Hashing It Out
Hashing It Out

Episode 98 · 11 months ago

Hashing It Out #98- ETH 2.0 Panel After The Launch

ABOUT THIS EPISODE

Jaye and Corey lead a panel discussion about the Aftermath of The Beacon Chain Launch for ETH 2.0. Panel members are Ben Edington (ConsenSys) and Gregory Markou(ChainSafe)

Links: ETH 2.0 Panel

Jaye Harrill

Ben Edington

Gregory Markou

Sponsor Links

The Hashing It Out Social Media

Hey, what's up so avalanche, let's talkabout it! LIT'S AN AVALANCHE! The snow comes down real, fast, fierce gamesmomentum, but I'm not talking about the natural disaster or I it's not reallydisaster iguess of no one's aroung but anyway avalanche. What is it you've heardabout it now, Yo getting here some more? It's an open source platform,volaunching DYCENTRALIZED FINANCE APPLICATIONS RIGHT DEFA! That's what your want developers who build on avalanche caneasily create powerful, reliable, secure applications and customblockchain networks with complex rule sets or build an existing private orpublic subnent right. I think what you should do right now is stop what you'redoing, even if it's listening to this podcast stop poull over go to the gasation. If you need to go to a subway, there's a subway like everywhere,there's always a subway all right right, there's always a croger! Just stopin aparkand lot somewhere go to avalav, OALAB STOT HORK TO LEARN MORE! All Right! STOP GO TOALVA LAB! That's AVA labs, labs, Dhat, ord, now, enteme cn work welcome to hashing it out a podcast forWetalk to the tech, innovators behind blocked. In intrastructure anddecentralize networks, we dive into the weeds to get at. Why and how peoplebuild this technology. The problems they face along the way come, listenand learn from the best in the business. You can join their racks. Everybody welcome back to hashing anouts for, for the first time on video. This is going to be the first videocontent that we produce, which is going to be an FTO panel of storts. We did one ofthese before main Neth launched on a theryum two point: Oh for the for the ayear. Now I think was it a year. Was it death longer Yeah Holy Cal than feellike that? Long at all, we we still underlockdown weren't Wi. Now this is before wow, okay. Well, we did one of these along time ago, apparently, and on that one we drank wine and there is a winetasting appropriate with. So today we're doing some of us are doing thisSaimg, depending on the time zone. I'm drinking nineteen crimes, cally redthat has a picture snoop dog on it, and I will let others let let you knowwhat they're potentially drinking or not drinking as Mansidue HIMSELVF jy.Why don't you to yourself hi, I'm Jay, I'm drinking coffee fromthe local nw, it's a little early for me. I don'thave a problem drinking in the morning, I'm just like any coffee. My name is Jay. I've been in the spacefor a long time as you m. As many of you know, I know what I'm not doing this II workwith Quan stap and we do security there yeah and did A. I think what we had done is wetalked about Hese two few months back as part of a meetup and it's good tolike kind of keep talking about to eat to, and I do roughly monthly little gettogether so keep an eye on my twitter, and you can check that out. Excited Otin here o put that the description, O miksure try to try and t fit as manylinks as we can at least Ju's with us to take notes and make sure that we putthe things in the description that we say we're going to great sa Wat's up, hey, hey yeah know I'm drinking water,it's like one, and I got a lot of work to do today,...

