A preview of MEV and mev-boost for validators with Remy, Yorick and Ladislaus

Transcript

hello everyone my name is remy i come from the east acre community and i'm i'm joined here today by yorick who's also um from the eavestaker community um and i'm also joined today by latest loss which is the cryptonomics kryptonomics enthusiast and he's also intuition hello guys how are you guys doing good hey guys great yeah thanks for having me thanks for the invitation it's our pleasure today we're going to talk about mev what it means um for solo stakers and like how you guys can even start trying mev on your own node and maybe on robston we'll see how it goes uh with uh yorick there

but first the big question um which we have to address here what is mev uh maybe yorick you can tell us what med is sure and before i get into what mev is um maybe this is what mav is so we got this tweet today um massive mev opportunity and i'm if i'm doing the the math here really quickly this is about 2000 eth paid out today in arbitrations alone so that's what we're talking about what is this who just got two thousand the ether as a miner crazy get some of that that is crazy it's good for the chain is this bad for the chain i can say it is a very

controversial topic so high level mev is maximal extractable value that's what it stands for it is payments that are made by ordering transactions in a block it's something that someone who creates blocks could extract right like in this case it's arbitrage um which arguably is a good thing but you could also front run someone um that's arguably a bad thing or a sandwich them also arguably a bad thing uh lotusloves you wanna add a little to that yeah i think the that's the the gist of it and then you could sort of also think of it and say uh it's not only the block producer but there's like many other actors within

the mev rims who sort of extract or let's say benefit from ordering transactions or inserting transactions so yeah that's that's maybe to add to sure and i guess that leads me to the next question is like how do you capture mev what are the different tools available like how can solo sticker get started um with mev right so we're gonna talk about flashbots mav boost should we maybe briefly talk about what is flashbots who are these people sure so so from my observations um i would characterize flashbots as some sort of

research collective actually so um and they are specialized in this topic mev and mev extraction on on many different levels so they really try to tackle the negative externalities of of mev to the ethereum blockchain and from what i can see they're quite uh ethereum value aligned which is which is pretty positive they cooperate with the ef and the broader community so um and they've also been like uh pretty um like the main builder so to say in the current proof-of-work like mev extraction game

right so um flashbots as you say the research organization mev for the longest time was happening but wasn't really visible nobody was surfacing this it wasn't clear that it was happening it was mostly mining pools doing this to boost their income and people weren't really aware that was happening and flashbots um kind of democratized it and by doing that also greatly reduced gas costs because now you don't have mev bots battling anymore it's been like 90 of mvp now in ethereum happens through uh flashbots so we're looking at a diagram here a system diagram of what this will look like post merge is what

we're talking about right right now the only way to extract mav is you have to be a block builder or be close to a block builder physically as well with latency in many cases right but you have to have a relationship what's the relation with the merge just to make sure that everyone understand this correctly sure so with merge block building changes from miners to stakers and so that includes you know people with their little um knock or raspberry pi at home and one single validator and they too can um now participate in this they don't necessarily need to go through a pool like in proof of work you literally with a single gpu you're not

creating blocks on your own it's impossible you have to go through a pool right this is not the case in proof of stake you can 100 uh build blocks on your own you get one every right now every two months on average right um that'll go down as more validators uh join it's just random uh when a block is built um it is absolutely possible to take advantage of this med and that gets into we don't want to go too too deep into the woods here right because it's easy to do um there's things you can google such as what is proposal builder separation in a nutshell say you are running a node

today on the beacon chain right you got your consensus client you got your execution client like you say you have a lighthouse or a tech who running and you got like a a nether mind or a basu or an aragon running you have these clients and they function perfectly happily well you would now run a third client a really lightweight tiny one called mev boost all mav boosters is a little connection piece of software it connects your consensus client like your takeover lighthouse to a relay and the relay in k in turn you see this here on the screen i hope gets blocks from a builder the builder

gets its blocks from searchers searchers are people who run you know big machine learning ai rigs to figure out the best mev strategies and possibly private orders they build a block give it to the relay the relay now goes to med boost and says i have a block that will pay you this much but i'm not telling you what's in it so the reason it doesn't tell you what's in it is if it did then you could just inject your own stuff and you know start front running people and we don't want so it'll tell you i pay you this much for it and do you want it that means this relay currently is

