From conversation with Kelly Wilson:
I had a discussion last week with Nathan and Steve Tolman around scale and performance. When you can get back to scale testing we need the following questions answered:
1 - We need to find our failure point. What I mean by that is how many transactions can we write to the ledger before it just stops taking transactions due to its size. We don't have to write transactions as fast as we can we just need to know we stop functioning if the ledger hits a certain size.
2 - How much can you front load through the genesis file? This is what was tried before by having a domain_transactions_genesis file with 10,000 NYMs in it. If we can front load the pool by loading up the genesis file it would make it faster to get a large ledger (in theory)
3 - Can we get to 10 million transactions on the ledger? (We have logigear working on some scripts that will help in this area. I will get more details to you on those scripts early next week.