Optimization: identify indexes that are heavily used that might benefit from location on faster storage devices
Postgre SQL supports putting tables and indexes on differing storage devices. This is an optimization task, to determine which tables and especially indexes, that it might be useful to store on a faster storage system (e.g. SSD or even even optane storage). The results of this issue would just be recommendations about how to optimize hivemind for maximal performance on servers that can be configured with multiple storage devices for the database.
Recommendations should be in the form of table/index name, data about size and relative size of the table/index relative to the database at the time of the recommendation, and some data about API queries where performance improvement was noted as a result of the change in storage allocation (with test conditions, such as types of devices used for the test). Most hiveminds will likely be in the cloud, so ideally the recommendations will relate to storage devices that are available on multiple cloud providers.
Ideally, we would want recommendations for at least the following configurations: 1) multiple storage devices of the same type, 2) multiple storage devices which include smaller high performance storage (this latter is probably the most interesting avenue of study, in particular would be good to have a configuration for a system with a combination of HDD and SSDs, or slow SSDs and fast SSDs).
Note that hivemind database queries are currently being optimized, so this issue should be blocked until initial optimization using a single storage device is completed, to avoid outdated results.