...also yeah just the last time. I speltone all over my keyboard, so I don't want to go through another keyboard, as you were freaking out about yourkeyboard. As someone was trying to like walk us through a wine tasting Hehad and Jy, was I told Jay, don'tsend me questions and gave me like the hardest question about Sfi, which I hadto Google, what SFI was yeah excited and for Lo Odona, CTO andlone of the cofounders a chain safe, and we run and maintain the loadstarclient, which is running a full sink right now onmain net. We just started advising a way to run about ARY client. As oftoday, which is you know, for the sake of e December tenth awesome bid, I am Ben Edgington got myself a Californiann Mal Beck, it's six o'clockin the evening here, I'm in the UK, so definitely ine o'clock and I have been with Consensos for justover three years working on Iftheyre into formost of that time before it wasI ere into, and it was just shouting and scalability first in a research capacity, and I'mnow product owner or product manager, for our tecku Eto client, which ar sortof target market or niche that we're carving out is mostly amongst theinstitutional stakers. But it's a perfectly fine cliency throwin at home.I've got one running right now. Earning me rewards, I hope,Wi better check. That's part of the discussion that I'mgoing to bring up is a few a few of those things you justmentioned, and I guess I'm Corit Petty. I work for status an so status. We runthe Nimbus, client and meand J part of the Hashkat out podcastand where we talk about generally anything to centralizationblockchain, whatever really want to talk about.It's interesting to us. We try and dig in as much as we can and we're not veryshy about being technical in any way. So welcome to the show. It's basicallyUS talking and I'm going to kick off with thery of two mainnet watched, so thebeacon chain, as a deciver fir, was officially so like what happened wasenough money in the theorof one deposit contract was deposited which reachedpastor threshold, which launched the genestis block of the FTO beacon chain,which is the first phase of a therium two point out, which is the effort ofAtherum to migrate to proof of Sake and increase or change the architecture sothat the number of transactions can rise drastically and the burden on anygiven validator is minimized. So there's a lot of other than now yeatthreaten in out there's a lot of implications of this type of stuff. Wecan talk about some of it or not, but what I've kind of wanted to talk aboutwas like now that started. What do we expect? What has been ourexperience so far as people who develop Han run validators? What are questionsthat people should be aware of or be askingand what are the different client teams? ECAUEIT's not fully representative?Here we an there are more client teams that are running very good software, like what are we focusing on as client teams as we Continuu to develop and get readyfor further phases of a therity point out. So I'll kick off. I'm running I'mreading four valutors myself at a single note, which is something that Idon't think a lot of people understand. A single node can run multiplevaliators and there are consequences of that. Have youall experienced that sofar when running support for the people who arerunning our software yeah. It's interesting this running,multiple Valod, validatas pernode thing.

I think the very original concept was.You know one CP, one vote, which was you know, going right back to Satoshi'svision, that kind of got lost in proof of work, and so you have one validator pe forUnit, but the deavs we were to cunning and we found ways to run multiplevalidators pernod. So adding validators doesn't really have much work load toyour once. You got to beak a node running, you can add. We are happyrunning two sand plus validatas per beker node and it doesn't ink increasethe load a great deal. So in that sense I feel we kind of Miseda step on thevision of one CP one vote, but it's it seems really hard to achieve that S.it's it's a decent compromise physician, because Runnyo baconode is pretty lightweight, so they can. You know we can distribute and get the decentralizationwe want. In any case, that probably doesn't ask you questionCory: likeas toay, just r CN note the scaleisn't an issue of like running multiple validators doesn't increase the load onyour machine very much, but does have consequences right like say I need tomake it upgrade, and I would like to I need to restartmy note or for some reason another. I have a power doutage, an eed to Resenin MI TMI mydo goes offline. The amount of work that I miss gwall off line isDirectlyr riportional to how many valigators are running on that singlemachine. I was going to say all of that sounds like a lot of slashingconditions. Just inesing Asic right, that's going to be it's going to be he the amount of money you miss byproviding quality attestations. You know for a certain amount of time, ifyou're offline you'll lose that money as well, but they're not nearly assevere as a slashing condition, which is what happens when you do somethingwrong, which is good t s, something I'm going to get into later right, that'ssomething that I really want to get into later, because we've seen italready happen within the ten days so far of people running clients and butmost of the people who are running clients right now should know whatthey're doing gret do havein like giveany input onthis in terms of like yeah Teri, so port yeah so from aclient team perspective we're in a bit of a unique kind ofposition. So, for those a know, we're like a full type, scrip and momentationleveraging some assembly ship stuff to you know, get wasam out there andbecause we're not, we don't have people actually like running validators. Nowwe run Ouroun and I'll get into that n a second, but because we don't run ourown doughters. Our support side is actually from our utility libraries solike the deposit Youi that I'm assuming almost everybody used to deploy like,for instance, was using our browser, bls library and, I believe, also as aSE library. Last time I checked so the ceralization library in the keygenerating libraries, so you can generate so like our focus was thereworking with, like you know, tef folks and Carl, to ensure that, likeeverything was offering correctly so we, our support, was like before it alllatch right. We were like helping make sure we're conforming and stuff when itcame to actually riing the validators. You know like a we run right now, we'rerunning a light host tat up we're going to migrate to load star once were likeconfident and ready to get that going to show public. You know we're goodwith that, and some of the things that we've noticed that have been likeinteresting is especially with the upgrade stuff. So, for instance, ourUPGRAD biggest issues we running is actually aure eath one note falling outa shing and the ability for, like you know, we run GEAT, simply dod. That wasour choice and we have another mindbackup, but you know like it falls out a sinkfor you know what we end up happening. Is Our machine hit? Actually, we peekedon Ram. I EAN we ran out of Ram, so we had to do a hot swap and get thatupgraded, which had some downtime and...