trusted is trusted by the staker here the consensus client validator for creating a valid block and it's trusted by these builders for fair routing that it will fairly route this to everyone right there are ideas in the flashbots github which is where we're at right now as to how this trust relationship can be um become less trusting but that's gonna take time you know in the beginning these relays are 100 trusted entities have to be uh so long story short if i can just you know say you have a proposal your consensus client here gets

a block from the execution client and one or more blocks from mav boost and then just chooses whatever pays best and then that makes it to the to the chain what did you want to say a lattice loss yeah i that's that's um pretty much what i alluded to earlier so you can see that there's several more actors than just the block proposal who's sort of the validator who's in the last uh entity in the in the right process here yeah exactly so um and then you have like uh you're starting with builders and sort of they are the ones who were sniping so to say for um for economic activity for lucrative economic activity on chain to sort of bundle transactions in a way that they benefit

from it and they send it to builders who are then even more specialized in order to have relationships with many um with many builders and then sort of uh create a blog craft a block and then they bid for the blog space they they sort of um yeah they they buy the block space uh the validators offer them and so there's a lot of like just mentioned a lot of um trust involved which sort of currently is is is solved um like with economic incentives so we're still like we could argue well um in blockchains we we love crypto economics so we would love to serve as

much as possible like with math but as you can see like for the moment like the as we've seen the ethereum network grow over the past years um with all this economic activity coming up um for now we just for the for the moment we try to not economically incentivize different actors to mitigate like the the bad parts of mev right so the bad parts being things like we glossed over it earlier right so say i have a couple trades and i find that somebody's buying ether at one price and it's being sold somewhere else at another price this is a classic arbitration uh opportunity and that's arguably a good thing but the other

thing i could do which is blatantly illegal by the way in the traditional markets um but because we're not regulated people do this anyway in crypto the other thing i could do is if i see that you're say selling eth i could front run you right i could do my own transaction first so your transaction price changes and then i do a transaction after again um so i'm basically taking money away from you because i know that you're going to execute this trade so i trade against you before you even execute the trade that arguably is the bad part of mev right this drives users away users will not want to transact in an environment where this is possible and so i believe like there's a bunch of

different solutions to this issue but um i just want to come back here uh to this diagram that you're still showing on your screen there's a bunch of different actors and a bunch of different boxes here and maybe people are getting confused because there's so many different um roles and separation of concerns but there's some good reason why let's say for instance search searchers are different or like away from like the validators right yes because this this takes a ton of compute to do right you can't do this on a solo staking rig not only does it take a ton of compute it also takes a ton of research so you're now looking at typically

artificial intelligence machine learning rigs that do this you're looking at potentially quads being employed once being high powered mathematicians who work in the financial realm who come up with strategies to find the best you know value extraction here and then offer it to the guy at home with their pie for right the demarcation line is right here i wish i knew how to do uh drawings in google i don't know how to markup a site i'm sorry um this bit here this runs in your node right the consensus client and value to the execution client mav boost all this runs in somebody's node and then all

this stuff here that's somewhere else those are largely centralized entities you know not sure how many we will have hopefully many you know maybe a dozen plus but these are largely centralized entities somewhere out there once this block makes it to the ethereum network validators will validate the block so it still has to be a valid block for it to be accepted into the chain that doesn't change the consensus doesn't change and yeah so we talked a little more about those uh searchers which are very specialized and need a lot of compute power maybe we could talk a little more about those builders and relay what what are these guys doing exactly

yeah so the first question you asked was actually great but because the the intuition would be like why why do we need builders uh anyway so why can't they just send their abundance to let's say the the reader directly or let's say to the validator directly and then you see okay a builder especially as as thorson alluded to it's also highly specialized that they they have many uh relationships they have a lot of relationships to many uh individual builders and so they send them like tons of bundles um and and and sort of which are profitable for or supposed to be profitable for the searchers and so this is so the first

the first argument for for why our builders here is is likely well it's some sort of a dos protection so for for validators that's this like the first thing and then of course you need the builder or many builders ideally um they yeah they sort of they um they have their algorithms running too like they sort of as they construct the block they have an interest to also construct the most profitable lucrative uh transaction from all the the bundles they get from the from the searchers so that's probably what i'd add here

