Kevin Bates

Hi @jtpio. I'm sorry, I don't _currently_ have the bandwidth to revisit the jupyter-releaser stuff (I've used it once, probably late last year) and wouldn't want to mess things up....

@jasongrout - are you suggesting the hash computation be performed by the client (lab)? That would prevent adding the hash value to the model and dealing with the impact to...

It seems like you must have some network issues. `check_is_active_rm` merely hits the RM's `cluster` endpoint. Once you get the RM instance (after 20 seconds), do you continue to incur...

Good points @dimon222. I definitely think _if_ we were to add any kind of caching, it should be optional (i.e., configurable). I think understanding the current scenario is more warranted...

I agree - definitely a Hadoop issue. I suppose we could try to catch and log (or convert) as text, but the calling application can just do the same so...

Sorry for not seeing this sooner David. This is a bit perplexing because this particular message about a missing `local-provisioner` is essentially an invariant condition since the `local-provisioner` is packaged...

Hi David, I suspect the two Windows issues are related since the fatal exception is occurring within the Windows-only portion of attempting to write the connection file in a secure...

FWIW - this is what Enterprise Gateway does. However, I'm in the process of updating it to use a [single response address](https://github.com/jupyter/enterprise_gateway/pull/877) to which the kernel (launchers) respond, rather than...

Thanks for the response Johan - glad to see this is a single response address proposal as well. >> one suggestion I would make is to not overload the use...

> How will the client know which kernel is communicating its ports ? do you need like a kernel_handshake and a kernel_token? EG sends the kernel_id to the kernel and...