uploaded files.

[dltn]

Hi @YinSonglin1997, I'm curious to learn more about your setup – were you hosting the inference server elsewhere? How did you disable IPv6?

Re: API keys, they're pulled from environment variables. I'll add some documentation around that.

@MontaEllis, I'd like to help you debug why app/main.py can't process files – could you create a new issue with details?

[YinSonglin1997]

llama inference start --disable-ipv6

[pitimespi]

Is there a way to have llama_stack_client include a message about ipv6 in the ConnectionError message?