Docker daemon is running, but not detected while gptdeploy run

[adj009]

I am trying to run the created microservice locally using Docker. I am getting Docker daemon doesn't seem to be running. even though Docker is running.

I am on MacOS Ventura.

any way I can provide more info/ logs?

[joschkabraun]

Great that you tried GPT Deploy out!

We are updating the way we check if docker is running in here: https://github.com/jina-ai/gptdeploy/pull/54 Once it is merged, can you try it out?

[adj009]

Great that you tried GPT Deploy out!

We are updating the way we check if docker is running in here: #54 Once it is merged, can you try it out?

sure. waiting for that merge.

[adj009]

@joschkabraun i updated the pip module and tried to run the microservice.

[joschkabraun]

Okay, thank you!

Is your docker.sock here /var/run/docker.sock?

[adj009]

Okay, thank you!

Is your docker.sock here /var/run/docker.sock?

great catch. I was using colima for Docker, so it had its own Docker.sock path.

I had to make a symlink to /var/run/docker.sock

sudo ln -s /Users/home/.colima/default/docker.sock /var/run/docker.sock

---

Now, I am in different waters. :)

I dont want to stretch this issue here, if you want to close this, as the original issue was resolved.

[joschkabraun]

@JoanFM do you have any guidance on how to debug this issue?

[JoanFM]

can u run this with JINA_LOG_LEVEL env variable set to DEBUG?

[adj009]

can u run this with JINA_LOG_LEVEL env variable set to DEBUG?

[adj009]

also full output when I cancelled the run execution.

[JoanFM]

I would first debug the issue by locally instantiating the Executor "CHATGPTTOHUMANEXECUTOR". I bet there is some package wrapping some C code which is trying to do some wrong malloc or trying to deallocate already freed memory.

I would also try to look into the code or paste the code of the Executor here for it to be analyzed