and what about the relay what are these guys doing exactly or why do we need them right so the relay has a couple really important functions here one is the relay makes sure that the block that the builder built is actually a valid block so it's correctly formed right that's important if i get an invalid block as the validator um i don't get rewarded i basically it's if i put an invalid block on chain um the chain will penalize me as if i had been offline right because i didn't do my duty i didn't give it a valid block um i miss that proposal so that's that's towards the the validator there that's an important function

um towards the builders and searchers a really important function is that the relay be up right um it needs to be highly available um at all times they don't their their stuff is time sensitive right next next uh if there's multiple relays which again we want right if they submit their their built block to these relays and some of these relays don't um send the block to validators correctly then the block might not make it into the chain so then all that work that they just did to create the most profitable block it's it's not it's it gets thrown away right we're on to the next block we need a new bundle um the old one is no longer true

new transactions have come in um this is no longer the most optimized strategy and maybe one more important argument uh or point to add for the question like why is there a relay apart from what torsten has mentioned is um we have to we have to mention that um transactional transactions between all those different parties we see here uh sent in in plain text so that's coming back to like the uh like the trust assumptions here so basically uh a builder could steal mev from a searcher because he sees like where's where's the prophet uh where's

the prophet like he sees it on in plain text and then right if he's if the builder sends it to the validator like today the miner and and post-merchant validator i mean the the the validator could uh just unbundle sort of the the super profitable block and then just redirect the mev opportunity to himself so uh it's important to understand that this is this is a sort of a yeah not another idea yeah it's checked and balances and so the this relay like why is it here is to sort of flip this trust assumption that the validator could always steal like and unbundle transactions from a builder and sort of flip flip that and sort of

make it this um yeah this blinded block as as it's called and sort of um just sign something blindly in terms in terms of the block header and i don't want to be too technical here but uh offer other than than getting the block later um with all the with all the transactions in it so there's a lot of trust there like all of these actors have to trust each other's uh on some levels um is there anything we're missing in terms of like trust deception assumptions um like like what are the critical paths or like the possible

issues with one node not trusting another or like trying to screw another right on the right hand side here right they can start stealing uh mav from each other right like the builders could start stealing mav from the searchers and the relays even could start stealing any heat from the builders as i understand this setup right so that's on the right side here here on the left hand side where we have the validator they just really need to trust that the relay was truthful about two things

one the block is indeed valid otherwise i don't get paid at all and two the amount i will get paid for it because they will tell me that right but of course that doesn't happen until the block actually gets unchained right and so like what happens like if a really gives you an invalid block for instance if a relay gives me an invalid block i put that to the chain then every other validator sees the block now and um at that point while i just uh i just sign the block and then actually the relay places it on chain so i don't

ever get to see the thing until it's signed right once it's on chain everybody can read the block if it's not valid it's seen as invalid and then it's oh torsten's validator put an invalid block on chain well i guess that's an offline penalty and we have an empty block on to the next slot so just a regular protocol just a regular normal protocol you gave us an invalid block situation you know nothing terrible happens it's just i get penalized instead of um i get my block proposal plus med plus priority fees none of that happens now you know yeah and so i guess eventually like if validators notice that their relays are

not offering them valid blocks they might want to choose different relays right so that's your choices so that that there i'll flip it around to you remy um leading questions i've been told that eath staker might do some uh relay vetting any truth to that um that's what i heard too so if sticker is going to compile a list of relays so eventually if soul stickers don't know which one to choose from we will try to provide you guys with um a good list of relays that you could use with your own um mev boost software yeah and yet again that that is that sort of highlights the the notion that

at the current stage of how like protocol developers are tackling this mev problem so to say the idea is to yeah encourage competition competition among all participants here like competitive competition among searchers builders as well as relays and i should maybe also say a lot of slides this is a good point right this fits here there are ideas out there such as the shuttered beacon chain which you can find on earth research which say we should not just roll over and say mev is inevitable let's just make it less bad we can actually make mev impossible

if we have encrypted transactions that only get decrypted in the block after right things of that nature this is usually controversial when you look at the if research article for it you can really see tempers flaring on both sides on the side that says this need not be solved on protocol level let decentralized exchanges handle it and on the side that says this needs absolutely to be solved on protocol level because this will keep large actors from adopting uh ethereum for their transactions if they're at risk of getting sniped right so that's that's the outlook right so for the moment let like open market

