Surface insufficient txn fees/ unavailability of snark work

[deepthiskumar]

The number of transactions included in a block depends on availability of snarks and whether or not they can purchased using transaction fees. block producers log this information as a summary and in the detail in the following two logs

1. Number of proofs ready for purchase: $proof_count Number of user commands ready to be included: $txn_count Diff creation log: $diff_log

   diff_log: [
   0: {
    available_slots: 46      
    discarded_commands: {
     insufficient_space: 0       
     insufficient_work: 0       
    }
    discarded_completed_work: {
     extra_work: 0       
     insufficient_fees: 0       
    }
    end_resources: {
     coinbase_work_fees: [2]       
     commands: [2]       
     completed_work: [2]       
    }
    partition: [1]      
    required_work_count: 0      
    start_resources: {
     coinbase_work_fees: [2]       
     commands: [2]       
     completed_work: [2] 
   }
   }]

2. Detailed diff creation log: $diff_log

Every node can compute this based on its mempool and so having a cli command/GraphQL that surfaces this information will help to better understand whether it is insufficient transaction fees or snark work unavailability that may be reducing the throughput.

[aneesharaines]

@deepthiskumar and @A-Manning to sync on a different possible solution

[deepthiskumar]

updated the issue description based on the discussion with @A-Manning

[mrmr1993]

For usability, might it make more sense to expose the data in client status or in another client command (e.g. client recent-fees)? For example, I could imagine output like:

Transaction fees
================
min fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
max fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
max fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
Snark work fees
===============
min fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
max fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
max fee: X.X (prev block) | X.X (prev 3 blocks) | X.X (prev 5 blocks)
Snark work
===========
available for next block: N
min fee: X.X
max fee: X.X
avg fee: X.X
works included: N (prev block) | N (prev 3 blocks) | N (prev 5 blocks)
Transactions
============
available for next block: N
min fee: X.X
max fee: X.X
avg fee: X.X
Transaction pool
=================
total available: N
min fee: X.X
max fee: X.X
avg fee: X.X

If this were generated from a GraphQL endpoint like we do with other commands, then this could easily be accessed programmatically from there too.