...the biggest thing we found actually waslike the guest note, took too long to sink and we had to fall back to like aninferior like service to actually get us opthing going again. The DeaconNotes sinked pretty fast. So when it comes to, actually you know needing toperform that upgrade weve found that to be really really easy. We run a doctorsetup and basically that allows like it's quite literally Doctor Down Dr Poland you pull the specific images and then doctor back up those specificimage names, and we run that in one one command and we're talking like seconds.So I found the downtime with actually running validators. Actually quite nice,it's not that bad. The biggest issue we had. We missed some attestations duringthat period. becane. You know our learning services actually weren't setup correctly, so we found out we weren't covering a certain edgecase,because there weren't logs, that we found out after the fact that you knowthere's no sorry, I might have got a phone conl ofmy cut ut, but there is no long actually being omitted, saying likeyour valator missing that testation, because that's not something, that'sactually logable and not something. We can scrape off grep out of the logsinto data dog, which is our monitoring service. So that's something I'mworking, nowt ith, the lightus guys is like. How can we better, you know,figure that out and get like more verbos logging to know like hey thisdidn't work. You know, like your actess stations aren't going through. I M anthat's kind of the biggest challenge right now that I'm going to probablyassume people are having because from what I've found it. A lot of people arejust you know on you, know, beacon, chain or eat, beacon, scan or whatever,like just looking. You know. One hundred you know I'm like lookingthrough the list. Did I miss anything in the last like hour, because that's a bit, I think that'sthe biggest challenge right now right and it's like how do we handle that? Ithink thats O and enough there. There aren't other better waysto to notify that, like I remember when I was first frunning any one minding note like I would try and pars out any like specificevent logs just so I can kind of leave it and then come back to it and seewhat's going on, but I think most people, maybe they might know how to do that, butthere's a good chance that they don't yeah. We've got this in Turkoo, so itwill in those when it should make an atti station and that atestation shouldshow up on the on the chain on Jane within thirty two slots, which is sixminutes and if it doesn't, then we have metrix to to drack this and we canalert on them through all these or of Promedias and and Graphano alerting. Soit's is possible to do is delayed, of course, because you know you have towait to see that appears on chain and we can tract the distance thatartestations are included so it' ar included very quickly or they delayedby a few slots. We can also track that mmetric and that's a good indicator ofthe helth of unit node if it's Shorton memory or CPU that that will increase. Let's take sa that actually thinkgoheabefore get to this conversation, to keep this, maybe as selfcontained aspossible, let's try and give a short overfe like it's. So as a user, I would deposit thirty two tinto the T: One Deacon Chain, deposit contract. After that gets accepted, I thin havebasically keys associated with a Valad dater on the F to beacon chap acompletely new blockchain, but same asset talk, maybe Wel Git, I that later,but it's a different blockchain. It's fur on death, one that means like okay, you have avalidator on F to awesome. What is their responsibility of asingle valitator? What are they doing and what are they trying to make surethey're doing correctly, who'd like to answer that correct have ot it yeah, I mean like a volder, is kind ofdumb to some degree right, because you know all at's really doing is making aget request to the Beakonot, and it's just saying telling asking the BEAC anode like am I su? You know for a set of information about the current stateto know. Am I supposed to sign...