forces and also users decide how to tackle it and then like later on as you mentioned earlier there might be in protocol solutions like proposes billiard separation in protocol and then maybe on the very far end of the time scale hopefully like um moon math like encryption transaction encryption technology coming into play so this is a nice deterrence sort of that that we think about it in in the greater github community um but for the moment like the picture we're looking at here for quite some time is is the like sort of best solution for the moment likely which we have so there's like a lot of research being done on all of these um ecosystems

and like all of this different structure that we can put in place to make it fair for everyone or interesting um for everyone in terms of cryptonomics and i don't know whether we've highlighted it enough like the the green rectangles here that's pretty much like all it takes um for for solo stages for solo validators right so um there's like if you're running a note at home and you're not even validating you don't need to run maybe boost the software like you're totally fine you follow you follow the chain um you have a complete overview over the network totally fine and then even a staker a solo stake or a validator that does not necessarily have to run mav boost the software it just there is

an economics incentive on the table for the validator to run meb boost and sort of like taken take into account also the maybe like the the trust assumptions we've mentioned earlier but there's an economic incentive so and then even if like uh what what wasn't also mentioned earlier even like if there is an invalid block it's like it's not the end of the world like the execution engine like your local consensus local execution client will always provide you with order transactions like it's it's like just as it is it has been uh like for forever right you will always have your your execution client guest nether mind whatever base you to to construct a

block and so of course you know you won't know that the block they got from maybe boost is invalid until you signed it and placed it on chain correct so use yeah so let's say let's say i'm a validator i want to get started with mev i want to join this um to be ready or eventually after the merge i want to take part in this what's in it for me what what kind of reward am i looking at here and i think we we might be going back to the example that we saw initially well that's a crazy example right let's let's let's just put that out here uh 300 e 700 eve um i'm not sure that will ever happen

again um somebody got a nice payday there for sure right um that's what's in it for you let's say that's not that amount but there have been some you know very shall we say optimistic guesses let's say that your current four and a half percentage is that what we have right now apy ish is that where we are yeah might with priority fees and mev um almost double mev largely is seen as maybe a 10 boost on your um on your earnings so not 10 points but 10 so if it's 4.5 before then now it would be five with

with mav-ish right something like that um half a point extra not wrong with that at all right and then you get your priority fees on top of all of that on average and so people can expect to have a little more reward with a little bit more reward exactly which gets paid which is immediately liquid by the way and gets paid to the same suggested fee recipient that priority fees go to so unlike the issuance you get so whenever i test correctly i propose correctly i get eth into the validator balance which is locked i can't get at that right until withdrawal sometime 2023 we hope but this stuff mev and priority fees is

immediately liquid goes to whatever ethereum address i give the client could be my hardware wallet could be a software wallet it could be a multi-sig it doesn't matter right yep very interesting here um so what has the impact of the flashbots mev uh been on ethereum maintenance so far blood so you want to talk about that it's more like the economic and gas yeah you but you touched upon it earlier already so so basically the like i guess it was in 2022 like when these famous sort of gas wars they were called as d5 sort of the d5 summer went on and we saw a lot of economic a lot more economic activity

on the chain and a lot more settlement value on ethereum like we realized okay this uh these uh searches and and and and bots which they actually where they were like increasing the the gas price because they were bidding like to get their transaction included into the next block by like by the means of of of the gas price so ethereum got clocked like it was terrible to use it they a lot of users were excluded like those were the negative externalities back in the day before like flashbots and then and this was like in 2020 2021 i guess it wasn't i guess it was like mid 2020 to late 2020 yeah okay

and so this is sort of when uh i mean flashback research or a research on having mev has uh yeah was there before but this was like when like a a tool like mev guest like sort of this side car to to the to the execution client geth was uh put into place like or in use and what what torso also mentioned is like nowadays like 90 i think 90 of all miners or even more than 90 percent of all miners use uh mev gas so this basically it's not 100 similar to what we like explained earlier which is the postmatch setup but this uh flashback that's the sort of the flashbots

impact uh or basically what would they have uh yeah um not mitigated but that was really much uh backed by them sort of to to have the overall ethereum network like be a lot more usable usable um because they they they program their software and it got used so one analogy people might hear is the dark forest that the mempool basically where all the transactions are is a dark forest with predators lurking those being the people who want to extract mev and you never know when your transaction will be taken down by a predator um flashbacks in some way you know shone a light into this dark forest and made

