profile_traits model does not perform well in Amazon Redshift

[josh-spratt]

• We have a large identifies table (~3 billion rows / 802 GB)
• We have basic query monitoring rules set up on our Redshift cluster to make sure we do not have any queries running that can negatively impact the performance of the cluster
• Any query writing more than 2000000 MB to disk is killed
• Without improvement to the performance of the profile_traits model, we will not be able to use this dbt package for our Segment Profile Sync data