The extension finds the models, but cannot connect to them

[AlbertoSinigaglia]

Before submitting your bug report

• I believe this is a bug. I'll try to join the Continue Discord for questions
• I'm not able to find an open issue that reports the same bug
• I've seen the troubleshooting guide on the Continue Docs

Relevant environment info

- OS: ubuntu 22.04
- Continue version: latest
- IDE version: 1.98.2
- Model: any
- config:
  
{
  "embeddingsProvider": {
    "provider": "ollama",
    "model": "nomic-embed-text",
      "apiBase": "http://127.0.0.1:11434/"
  },
  "models": [
    {
      "title": "Qwen 2.5 Coder 32b",
      "model": "qwen2.5-coder:32b",
      "apiBase": "http://127.0.0.1:11434/",
      "provider": "ollama"
    }
  ],
...

Description

The extension can find the models with no problem:

However, as soon as I pick any of them, and start chatting, I get this warning:

And if I press "start ollama", I see that it tries to run open -a Ollama.app.

Now, the setting is the following:

1. I'm running VScode from a Mac
2. I'm connecting to a remote host via SSH
3. Ollama is installed in the remote host (so say, if I open a terminal from that window, ollama list works fine)

To reproduce

not sure(?)

Log output

[Extension Host] Error: Unable to connect to local Ollama instance. Ollama may not be running.
    at customFetch2 (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:122240:23)
    at processTicksAndRejections (node:internal/process/task_queues:95:5)
    at withExponentialBackoff (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:118027:27)
    at Ollama2._streamChat (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:130989:26)
    at Ollama2.streamChat (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:122528:32)
    at llmStreamChat (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:607168:23)
    at Th.handleMessage [as value] (/Users/albertosinigaglia/.vscode/extensions/continue.continue-1.0.4-darwin-arm64/out/extension.js:619543:29)

-----

Unable to connect to local Ollama instance. Ollama may not be running.

[edashbolt]

@AlbertoSinigaglia for me this issue was because I was on a corporate network (ZScaler) and running Ollama in a docker container (using WSL) that didn't have access to the certificate authority. I fixed this by copying the ZScaler Root CA cert into my linux environment and running update-ca-certificates, then adding the volumes to my compose file like so:

ollama:
container_name: ollama
image: ollama/ollama
ports:
- 11434:11434
volumes:
- /opt/ollama:/root/.ollama
- /usr/local/share/ca-certificates:/usr/local/share/ca-certificates
- /etc/ssl/certs:/etc/ssl/certs
devices:
- nvidia.com/gpu=all
extra_hosts:
- "host.docker.internal:host-gateway"
restart: unless-stopped