visible what's happening there and then kind of made it just not as economically feasible anymore for these hidden predators to just do their thing instead they they now work in the flashbots ecosystem for most of them right you said for most above 90 is is the number i've heard um i'm sure we'll get in the comments some notes that were completely wrong and would love some links to how we're wrong good point there do you expect the a lot of these uh validators are the stakers um after the merge to actually set up flashbots or any other meb solution like do you have any kind of estimate i have

my own yeah i think it'll be quite popular i know that the the larger staking pools will require it of their um node operators and then for solo stakers or people who run like a rocker pool it'll be fairly easy to set up an evb boost is really lightweight it's not hard to configure i expect that people will go free money you know now you might have a strong ideological dislike of mav in general and you're like uh half a point is not worth my soul um

that's fair you know but i think for most people it'd be like uh maybe i don't like it but it's free money good point yeah there will be ethical questions like moral questions there but for most people we expect them to actually set up maybe and use it as part of their value their setup right and as far as these these ethical questions go um again the large majority of mav or so flashbots tell us this is also controversial as to what's actually true is arbitrage is after the fact you sold a barter thing and i can get it cheaper or more expensive somewhere else and so i will arbitrage against you that's a good

thing that means that markets stay efficient you know that's what's supposed to happen yeah we we basically we touched a lot upon the the negative externalities but bosnia's totally right like there's there's a good part of of mev uh in terms of like making markets on ethereum or on blockchains in general more efficient and that's like why we also like benefit to some extent from from searches activity um yeah right so what does an efficient market mean for me and a lot of sloughs please get me right here because i am not the economics guy right i just run nodes but i can give an example of where i see an inefficient

market i have a token and it exists on a layer two and i now want to um move this token into ethereum this happens for me about once a week right what i find is depending on the layer two the premium of swapping that into each on that layer two is so high it actually makes sense for me to bridge it to ethereum first wait seven days and then get my eath um yeah and so in a more efficient in a more efficient market um arbitrage would make sure that the token to eth price is just about the

same no matter where i go and where i transact again i also feel i have to i have to highlight again like mev mv boost it is a side car to continue as flashback says it so like if we if the like the the the the listener things like we're talking like uh it harms the network of mev maybe we we still have to highlight its um consensus is not uh census is not impacted correct it's not impacted and basically like the the the idea is like the the more economic activity you have and then uh you just you just mentioned a case where

markets can be inefficient and then sort of the there are actors who who try to compensate for it but um right uh the chain like runs along um 100 does so it's a hundred percent about user experience right how how is my experience trading do i have an efficient market and am i at risk of my my trade getting sniped by someone else i guess this will lead us to our next question about the actual mav boost side card do you think it'll be ready for the merge uh york oh boy um there was a big discussion about that at acd just now last friday and um i

think the the consensus was it's not the end of the world if we merge without it i'll reserve judgment i i'm really completely 50 50. i think uh the me boost itself is a simple piece of software and it's ready and it's working but the whole builders and searchers etc ecosystem and relays behind that that might take a minute to build you know and then switch over to does it mean like after the merge we'll see maybe a drop in mev um like a total drop in all this extractable value getting we might see a drop in um

extractable value that is taken by flashbots which means that we would see bot wars again and gas wars again in that case that makes sense yeah but actually maybe to just that one little thing so mev boost uh like the post merge construction is enables like builders apart from flashbots so flashbars is not only the research collective they are uh like in proof of work like in your ethereum uh mining world today flashbots is the main builder so let's say which when we look at the diagram and sort of with mev boost they basically have developed a tool

um which allows for more data so we expect well as far as i can tell like we expect as lucrative as mev potentially is as they open up the market with this tool we see more parties come into play apart from flash stops being builders and sort of yeah so yeah it's it's it's a basic open open markets condition which which sort of um right we have to we have to keep in mind so if i if i can revise my estimate i think given that people like money i'm gonna bet on mav boosters ready there's going to be one relay and maybe one builder flashbots

and it might take a few months for this ecosystem to build out fully but there will be something to use with hopefully there will be yes all right um so as we see on this diagram right now uh the mev boost has to talk with the consensus client when you're running your own validator setup so i guess whose client is going to be ready or who has an implementation that we can start playing with so you can play with lodestar and lighthouse right now there's a third and i forget which one it was uh

