Staking Pools

As @werkof said pools will probably have higher requirements than the individual core nodes (those relative rich ones that earn enough ADA for paying the bills) for the infrastructure for the 24/7 availability.

Stake pools without an excellent uptime (it’s what the gaming theory and incentive paper absolutely want’s to achieve) will not only set at risk their own profit, but also that one from users who delegated their stake.

Fortunately, it was already said that these revenues - in case of missing block creation as elected slot leader - are lost (burned) and will not be divided up to other pools. This should favour that active pools will act as a team and not like adversaries.

However what happens if one “powerful” individual or group wants to become a new staking pool? Would it make sense for them to target one or some of the active ones and DDOS them down just when they become elected as slot leaders? When the victims start losing revenues they have their own costs plus a lot of angry requests from people who delegated to them. In such a case you can try to invest and find efficient defensive solutions or simply leave the field.
The very best incentive in such a case would be a well thought (moderate) revenue for pool operators.

When all pools communicate via broadcasts (that’s why 100 seems to be a good compromise) it would become harder for attackers to target certain fixed IP-addresses. The mentioned RINA communication is far away and a very long way to go. Very interesting in order to achieve performant, scalable communication between nodes by also having some sort of anonymity and protection.

but to be honest, it’s pretty much of guessing as long as the staking and incentive mechanism for pools is not fully defined and published.

I am expecting the Daedalus Wallets to be switched to some automatic pool selection mode at the begining, to initially secure the network. However, there are some other effective methods to switch to Shelley progressivelly, but this above is the easiest. Means, 1st step is to register the pools, 2nd is to switch wallets to the automatic pool selection (randomly select a pool, to prevent whaling), 3st stake holders select a pool or set up some core node.

2 Likes

See this is why I’m so hooked about Ada, the speculation over the future potential is so exciting! I can’t wait for it to become fully operational!

I think that the somewhere mentioned 1/k rule (beside w/ some other unknown ones) are for that. We saw some graph of some model’s result.

Yes, it is well known, that what we are writing here is only guessing.

this sounds like a smooth and smart idea to discover the best setup.
Also it would make sense for users to differentiate their stake to different pools.

Almost every cloud operator has docker service, and it makes sense, however nobody stops anybody to build theor private environments on premises. Frankly, I prefer cloud based solutions but the drawbacks, I can see, are the services price and the very few big providers (AWS, AZure, Google, IBM, HP, Rackspace). But, luckily, the nr of cloud providers that offer Docker are growing.

I undoubtelly thought, that it can be achieved by creating individual wallets for staking, and you just assign a wallet to a pool, but I can be wrong.

Is there a way to require/periodically verify a minimum level of “network, hardware, the variety of operating systems, DDOS prevention and monitoring systems, …” so that we have a uniform field of stake pools? If we fix the minimal configuration somehow, the pools can go ahead and compete on revenue generation. But the community will know choosing the cheapest option doesn’t compromise security/quality.

This is why Cardano is a superior blockchain. They think about incentives hard and deep. Simply genius!

Probably a stupid question, but is there a way to provide docker services on a decentralized platform?

I bet those who would like to earn ADA staking individually wouldn’t mind keeping their PCs on 24x7, if it was easy to do so. At some level, I suspect running your node is more cost-effective than delegating to a pool. If the protocol could check/verify your bandwidth and other quality metrics periodically and approve your node for minting, you could be better off minting blocks on your own. Especially if your energy costs are subsidized/cheap.

Somebody worked it out, assuming typical costs and rumoured rewards, and you need A LOT of ADA to make it worthwhile. And I’m getting the impression actual rewards are likely to be lower than was being rumoured for quite a while. If you’re interested search the forum, this has been discussed several times.

1 Like

If you meant decentralized platform as “public blockchain” then at the moment, it’s impossible. Might be some decades later.
If you meant it as “cloud computing platform” (which is decentralized already in the service’s POV) then they offer it already.

In theory, it’s possible w/ the current version of Ouroboros, as the selected stakeholders are already known at the beginning of the epochs. So, no one can stop you to shut down your PC and just turn it on and do a full resync (validating and downloading the missing blocks) before your turn (slot) to generate the block. However, the newer protocols will prevent this by obfuscating the list of selected stakeholders, means you will only know whether it’s your turn or not after the last generated block.

It just applies when you have enough stake to mint. So, it can only be properly calculated when the monetary incentives are known. Rough calculation is. If you have 1.4m ADA for stake, then you will be selected once in an epoch (5 days), means if the rewards for a month is smaller than your bills/costs then it not worth staking but delegating to a pool.

UPDATE: Just seen @vantuz-subhuman’s post, so pls check it for detailed answer.

1 Like

Who owns the servers in this setup? Did you mean “distributed” rather than decentralized?

The reason I asked my earlier question is that people tend to be viewing AWS as a centralization risk for running stake pools. So the natural solution for that would have been decentralized cloud computing, somewhat akin to what filecoin is doing with storage.

I saw that earlier. Unbelievable that you have to have 1M to have a chance at minting, but that will probably improve network bandwidth as people will delegate to stake pools. If I remember correctly the majority of ADA HODLers have between 10-10K ADA.

Do not get confused,. Sometimes, the distribtuted and decentralised terms in CS can be used interchangeably. However, I referred it to the decentralized cloud compute components that are cluster based, which are decentralized (in non-governance context i.e. In computation components context). I hate typing on tablet.

I would say 1-2% of the rewards for delegating is much better choice if you are not a whale. It is kind of trade off of convenience versus earns. I am old enough to prefer convenience nowadays, but if you woukd ask me 10 yrs ago I would have had said the opposite and not because of the rewards, but the technological challenge by being a hardcore Linux geek in that time.

That would be two levels of decentralisation, an extremely complex structure with, I’d think, unpredictable communications issues.