As the number of users on the network grows the amount of work required of indexer nodes also increases - this is obvious. What is less obvious, but equally understandable is that this relationship starts to look more and more exponential.
I.e. consider that “my likes and activity on the network influence what you see in your feed”.
Solution (Long Term)
We need to begin to shard the compute carried of indexing services amongst the pool of work to be done - this however is not what this proposal is about.
Solution (Short Term)
I would however like to propose that we increase the max delegation of 7mm for Discovery Nodes to match the 10mm for Content Nodes.
Network IO and Storage are cheap as compared to Memory (RAM). Discovery Nodes are very memory intensive. And as the aforementioned self reinforcing loop grows with each user, as do the memory requirements for table joins and materialized views.
Furthermore, based on the amount of requests on my nodes alone, it appears that the network as a whole would benefit from an increase in the count of active Discovery Nodes.
Perhaps by simply adjusting the delegation maximum to be equal (with that of Content Nodes), we would incentivize more Discovery Nodes to be registered.