...something right, whether that be an actestation orproposing a block? And those are the two main rolls rightit's a testing andproposing a block proposing block is a higher reward, then a testing. I then,as we discussed earlier already like a testing missing, acttestations, isn'tthe end of the world. You know the slashing conditions come from, you knownot actually being able to prosthese blocks, doing double spans and suchlike that or not double spen, but rather like a double proposal orwhatever, so the vaderies themselves are aren't goving a lot of the ligework, at least in my opinion. My point of view, a lot of the likework reallyrelies on easier beacon, note able to to produce enough data so that you knowevery so how many seconds you're polling for that beaconnode the vaderspulling that beaconote to know. Is it ready to do something? That's my like really distulbed versionof it. So you have. This have te spell to this.If this beacon dote right Te, can no software this running, that's gatheringinformation for multiple sources, it's thinking with the other beacon notesofthware the network to make sure that they're understanding what theinformation that's going on just within the beacon shape and it's pullinginformation from Etheriam one to make sure that it's understanding whatvalvator sait is currently alive, like what? What is the active validater said so that it could e could performappropriate randomis and doing that is a lot of networktraffic. Eventually itwill become more network traffic, as we subdivide thenetworks and the Sharts, and they have to then manage switch hot.U To different stuff that works very very quickly and manage to appeersassociated with all those different things later conversation. That's forus for another Hassye out later on down the line, so I need to understand allof the network traffic across the different networks and make sure that when the network wants me to dosomething, my beacon chain said: Hey validator, sign. This message proposethis block with this information et Cetra. It does that in Sifitsit to network in a hopefully reasonable mount of time.Tbat then the rest of the network can can attest to those things or make surethat it does h't correctly right. That's why what then I'm saying as thenumber of valiators I, the number of Vali dators on any given node can scalereally quickly, because it's not doing a lot of work. It's just signing shitwhenever wants to whatever it needs to the real work is in the no softwarekeeping up with all the different netfork information consuming itputting in the right place, keeping track of like all the things that have been done bythe validators. You don't Redo it and alearning. When something is wrong, is that DAMAG? You agree with tefetsummary Yeah, so I mean I tell you the work ofthe validators announcing to the network. This is my diea of the world,so we've got all these thousands of beacon chain nodes distributed acrossthe world and they all have a slightly different view of the network. They'veall seen different network traffic over period of time and they're all slightlyoter, think with each other, and my validator periodically is called uponto say this is my view of the network and to sign off on it and is puttingdown the sate that give you the right to y sign off or on that, and thenthat's brok us to the network, and you receive all the views from the rest ofthe network and then the beakon note can adjust its view of the world tobring it in line with everyone else's view of the world, and we need amajority of Validatas that are the Egree, because if somebody wants tospoof a view of the world, so you know cory wants to promote his own view ofwhat the network looks like. Then he needs to have a a majority, or at least a furtherthan network, to achieve hat. It's getting crazy. Your own Thori, I'm gonna press a little further onboth of you. Can you further distil all of what you said into a tweet length, and Iwill allow for up to...

