chainlink sgxChainlink Web3 Summit HackerNode: Being a Chainlink Node Operator

alex kwiatkowski chainlink ishygddt background chainlink chainlink sgx Chainlink Web3 Summit HackerNode: Being a Chainlink Node Operator
chainlink sgx chainlink price prediction reddit Chainlink Web3 Summit HackerNode: Being a Chainlink Node Operator
Awesome. So I think we have everyone. So were going to have a Node Operator panel, and people are going to explain to us their background and how it has been like, their experience operating Chainlink nodes and how they see the future of the ecosystem moving forward. So were lucky to have We are lucky to have LinkPool,, and Certus One. So its probably going to be a very exciting panel, and lets get started. So I think the first thing for this panel would be already for you guys to intro yourselves, what you guys are doing, and really tell us a history for your company. Hello. Im Hendrik and founder of Certus One. We are a validator company, and we also provide security software for validator cooperators. And we just recently got to the space of Chainlink. And we actually started in the middle of last year, with the start of the Cosmos testnet. This is how we got into the whole validation space. Before, I had been doing security research and software development, and my co-founder and CTO has been doing security research and infrastructure development. So hes got really a lot of insight into low-level engineering of infrastructure. And we try to apply that to build very secure validation and Oracle infrastructure. Hi. My name is Kris, and Im responsible for the infrastructure of At we started with-- first of all we started with the Cosmos, and we are validating several platforms of blockchain. We started with Cosmos, then [INAUDIBLE],, and IRISnet, Tezos, and so on. Our companys products are all proof of stake. We were interested in Chainlink operating in Oracle because we feel that for sure is the future to solve the Oracle problem. I came from IT. I worked in IT since that was-- now is pretty much 20 years. And I was working with my first company in Switzerland in the Cybersecurity Network Action. And I joined the bitfish in November, and Im here to bring and say all the best practices of the enterprise in cybersecurity to grow our secure biodata in Oracle. Im Jonny from LinkPool. So LinkPool is a company that was formed in 2017 primarily as a chain-staking organization [INAUDIBLE].. In terms of my background, I come from sort of a long background of operations software development, like dev ops, gone from a role involing software development, operations engineers, and tied up with a lot more [INAUDIBLE] based roles. Were very excited about Chainlink and also Chainlink engineering as well. I think that the press of Chainlink is [INAUDIBLE] in the blockchain space, so Im excited. I introduced myself like a few minutes ago, but Im still Peter, the founder and CEO of We are a team of four, located in Mainz, Germany. We have about the same performance as you have, about 20 something years in IT. So around 90-something we started. Theres lots of banks in Frankfurt, so we had lots of projects with banks. Our co-founders, they started working some 10, 12 years ago in another company. Thats actually how we met. We are data service providers, fast growing. And we are very intrested in blockchain technology because I dont think there will be many chains. And they link somehow. They need the access somehow. Chainlink seems quite a good idea so far. And we just started. Its nice. Its really great to have voices from different backgrounds. So for instance, you guys were validating Cosmos and [INAUDIBLE] networks. LinkPool has really been focused on chaining since the beginning, and you guys also come from very different backgrounds. We really have large diversity, and its nice to have this kind of discussion. So I guess a question I would like to ask is maybe, how do you see the next node operator ecosystem growing in the future? Like, how do you see competition between different node operators? Basically your insights on what you predict the ecosystem will look like in the future. Yeah, I think well see a tremendous growth because I think what Chainlink provides us is something really unique and is one of the first times that we can actually have more or less trustless data. And Oracle provides that. Oracles the data provider on chain. And I see the tremendous growth potential for all of the products based on it. I think were going to talk a little bit more about that later. But I see that especially, because currently were just providing a little bit of data on Ether pricing. Thats what we primarily do, where weve been headed, to be aggregated by. So we all contribute to this one price-finding mechanism on the blockchain. But as weve seen, Chainlink has built great partnerships, and I see us providing much more interesting data in the future. And with more data, theres more limits. So [INAUDIBLE]. The ecosystem-- I think that probably competition will exist, as on the other blockchain. And I think that the most aspect that probably, within the change operator, we will try to push the effort would be to build a secure field, as we are talking about the