Prototype WebKNOSSOS deployment on AWS ECS

[aaronkanzer]

Detailed Description

As the team size grows for the LINC project, more and more users will be utilizing webknossos.lincbrain.org concurrently. While our deployment on our EC2 instance has been sustainable thus far, it is hard to know how resilient it might be into the future.

This GitHub issue is to capture the need to fully understand if a deployment strategy migration to AWS ECS might be beneficial in the long-run until we fully explore annotation frameworks powered by neuroglancer.

Cc @kabilar -- let me know your thoughts here -- I think this might be a good proactive measure to fully understand sooner-than-later

[kabilar]

Thanks @aaronkanzer. This makes sense, especially given the recent performance issues we have encountered. Perhaps we can explore this option now for ~1 day and then work on the implementation after the NIH meetings in February?

cc @satra

[aaronkanzer]

Thanks @aaronkanzer. This makes sense, especially given the recent performance issues we have encountered. Perhaps we can explore this option now for ~1 day and then work on the implementation after the NIH meetings in February?

cc @satra

@kabilar Works for me! We of course can continue to bump up the instance type which is fine too (just a bit more $, but we can also curb hours the site is online)

Just want to scope out how heavy of a lift the transition to ECS might be in case WebKNOSSOS starts to slow down substantially.