i think tekku is ready as well so those three already uh nimbus and prism not yet as to after day 6 13. that might look completely different two weeks from now and when i say ready i'm talking about ready to be tested so for for lighthouse and lodestar for example these are so called prs pull requests on the github you have to literally compile the client from source to be even able to play with this [Music] yeah so i guess it takes a lot of efforts and a lot of take on knowledge it takes a little bit of effort and technical knowledge to test it it's not terrible

but yeah um and don't expect everything to be super smooth they're still all working on it and then again like uh if if they are ready like currently let's say on on the test nets like ken or robson like there's as far as i could say there is no mev to extract like so if you run the software and they are ready like you you basically need economic activity on test nets like there is not a lot for example yeah and you'll be able to create like fake economic activities just to test it if you have enough tx yeah sure yeah like yeah so maybe that could be like an interesting

like testing project for someone not there um for sure sure we want to like create a bunch of transactions and have some kind of uh search algorithm to try to extract mev and just test that the whole stack works on robson for instance yeah if uh if i i have to break a lance by the way for the for the shouting stone near stockholm uh we were told during the merge calls france troopsten and it's the it's the shouting stone where people were shouting for the ferry yeah that's what it is the rube stone ropeston maybe we'll find that fairy but exactly yeah so um absolutely if we have a rupestan whale listening who wants to start

creating lots of transactions that can be armed for test purposes please do awesome in the meantime the user will just realize like uh that the locally um created block will just be as the payload will be as beneficial as a block from mev so from mav boost so it's like this is it's test net environment and there's not a lot to see either um before we go to to demo um i wanted to briefly at least touch on interacts with um pools without getting too too deep into the weeds i don't know honestly it's like this gets too deep in the weeds don't do it we're gonna go there just for a second

so for uh for places that are not permissioned like rocket pool um basically because this gets paid to the suggested fee recipient the violator the node operator is trusted to actually send the right suggested fee recipient with this right there are some incentives there like slashing people's stake if they don't behave that the pools can use but largely this is a pool by pool thing to solve as to when are we running it and are we offering some of that gain to our stakers and then two how do we keep our node

operators honest so they don't just pocket the mev for themselves and so do you know how how some of these pools are trying to solve this issue or like do they have any solutions right now yeah so um rocket pool has a whole paper on it which i'm not going to go into in detail but basically they're going to use their odo to take people's rpl collateral if they misbehave we can link the paper if we like other more centralized pools where the node operators are permissioned and kyc'd they basically tell the node operators um you shall do this or else because if you don't then you get out it

and you lose your entire business and who wants that so they're using their reputation right exactly they're using their reputation to make sure this goes right awesome so so for users or for people who might own a particular token ask your staking pools like how do you handle absolutely and how are you going to handle mev now i might get any of that yeah am i getting a share of it like so these are really interesting questions from if you if you are a part of that particular ecosystem or all such tokens absolutely 100 right so if i have a token now um it's just issuance and i get a portion of that and the pool takes a cut right

whatever the pool percentage is um afterwards uh priority fees get paid mvv gets paid who gets that do stakers get you know goes that bad does that go back it do um the pool performance do stakers get that minus a pool cut or does it all go to the pool right meaning to the pool operator does it go to the people who run the nodes and and the stakers don't see it you know it's a good question to ask i expect that pools because these questions get asked will ultimately say no of course it goes back into the performance of the pool you know because they wouldn't be able to compete otherwise with other pools i think that's a bit then again we have we have we have mev boost like we have created a public market for

those mev opportunities where sort of we can uh hold pools accountable for like have you hidden mev from the user because in a certain block there was a certain opportunity for meb and it's it's kind of transparent it's at least more transparent in inefficiency and it used to be so um i think that's important again to absolutely yeah 100 so i guess it's time for the demo um uh yorick has a nice tech demo on how you can configure and use this on robson right now with uh lighthouse and harragon and iftar 100 yes so of course you know um i'm gonna use ethe docker

what else would i use um what we are looking at here let me see whether i can hide this video um what i'm looking at here is just the mev boost sidecar on rupeson connected to a lighthouse and we can you know it's not terribly exciting you can see here successfully got more valuable payload header that means um yes we received a payload from the relay that was more valuable and we used it how is this set up so i set this as manual it is enabled in docker right now in its merge get ready branch it is not uh exposed through the sd config

