Everything you need know to Spin Up Your Own Rollup - Uttam Singh

Transcript

[Music] Okay, awesome. Uh, sorry for this. I just present uh created my presentation last night, but yeah, let's go. So, hey everyone. How's it going? So, just a quick room check before I get started. How many of you have already heard about roll-ups? No need to be an expert. Okay, that's good. That's good people. So today I will start with like the basics of what roll-ups are and I'm going to discuss a few things like for example there are stacks available out there to like build your own rollups. There are data availability layers there are couple of other things like there are zk optimistic rollup native rollups. So I'm going to uh make sure that you understand everything by the end of this

presentation and at the end if you have any questions I will be here and if you have time we can discuss all those questions. So who am I? I am a senior deell engineer at alchemy. So for those of you who don't know what Alchemy is, Alchemy is like a complete onchain development platform. We provide rollup as a service, wallets, APIs, RPC, subgraphs, whatever you need to build your onchain application. And I completed my research from University Montreal. I am uh a Ethereum educator. I keep creating content around Ethereum. So you can give me a follow on Twitter. You can check out my content there. Previously, I've worked with companies like Flair, Polygon, and Binance. So here in this space for a while now. So four years ago the Ethereum community took a bold move to move to this rollup ccentric road map which basically says

that we will move to rollups to scale the Ethereum L1. So how these rollups actually work is that uh it processes the transactions of chain and then aggregates them and bundles them and then submits that bundles to the Ethereum L1 and by this it it inherits the security of the Ethereum L1. So all the major activity happens on these rollups while Ethereum which is the L1 blockchain becomes more robust and settlement layer. So there are different components while building a rollup. So unlike Ethereum L1 or any other L1 uh roll-ups actually don't implement

everything like settlement, data availability or consensus. They outsource the management of data availability and consensus to like any other L1 and uh we will compare different things and we will also compare data availability layers in the future. But as of now data availability is like rollup transaction the history of all the rollup transaction that is submitted to the Ethereum L1 and consensus is basically refers to like the integrity of global transaction that is like once the transaction are submitted to L1 they can't be changed. So this is what consensus means. So uh in rollups one of the important major component is a sequencer. So sequencer is responsible for bundling the transaction and submitting it to the

Ethereum L1. So there has been controversies that uh because most of these roll-ups as of now have just centralized sequences. So few people say that it's a bug. Few people say it's a feature uh and there are uh networks like espresso which are like uh for decentralized sequencing. So you can even use that. But even though these roll-ups have centralized sequencer, you can still force exit to the L1 if in case the sequencer is trying to censor you or not trying to push your transaction. So as of now there are mainly two types of roll-ups that are available. The one is the optimistic roll-ups and the other one is ZK roll-up. So optimistic roll-ups uses

fraud proofs and the ZK roll-ups uses validity proof. So as the name suggests, optimistic rollup assumes that everything you submit to the Ethereum L1 is kind of correct, but they do have like a 7-day challenge period uh where if you think that something is wrong, if there is and there is any objection, you can uh submit a proof. That's where this thing called fraud proofs comes in which is kind of a proof that is provided by like a verifier which attempts to run uh those subset of those transaction again and then make sure that it validates them. So you can see here that uh if there is no dispute then the dispute

periods ends and the rollup transaction state is finalized. But if there is a proof and there is something wrong then in this case like the verifier is like the validator uh stakes is gone and u and it requires like a new state change proposal and after that you can even see that if user submits invalid fraud proof then also that the fault proof is like uh dispute uh removed and the dispute period continues. So we have a end of a finalization state. So a few things about uh this is that um most of the time these fraud proofs sometimes are not permissionless sometimes it is permissionless because there is a thing called delay attack. So

what is that that people what they can do is they can submit spam proofs in this way there is no finalization of the state right so that's why it's important that you solve that error uh most of these roll-ups I think initially have solved this and have permissionless fraud proof but uh yeah and then coming back to ZK roll-ups uh with ZK roll-ups the thing is like uh very different because uh you have this thing called fraud proof so whenever you submit commit a straight transaction uh to like these batch of that batch of these transactions to the L1 uh the correctness of that proof is uh provided by this complex mathematical proof which is a ZK proofs or a validity proof. So