Oracle problem that is a key factor to bring data in and out from blockchains. So for sure also there will be building tools for people. Yeah, I guess that most would be around building a lot of tools, useful tools, and reliable to help Chainlink to be more secure and trusting for people. The question is an interesting one. So I think in the future, especially when there is a lot more co-developments with Chainlink, especially with threshhold signatures and removing the linear increase in gas costs, [INAUDIBLE].. And the development of tools like that marketplace, for example, how it will function online, such as reputation services involved. Theres also been reputation services [INAUDIBLE].. It can be very interesting, especially when you look [INAUDIBLE] in the future as well. I think in the future [INAUDIBLE] were one of the first few. And then in the future when more and more people get involved in Chainlink you can start seeing many, many, many thousands of products essentially in the future. And there might be lots of people running, and youre going to get more levels of node operators. I think once youre first starting out, you have the cheapest nodes on the networks, and the people will have millions and millions of data points from their nodes. And then the concept of paying for decentralization. Youre paying for the reputation. Youre paying for the best nodes. For example, if youre going into [INAUDIBLE] between two large localizations, you might not want 1,000 nodes all staking a smaller amount of money rather than a handful of nodes who are KYC, who have very high reputable stake themselves through use of [INAUDIBLE]. I think its going to be interesting how different people will contribute [INAUDIBLE].. So we are specialized in [INAUDIBLE].. I think the concept of generally being a free market and the future of node, the providers themselves setting prices on random selection of the articles. Its going to trigger a lot of competition. [INAUDIBLE] I agree with you, what you said about the trust thats growing, so that in the evolution definitely there will be growth. I wonder how fast. I think that is depending on blockchain in general. So we have overestimated the goal from the last half year out of our founders perspective. I think itll go up for a while, so another half year, at least. So it is not very [INAUDIBLE],, but I think it will grow, definitely. And it might accelerate-- actually, Im pretty sure it will accelerate. Deep in my heart, Im a blockchain fool But were [INAUDIBLE]. So the question is, where are the use cases of the Oracles? Where is the data? That is interesting. And wheres the [INAUDIBLE] data also? Where there will be renew state or high usage scenarios, for example? So we differentiate on a spectrum. But I think we are still a small community, so I wonder how it will look like in five years or 10 years. Awesome. So you asked a question I wanted to ask, actually. And Rick actually referred to the different use cases. I guess my next question would really be, what kind of data and what kind of use cases are you the most excited to allow in Chainlink? I actually already mentioned it. I am a huge follower on DeFi prices and contracts that arent terrible implementations of our products in that space. For example, providing data about flights, about the weather, and helping people about prediction markets and helping these Oracles provide data. You can build very exciting things on top of these that will allow much more complex financial products than the traditional finance market would allow. And data is required in the end, but results are, for example, [INAUDIBLE]. Yeah. So I think this is going to be one of the use cases which I personally think is most exciting, and we see a lot of data already coming in. Working with the Honeycomb market, where a lot of APIs are now being added, and well provide that data. So Im really excited that we already start to provide that as of now. The same. Im hoping DeFi is the key and the future. If we think that, for example, if I am starting to look at traditional finance, I think that at some point, we will need all these data points that is part of normal life-- for example, [INAUDIBLE]. But we can also think about building on top of the wage things, you know, all these kind of things. And its crazy to think there is a lot of things that people are getting from the them. Why not also-- I was just thinking now in getting some-- not exactly about betting, but that type of thing. Its a huge market. Its a huge market. I think a lot of people, when you talk about Bitcoin and money, they often say, [INAUDIBLE].. I think weve talked about DeFi for even insurance products, and I think one of the massive benefits of this is that, certainly, youve got a lot of those traditional and typical products available to people who have not traditionally benefited from these things before. For example, farmers who are in third world countries, who certainly have to be able to purchase decentralized agricultural insurance. Its a massive case of just being able to develop and offer products that genuinely help communities in places that might not have the ability to be able to the sort of reach or to help those problems before that. Even in the case of logistics-- he was just talking about it a little bit ago-- of being able to track and incorporate more IoT data to create a background in which it can function and [INAUDIBLE] was verified with Chainlink, the delivery of shipment was done and settled by small contracts rather then more traditional routes. I think even we can sit on this panel now and talk about use cases, but then in two, three years time, I think were going to be very surprised. I think its always one of those things changing, because there are use cases where you get external data contracts [INAUDIBLE]. I agree, but also I think there will be much for enterprise data. So we are thinking about specific data from private chains, from [INAUDIBLE] chains. Our engagement with the Energy Web Foundation is supporting things with carbon dioxide certificates, as an example. You mentioned logistics and supply chain. Always there is many parties involved. We are talking to a very large company about cars and what you can do with renting cars, having accidents with cars, getting tickets, and all sorts of things that have to do with especially these rental cars. So if you have objects in several parties, I think there will be many business processes, and we are very interested in business, pretty good business processes. So once they move their processes and their enterprises, move the process on blockchain, interacting with other parties, I think then there will be reporting. There will be questions. There will be [INAUDIBLE]. And there will be limitations. Right perfect. I see that you have some great ideas there. I think probably if blockchain is directly happening on the traditional network, as you were saying, then we need all the data thats available in the traditional world. Thats why the possibilities are unlimited. A question I would like for you guys is, if we look at the ecosystems like Cosmos, a lot of creators are building, tooling around the ecosystem, right. So theyre building explorers. Theyre building [INAUDIBLE]. But Chainlink had the great contribution from LinkPool, which is basically a marketplace where a lot of operators can just lease themselves, put out the jobs, and contractors are using. And market creators can easily choose whos the best to basically service the demands. The question would be really, what kind of tools, if any, are you looking to provide, or looking to see being built on Chainlink? And how do you feel like the ecosystem of tooling will be evolving in the future in our ecosystem? Ill focus on what our focus actually is, which is security tooling. And I already had a talk about it, but Im going to say it again. For us, the security of the keys is extremely important, because thats what we actually rely on when we want to securely respond to requests through Oracle. And the security of the key is one of the most essential parts of it because thats the key used to actually sign the response, and thats the key thats potentially going to guard lot of stuff that will lie in our nodes-- and a lot of responsibility and liability as well. So what we will be contributing is a key management solution and potentially some security fixes and contributions to the Chainlink node to make operations more secure, especially looking at future staking. For a number of sides, I think that recently we are moving on the development side. Our dev team is growing, and for sure we want to focus on creating tools. We are already developing tools for other chains that we want to give to our customer, user. Lets call it customer. One thing that we are strong, actually, is the community because we are organizing meetup all around the world. We organize a meetup in Bangkok. We are organizing an event in Brazil about DeFi, and I think that the community will be also a key because it brings in people and make people know about Chainlink. And this data that can be used is also a big problem-- pretty much this. We are trying to focus also in developing tools that we will have people have to understand also to make things easy as possible. So this is a question which has caused us grief. With the marketplace thus far, weve been trying to dedicate a lot of time into building [INAUDIBLE] process. For me, personally, I think a lot of the advancements in terms of tooling will be due to protocol development. For many, if you want to use Chainlink, you need to go on and manually find nodes. It can be a certain type of job for you, and you need to make sure that youve got a certain amount of nodes that support it and the bulk of the jobs list and then the adapters. All the jobs are defined by the node operators themselves rather than requested by the user. For example, the service agreement protocol. What its formally known as is the method where the power of what is requested is given to the end user rather than the node operator. So I think that picturing our marketplace and especially a pre-payment protocol, if you go on, you can just say, I want to create a new service agreement, and the service agreement will last 12 months long. And we will select a random list of 500 nodes will all meet your criteria of what you want on that service agreement. And then you request, I want this API, and I want to be transformed in this way and then written about this contract. So rather than node operators defining what can be done, its more that the power is then given to the end users. And those guys as node operators could be in a 12-month long service agreement where we get pre-paid a certain amount to facilitate a contract. And I think when thats the case, and then especially when mixing our threshold solutions where gassing isnt as much-- the gas doesnt scale, its going to be extremely exciting because anyone can just go on and then say, were creating a request. And then youve got a thousand random nodes that all meet your criteria, and youve told the node operators to fetch this API rather than the node operators advertising that they can fetch that API. So it turns. And for me, thats where the protocol becomes a lot more mature and powerful. Well, we want to spot something node operators with dashboards so inside we have all the data ready to be levered. We can host the dashboards. We can create alarms and provide insight. For example, if you rent out a node, if you tool your node, if youre running nodes for customers, you will need bidding at some point, generate some reports on paper, even-- or at least PDFs. There might be taxes involved, and there might be even an auditor if youre doing enterprise. So we are interested in what we can provide. It makes a lot of sense. Im in a position where basically all the questions I wrote were answered already by you guys. We have a great crowd, so maybe if you guys have any questions to ask, we can start a Q&A. [SIDE CONVERSATION] I was just going to say, whats the biggest obstacle to decentralization in terms of the data feed that youre seeing? If youre thinking about enterprise-- for example, insurance. Does that make sense as a question? So in the sense of, if were thinking about a decentralized data feed into a contract, were all providing various forms of infrastructure to try and meet that demand. But from your perspective, what are the biggest obstacles to ensuring that decentralization continues? I think one big obstacle is probably regulation and liability. Because look, we have threshold signatures. We have multiple people verifying some data. Who is in the end liable? And thats the problem with corporate policy, and we wont be able to provide a simple solution at the moment because where you before had one counterparty-- which was this API-- its now potentially 16, 20, or more nodes. Who are you going to sue? I think this is a major problem because thats what these corporates think about when they implement it. Right. But its actually starting ahead of time. Who is going to guarantee anything? Thats why we are centralized company, and my basic pitch always is events bridge to a new world, which is decentralized through the old world, which is centralized. So you actually need some single instance or something that promises the features of a service. For example, liabilities, speed, whatnot. And then, actually, who pays or does something if they dont deliver? This is a complex thing. Its a mindset question, I suppose. So for decentralized data, I think one of the biggest mistakes that could be made is even if you request 1,000 nodes, for example, it all points to the same API. Because if that goes down then its game over, and never mind that youve got 1,000 of them. So I think a big hurdle will be finding a lot of reliable data sources that provide the data points that you need. Because when it gets to a point where end users request data, its not like youre going to be getting 50 nodes or using different APIs individually. Youre probably going to get 50 nodes. You are calling five or six or eight guys concurrently, and then theres some form of application done between those APIs and then aggregation between all of the different Oracles as well. So at least in that scenario, the room for error, I think, is extremely minimal. If you think about, like I mentioned, service agreements, one is the notation of a service agreement is incumbent parameters, and those parameters are essentially like an SLA to that agreement, like when youve got to respond by the allowed deviation of written data results, for example. So I think a lot of those hurdles might be tackled, not by the protocol changes in the future, but I think in terms of even what we do as developers and adaptive support and identifying different APIs, we need to make sure that we use as many as we can. And that the level of decentralization, especially if were in very important contracts, is extremely high to the point that its nearly trustless. Its trust-minimized rather than fully trustless. And I think, at least in terms of the liability and if anything does go wrong, I think at least because this is all blockchain-based, if we got to a point where a company even potentially sued node operators, at least you have then got the visibility and the traceability of all the data and what was returned back. So the fault can be pointed to the party upfront. And I think like at least what Chainlink are aiming to achieve is get to a point where the middlemen-- the Oracles-- are never at fault. I would just try to visualize this topic. And I think probably if you are talking about insurance or [INAUDIBLE],, it would be something that would not depend-- we would change country by country. So I think the problem will arise at kind of a node operator not certified but maybe recognized by a country-- probably as convenient. Not the full solution, but probably it can work. I mean, I guess I would have a question. What is your outlook in event of the TEE? So basically when another operator has a possibility to run software in a trusted execution environment? What do you think the repercussions will be, both on a regulatory and maybe even a technical point-of-view? Actually, we havent explored it too much when it comes to regulation to that part. But looking into tech will obviously have different requirements to hardware. What youre doing is using Intel SGX, and that will obviously limit the platforms that we can use. But were seeing more and more adaptations, I think. Thats not going to be a blocking factor. I just see a lot of just exciting possibilities once you have a trusted execution environment built into the application. But yeah, its obviously got to be a bit of a limiting factor-- maybe even pushing some people towards using dedicated hardware compared to cloud because SGX is not that widely available in cloud. Its coming, but its going to take time. Im looking forward to seeing traction on non-trusted execution-based Oracles and data sources and aggregators before we get into that stage because the tech is sometimes just not defined yet. We are working mostly behind the scenes, I can say. As I started working for, I bring in the company all the best practices I was using before. So we are approaching all the framework. We already started to apply to our company the ISO 27000. We are using-- for example, for the [INAUDIBLE] model we use, its like a [INAUDIBLE]. And we are trying to push and push company-wise also the NIST framework, CIS. We are mostly working behind the scenes. For sure, we are using the best practice also for security measure for the key management. But we are mostly behind the scenes, so thats why also we are not talking so much about security. But yes, most of that is behind because our processors are important also as well. If you have the best HSM, but then you get compromised on [INAUDIBLE]-- or your employees say, ah, this is the SSH key. Thats it. Yeah. For trusted execution environments and regulations, I think its going to be an interesting one in the future. Because you only have to look at, for example, that cash being removed from Congress in the UK as an example there. So you know, Congress got a new banking partner, and somehow the cash got removed because it didnt [INAUDIBLE].. So I think if you take that same example and then start thinking about trusted execution environments and generalized computation with [INAUDIBLE],, youre going to get to a point where youve got contracts on chain where the logic of the contracts and the actual outcome and what is being calculated and the inputs and outputs are completely shielded to the point where only the people who built the contract know what the contract is doing. So you could get to a point where youre getting millions of dollars being paid, but why and how and when. Its all completely shielded, and no one other than the people involved in that contract know. So its interesting, but I think this is going to have to be embraced because the reason this technology exists in the first place is to give power back to the people and the creators and make sure someone has access to these fine products. But I definitely agree with whats been said around about certifications and improving your integrity as a node operator. I think thats why weve seen some teams starting to [INAUDIBLE] formal verifications of people running nodes. So you get a stamp and say, Ive passed this Google test, and going through certain certifications is something theyre looking at doing now as well. I think thats going to be a big market, and I think especially, for folks who are running a marketplace, I think giving visibility of that information to all the end users and the contract creators is absolutely vital. I actually dont have anything to add, so it is the consensus. I dont think the hardware is a thing. It will come there. It will be there. We are still in the early days, so well see. I think its always the people. So there will be politics. There will be regulation, governments, social contracts, blah, blah, blah, where theres many things to come, I suppose. I think thats all for this Node Operator Panel. So thanks again, guys. It was a great discussion. And yeah, awesome. Thank you very much for this session, and that closes it for the day for the Oracle Hypernode. So well see you back tomorrow. Web3 Summit HackerNode talk 2019 hosted by Chainlink. Jonny Huxtable - LinkPool, Cristian Lucera - Bitfish, Hendrik Hofstadt -, Peter Eulberg - Chainlink is a decentralized oracle network that enables smart contracts to securely access off-chain data feeds, web APIs, and traditional bank payments. Learn more about Chainlink Website Twitter Telegram If you’re a developer, visit the developer documentation or join the technical discussion on Discord