krallistic

Awesome Idea, i always wanted to dig deeper into physics but wasn't sure where to start.

Could be a nice bridge to move off docker :P

Note that they used a pretty old+forked version of keras for the mxnet backend

I highly doubt that given the history between both companies... But would be funny to force a docker dev to merge systemd pr's...

If you really need "real-time-ness" TSDB will probably not work. You could go with sparkstreaming or flink but for that amount that will be overkill. Kafka scales quite nicely and is relativly easy to operate (compared to spark etc). But thats only if you really need to scale, for your current usage even kafka is a bit overkill.

How small where the box? normally kafka dosnt need much (4GB/8GB) and just uses the rest for page cache which can be sacrificed for performance.

Its worth mentioning that kafka only supports guarantees for ordering inside a partition, not globally. (nifty little detail :) )

just to clarify does that mean you should already have a work permit, or eligible to apply for the common visa's? I would expect the first but i couldn’t find any clear definition in the past since it changed from case to case