feat: rebroadcast block and tx hashes to sensor connected peers

[minhd-vu]

• add something like --broadcast flag that will instead of just logging that the sensor received a block or tx hash, actually resend it to all of the peers

[minhd-vu]

After deploying, there seems to be some latency issues, look into:

• increasing the number of concurrent goroutines
• refactoring the database writes to prevent contentions

[minhd-vu]

There may be tx pool partitions?

[minhd-vu]

Consider also checking the datastore costs after implementing the block cache.