What would it take to be a node


#1

I love the idea of complete decentralization like torrent network. But in case of Cardano what would it take for me to be a viable node?

Will dedicated PC running 24/7 with 1Mbs link be sufficient?
How much ADA would I need to have for POS algorithm to chose me?

Thanks!


#2

I believe that hardware would be fine, but unless you have a very large amount of ADA, you will be chosen very infrequently, so it may be better to join a pool.


#3

Seems counter-intuitive. I thought that the whole point was to scale the network resources with each individual that joins the network rather than ruining 21 pools regardless of number of users using the network.


#4

More like - with each node that joins the network. Not sure how network would scale with an individual that just launches a light-wallet couple of times a day to pay for something.

But yes - the final plan is to utilise the resources of nodes to make network “harder, better, faster, stronger”. In comparison - no matter how many nodes\pools there are in bitcoin network - it does not scale accordingly. Nothing changes in the bitcoin network capabilities when a node joins.

There are planned to be at least 50 - 100 pools on initial stages, and after this - anyone will be able to launch its own node\pool. The @HazelMazel3’s answer was rather caused by the way you asked the question.

Anyone can run a node. Minimum number of ADA required is not yet publicly known, but it’s supposed to be relatively small, so running a node would be quite accessible.

PoS algorithm may chose you even if you have 1 ADA. The difference is in probability. Probability of your 1 ADA being selected among a total stake of few billions of ADA is rather small. So the question is not even applicable, cuz the correct idea is:

You need to have as big a share of ADA, comparing to the total stake available, as often you want to be selected. If you have 1% of total ADA staked - you will be selected 1% of the time, so approximately every 100th slot, so about 4 times a day. If you have 0.001% of total stake - you will be selected 0.001% of times, so approximately every 100’000 slots, so about 15 times a year.

Depends on PC, but 1 Mbs if stable - should be enough for now, imo.

@HazelMazel3 is right.

@Vashen, running your own node is noble, but it is true - if you don’t have quite a big stake share and not running a pool - then your reward will be small and scarce.

But the main idea is that you would be generally better off running a staking pool, than delegating to someone else’s pool. So if you’re planning to running a node - it’s just logical to also register as a pool, and then people would delegate their stake to you, smoothing down the scarcity of your slots. The whole system is optimised towards lots of small pools.


#5

Thank you again for a detailed explanation @vantuz-subhuman !


#6

Actually I have a follow up question for you @vantuz-subhuman
You have mentioned in other posts that the there are incentives to promote large amount of small pools. How much ADA does one need to be a viable small pool? 1000? 10,000? 100,000? 1,000,000+ ?


#7

Detailed research paper from Prof. Elias Koutsoupias and specific details like this will be available in following months (may-june prolly).

According to this:


#8

Also:

Delegation

Research paper - Vincent suggested re-evaluating the need for non-malleability, this might lead to simplifications in the address scheme. Apart from that, the paper is basically complete. Design document - Phillip made some additions.

Incentives

The main focus of the past week was the research paper. The experiments finally look very promising - stable behaviour that leads to a desired configuration. However, on the research side of things (i.e. the actual paper), the team still have to prove some theorems to align the theory with the experimental results.

From the latest report: