JKG - using prespawned kernels

[Kevin Bates]

Hello Community,

I'm interested in using JKG's kernel pool of prespawned kernels that can then be used from a Notebook/NB2KG client.  While taking a closer look, it seems like the prespawn_count parameter (and pooling functionality) is closely associated with seeded kernels.  That, coupled with the fact that I only find reference to prespawn_count when notebook-http is discussed, leads me to believe that the use-case I'm looking for (i.e., the ability to 'attach' a prespawned kernel to a notebook's initial startup) isn't a current use-case of kernel prespawning/pooling.

Could someone please confirm if that is correct or provide further insight what I'm missing?

Thank you.

[Roland Weber]

Hi Kevin,

until the notebook is known, we cannot tell what type of kernel will be needed. The notebook-http mode has a notebook in the configuration, so it knows which type of kernel to start. To achieve what you are thinking of, you'd need separate pools of prespawned kernels for every kernel spec. It doesn't come out of the box, and I don't think it should - too wasteful on the resources. But you could implement your own kernel manager to do it.

cheers,

  Roland

[Kevin Bates]

Thanks for the response Roland - this is exactly what I was looking for.