ui because it's still so experimental so if you want to play with it you will have to make some changes manually what you do is you go and edit the env file you see here the command nano is a text editor and then down here you see mav boost equals true by default this will be set to false set that to true that enables mav boost but now you also need a client that can actually do this up here this needs to be a client that currently supports mev boost as we're recording it that's lighthouse teku and um load star and then you need this

mev-boost.yaml in here as well you need to add that manually right and then now we're getting geeky we're going to go all the way down let's say i have lighthouse uh there it is by default lighthouse will build binary this will read dockerfile.binary this will need to become dockerfile.source and then this thing up here tells us what are we actually building currently this isn't even in a branch this is in a pull request so i'm saying my target is pull request 3134 for load star here i see similar pull request 3969 right once that's done

go save that file and now just do an fd update with this peep targets that's really important because otherwise sdo update will overwrite these uh changes you just made uh it will not keep these source built targets it will then source build this stuff um i can take an hour and go walk the dog or something and after that you can just start um ethy and take a look at the logs for your um like for your meb boost here and you expect to see if it runs something like this right that it registered validators and that um

eventually if you get a proposal that it got a more valuable payload header and that's pretty much it is there anything to see um from the mav boost um client like in terms of logs or anything like that this is the mav boost client this is the enemy all right this is mav boost itself you're looking at the mav boost sidecar that's it that's all you're gonna see and i guess you have to configure a relay too right yes so if we go back into dot env um the relay is pre-configured for you builder relay routes under flashbots.net if we go back to the flashbot site here you see that they have two test nate relays builders up right now one for

kiln one for rubstone and so ideally you would like comma separate several relays several relays correct but we don't have several relays so you know here we are like from from a user experience standpoint you would sort of make want to make sure that you are at least connected you you know there's some logging on i'm connected the cycle is connected to a relay right correct and you will see that in the relay itself um so let me just do the head of the mav boost log so when you look at the mav boost log you will see here

listening on that's all mev boost up to here nothing has happened yet register validator okay that means it worked if you don't see this if you do not see register validator it's not working right something's not configured right you don't have the right build there's a bug somewhere a parameter is missing whatever it might be if you see register validator it worked that's the bar data client talking to the consensus layer client talking to mbb boost and registering the validator with the relay to say so why do i need to do this because the relay needs to give me a block when it's my time to propose so the relay needs to know who i am what is my pup key otherwise i can't give me a block it won't know that

it's my time to propose interesting is there anything else um to show about this um it can get incredibly geeky i can just briefly show you what we're actually doing in in the in with the consensus client and the validity client um one second um in the validator client i need this private tx proposals so this basically says i am okay with proposals that are when i'm blind where i don't see what's in them right if i don't have this it won't work for the consensus client

i'm doing this dynamically i have this bit here builder and an http mev boost this is what connects the lighthouse to mev boost i think that's it so i think i should mention i haven't documented this and uh right now i'm reticent to document it until it's no longer in a pr um i want to document it but only when it's in a released client right now basically this video is the only thing that tells you how to do it and and that's a good point and we'll try to document how to do this with like different guides and different setup um solution because like swimmer's guide exactly and not everyone catches you and

so on if docker even if it's a great solution um but we'll try to help you guys set this up on your own system uh you'll just have to be patient with us like you you just saw there's a lot of rough edges and manual steps that you might need to take if you want to prepare for this for now i expect that as we get into girly this will be a lot more ready right um it will be a lot easier to set up um the main guides like like somewhere in coin cashier will be ready and we'll have this documented um eth docker and things like um steerium will have it in there um normal config flow it'll be a lot easier but for now very manual

so i guess that's pretty much it is there anything else we we missed on all this mev talk here probably it's a huge topic right so you the the it's you can go arbitrarily deep into into a lot of aspects here but um um yeah we still not only fairly early in the developments of of uh mvb boost but but also in discovering like what is mbv and where's mub going i guess that's uh that's where we are i think my closing remark is uh thank you for giving us a full hour of

your time if you're still watching this video if you're still with us and if you made it through that's uh that's just great on you you'll be more prepared but thank you guys uh thank you for joining us here today on this nice med topic and this small tag demo at the end here we hope you enjoy um we'll try to follow up on this we'll we'll do more of these videos and we'll try to help you guys uh with your own uh sticking setup so hopefully you'll learn a little and we'll see you next time thanks bye