Move to sequential UUIDs?

[bchrobot]

This is a way-in-the-future performance optimization

See:

• https://www.stevenmoseley.com/blog/uuid-primary-keys-django-rest-framework-2-steps
• https://www.2ndquadrant.com/en/blog/sequential-uuid-generators/

[dcdanko]

In my opinion this sort of performance improvement is probably 'out of scope' for Pangea. While it would likely result in faster insertion it also has the potential to increase code complexity in a fairly opaque way.

One of the goals for Pangea is to make it easy for computational biologists to contribute and maintain. With this in mind I think we should take an aggressive approach against any kind of code bloat.

The reverse is that we will likely have a huge number of objects in our database even early on. If inserting these objects presents a serious drag on performance Sequential UUIDs might be a good approach.