for those of you who don't know how ZK works is that there is a verifier, there's a prover the verifier creates a proof and the prover and the prover actually proves that that it is correct right so it's like and you don't reveal any of those information so that's how like the validity proof works in and with zk the problem has been that it has been expensive to use and it is very hard and expensive to like create these proofs but as the time passes by it is becoming more and more easier to like create ZK proofs and it is becoming co low cost and recently like uh saxen team uh introduced this realtime Ethereum proving which is like ability to proof the Ethereum mainet block in like less

than 12 seconds and uh Google is planning to use like ZK proofs as well in their identity wallet or something like to showcase your identity wallet without revealing or showing your ID card right so ZK space in overall is growing but that being said all the Top five rollups by TVL as of now are optimistic rollups. So what we might change see in the future that things are changing. So this is like the architectural diagram of how any optimistic rollup works. Right now this is an example of like op stack rollup. So as of now you can see that there are this is the rollup and then here is the sequencer which is try which posts all the sequence all the transaction that bundles to the Ethereum L1 and then

there is a smart contract called at L2 output oracle contract which actually contains the straight route of the rollup against you know any withdrawals and bridging is done which is like a 7-day withdrawal period and the correctness of this contract is determined by the fraud proof game that we just saw like a 7-day challenge period And when we use ZK roll-ups, the correctness of this contract is uh provided by uh validity proofs, right? And validity proofs, the thing with fraud proof is that you don't need fraud proofs all the time. But with uh ZK proof, like validity proof is generated for every transaction, every batch that you submit to the Ethereum L1. And there are four other components as well that are responsible for other things. And

oper is like a thing which submits those transactions to the L1. And uh this is the blob. So if you're using Ethereum as a data availability layer, you can see here that uh in the block there is a specific thing called blobs which came after EIP4844. So you have a specific space to submit those transaction to the Ethereum blockchain. So you know Ethereum is now moving towards and designing theirelves to become a more robust settlement layer. So you you can see here that there is a space for those roll-up transaction to be submitted that is called blobs. So this in general is like um how optimistic roll-ups work in. Uh so what are rollup stack? So rollup stack is more like a chain framework

that refers to like set of software components. So think of it like an SDK whenever you build an app you you already have those tools and those protocols that work together to like create a function rollups. And the advantages of like using a roll-up stack is that there is standardization. So they already provide like set of consistent tools and interfaces and making it easier for like developers to deploy those applications and it abstracts away all the complexity. So you don't have to like deal with underlying complexity. You don't have to start from scratch because it's hard to build a blockchain. It's not at all easy. and modularity like they have uh modularity frameworks so you can you know customize your you don't have to use ETH as your gas fee and the most important thing is like ecosystem

support so whenever you choose one of these rollup stack you get ecosystem support from them which is like community support extensive documentation sometimes they also have like monetary benefits as well and uh all these stacks are uh designed with interoperability in mind which is making it easier to connect with other blockchains so if you're building your own blockchain and you're not using rule of stack you have to figure out interoperability you have to start from scratch so it's it's hard to build like uh basics from the scratch uh even blockchains like sonium which is like a blockchain of rollup of Sony they also used even optack to build it so I built this comparison chart to compare these two stacks which is like optack and

arbitram orbit these are the two famous stacks out there and most bo both most of them both of them are for optimistic stick rollups for building those. So in OP stack it's like they have MIT license and then arbitrum orbit has business source license and uh the priority ordering what does it means is basically ordering means how your transactions are ordered when you submit it to the blockchain. So OP stack has this thing called priority fee which is basically saying that if I give you this priority fee make sure that my transaction is on top of the block. So this is like a priority ordering but in arbitrum orbit they have first come first some ordering is is that that whenever you submit a transaction so whoever serves the transaction first submits the

transaction first it is going to order in that way they do have like time boost which is similar to priority ordering and then optack default time is 2cond but with flash blocks which is like an extension uh just like me boost you can cut it down to 200 milliseconds and then default time is in arbitram mode is 250 milliseconds and then as we discussed optack has super chain interoperability and then arbitum orbit has universal intent engine with opt you can't use your personal token as a gas fee you have to use ETH but with arbitum orbit you have that feature that you know you can use your own token as a gas fee providing utility to your token and uh both of these supports like forced L1

inclusion support so even the sequencer is like trying to censor you can exit via forced L1 inclusion And these are a few chains that uses these uh rollup framework. So now what is a D layer which is another component of the rollup ecosystem. It basically refers to that all the transaction data that you submit to the L1 is available. So it is important because it allows nodes to like independently verify those transaction without trusting any other entity and the roll-ups transaction data is submitted to these DAR which is available basically for anyone. In some of the cases this data availability

layer can be another blockchain but in some of the cases it can be just uh a network. So what are the data availability layers that are available right now in the market? So, and this is all like I'm not biased towards anyone because I don't work anywhere. So, I'm just giving you uh proper information and Celestia is like L1 blockchain as we discussed like Celestia and Ethereum is a specific L1 blockchain while DA is kind of a DA service like a network and uh important thing here is that you can check out the block time and the time to finality of these block and DA sampling and then data throughput how come like how much data throughput it is giving each second. So TLDDR here is that if you want like a if you're looking for

like security concerns and if you want like a secure data availability layer then you always choose like Ethereum because Ethereum is has been here for a long time it is like the most secure but if you're looking for something with more flexibility you want like uh faster finality so you can use like any other DA like Celestia DNL and then you can compare what are the chains available and then uh you can see that Eclipse using Celestia ID is used by like Mega this is one of roll up that is coming out faster and then uh Ethereum is like all the major rollups which is um base and optimism and arbitum coin arbitrum and world chain uses ethereum. Uh there is a quick difference by the way we will discuss in the future

that if you use another D layer other than Ethereum then you are you are called as Validium and OPM so you're not really a rollup. Uh there are rollup stages that we can discuss. So stage zero is basically on a full training vase. That means that if your rollup is in state zero uh it is effectively run by the operator but still there is like a source uh available software that allows you to like reconstruction of this state. So there is still transparency and verification and stage one is like limited security which is like enhanced rollup government. So the rollup is governed by a smart contract and there is a security council that is in place for like bug resolution but there is like uh fully functional proof

system with like a decentralized fraud proof. So anybody can submit those fraud proofs and validity proofs that we just saw. And the stage two is like no training wheels at all which is like fully decentralized and a smart contract like fully manages the rollup. And at this point like all the fraud proofs, all the validity proofs are like permissionless. anybody can submit those uh proofs and there is a still a security council placed but it is in case of like extreme governance attacks or anything like that. Uh so we don't we only have like 15 minutes and 10 minutes left. So I just want to make sure that I give you I I shot out all the terminologies that are available right now. So base rollups is kind of like a new concept that is by Justin Drake a researcher at Ethereum Foundation. So uh

it is a network where like the base layer which is the L1 which is the Ethereum is in charge of the sequencing of the transaction. Sequencing what we just saw how do you sequence your transaction. So as of now like there was a single sequencer who was doing that in all these rollups but with base rollups it is done by Ethereum L1. So searchers builders and proposers all that PBS proposer builder separation game that exists on like the Ethereum network that is going to perform the sequencing. So in this way like that centralized sequencer problem is going to be solved.

And then another one is like native roll-ups and native roll-ups by contrast are built to be like executed and validated natively by Ethereum itself. So instead of relying on like external verification native roll-ups use like an execute pre-ompile that allows Ethereum's code protocol to like re-execute and verify the transaction of these rollups. And uh as of now like there is no native rollup and there is one bass rollups I think tao. Yeah as far as I know. And then another one is ultrasound rollup. So ultrasound rollup is that when you combine the bass rollup and native rollup then you get like ultrasound rollup. And uh zk roll-ups we just discussed that uses zks or like uh fraud proof validity proofs as their

proof system and then optimistic rollup utilizes like fraud proofs as their proof system and then you have validium. So this is what I discussed. If you used external DA layer and if you use ZK proofs then you are known as validium. If you use external DA which is not Ethereum and uses optimistic proof then you are known as optimum and stage two rollups again it's like the end goal where there is permissionless fraud proof anybody can submit those fraud proofs and minimal oversight like only a security council is placed and then there are different layers data availability execution settlement and consensus. So I hope you guys know now what to do when you want to launch your own rollup. You have to choose all these things. Data availability layer uh what

framework do you want to choose and how what kind of modularity you want and why I run a rollup. So so launching your own rollup provides cost efficiency and increased scalability and customization because you can customize yourself and you are actually leveraging like e security and making it cheaper for your users. So it is more faster and reliable as you can see like world has 50% of op mainet transaction which is like which makes sense for them to like have another layer and you know uh have more efficiency and cost effective for your users and then customization and then financial incentives for sure. So whenever you run a rollup you get this sequencer fee and me and like if you use

native guest token you also have like uh you know people are using your token so uh it has a utility. So there are a couple of financial benefits and we have seen like uh things like Sony has launched their own rollup and then you know Coinbase launched B. So you know people are moving towards like launching their own rollup and why go with our rollup provider. So RAS platform is like a rollup as a service which allows any team to like run their own chain regardless of their technical expertise at low upfront cost. So why RAS? So if you want to like launch a rollup you need a full infer engineering team to like maintain that chain like 100% of the time and this also makes sure that your chain is secure and also in-house costs are always higher so you don't

need to worry about ongoing maintenance you can have uh another team who can manage your rollups while you can focus on like providing better UX to your users and you can also make sure that you are focusing on like application builder so you attract those builders right you don't have to worry about the infra right and there is time and cost savings for sure and customization and modularity so whenever you choose a RAS provider they have this customizations and they provide block explorer support from day one validator tools API indexer or wallet services so whenever you're building a chain you need these wallet services right account abstraction so it is provided from day one so you don't have to worry about all that right and why rollups really choose alchemy so we power like bigger apps like poly market

and open C which are like the most used onchain apps out there and uh on day one like we supports different stacks like data availability layer we have extensions like rollup boost we have ops succent opent is kind of a thing where you convert like any optimistic roll-ups to a zk rollups and it's very important to like choose a provider who is like reliable why because there have been uh incidents in the past where a rollup has gone down because you know uh something happened with the particular rice provider and uh recently also there was NFT API platform that got shut down and it created like a chaos among the developer community. So they ran away and they wanted another solution for that. So you don't want to choose a RAS

provider that could go down or uh you know it could go bankrupt in like few years because you are here for the long-term game. So it's very very important to like choose a right provider and we have like best-in-class data tools and APIs. We have like wallet services, account abstraction support and we you can choose the interop that fits your rollup best and yeah so everything you need for your rollup is with alchemy. So I think yeah that's it. Uh thank you for joining. We do have our grant program for like rollups deploying on arbitrum orbit. So you can scan here.

So if you're planning to build a rollup with arbitum orbit you can apply to that. And then for apps and rollups who are building using wallet services, they can also apply here. And uh yeah, give me uh a message, drop me a message on Twitter. I'm pretty much active there. And I think we have time for questions. So if any of you have any questions, happy to take it. Yeah, thanks a lot. And we have few minutes for the question. Here I'm sorry. Thank you. Uh what about app chains like do you need to create a roll app to create an app chain or is a separate thing? Yeah, you can customize your app chains too like we do support

app chains as well. So if you want even for app chains you can use like an external data availability layer. You don't have to like choose uh specific data uh like Ethereum itself. You can use Celestia or anything else. So you can have customization as well for your app chain as well. Yeah. Cool. I think that's it. Anyone else having a question? You have the time as of now. Awesome. Then thank you so much. [Music]