...three tweets in that tweet length as if it was athread go greg? I knew you're going to do that. Yeah I'd say like there we go. VALIDATOR has two rought two jobssimply attesting to existing data, enchain data and proposing newblocks. Within these two scopes of work. Theyreceive all their information from T E beacon chain. They don't communicate toany external body, except for the beacon chain that you, the beacon,chaine notes, sory, that you have told it to communicate with. At thatpoint, it's up to the beacon crain node to disseminate all the informationgather, the network, the other communicate with all the other, notpeers on the network. To give that information. So ifout it a can poll thebeacon chain node and know what to sign, Wendto sign and submit that back to thebeacon chain, note to then be propagated throughout the network. That's really good yeah. What gritsaid validators propose a view of thenetwork and either agree or disagree with each other. That was a lot better. That is certainly weet. Okay. So, let's assume for a minute thatwhoever's listening, maybe less familiar with with es to then then maybe they should or maybe ethentheir experience or what have you? There seems to thee there's. Obviously,this difference: betwen node, Validator, beconoede Becan crain contract. What is the relationship between thosefour right? So let me see about this, sobecan train contract runs on if they're in one, and that is the the register ofall of the stakes that have been placed so far in thirty two eth increments. So that'sthe SORC of truth for WHO's eligible to be a validator. The beacon node contains all of thestate of the beacon chain and is the point of communication with the rest of thenetwork, so beacon those talk to each other and they watched the depositcontract and in order to watch the deposit contract you need and if hou'rein one mode, running, geth or open a theorym or nether mind or whatever, andso really all that's doing in the current context of Ithoin to isproviding a view of the beacon contract. The deposit contract to etherium to sobeacon modes are watching the deposit contract they're talking amongst eachother they've, maintaining the state and then hanging off each beacon mode.You've got a number of validatas which are periodically attesting to the state ofthe local beaconode and telling the network. This is my view of what'shappening on the network. Each valid data has a private key associated withit. So when you deposit in the contract, you create a private key which iswhat's registered in the contract and that's also registered with yourvalidator, so that that ties the whole thing up and closes the circle. So those areyour fulk components essentially perfect. I think that provides atlittle, more clarity for people, so anyone can just be like...

...they can just if they want to run a beacon note. Theyalso need to run an eth one know for to watch that and then they can go furtherand run validators as well. Ll Need e EastOne note from somewhere depending upon how much you trust other people, youcan consume it from someone else's eath. One note and that's that's a verydifferent conversation, but yes, you have to have that you have to have it. It's like aTheorin, one data as an oracle. Andto your beacon note where you get thatinformation n. How much you trust it is up to you, and maybe that's maybe it'sreasonable t talk about the consequences of not having proper data there right so like. If youhad abeth one data compromise in somewhare, Shir perform what couldhappen to your validator yeah, yeah, good point and one of the biggestchallenges we've had getting up and running as been people's eet Wonedou.It's just still not informator a lot of resources.Yeah Hoes exactly exactly its still, not light weight and, and so especially around Genesis,people weren't in sink years and so on, and and also initially the amount of data. If youstart up in any theory in tlu client, it has to show ov a lot of data in thein the early moment from you know, all the deposit contract data up to now outof the on note Ond. This proved to be a bit too much for some of the those wewere using. So we had to dial back on the rate we sucked information out ofthem just so they could stay responsive, but I think we're over that now. The only thing you can't do, if youcan't, if you can't see the deposit contract vy, your ethon notice, proposea block. So when you propose a block you have to is mandatory in theprotocol include any pending deposits up to, I think sixteen deposits, and ifyou miss out any, if you don't, because you don't know they're there or you'retoo lazy, then you will. Your block won't be valid. So that's basically the onlyconsequencis that s whenever you propose Al Block it wouln't be valued.You can still learn rewards through auter stations. Hat's got norelationship with the eath Oneersul for purposes of right. Now we want totalk about Slashi lashing issues later, but if you get ashed, that's a harsherharsher punishment to your validator, we'll talk about how you could slashlater, but is it possible to be slashed byproducing an envelop block because you have wrong or like maline data from Eth, one ONA go dreg. I that's a good one. I'm going to seeyes, but I almost want to go look up the steck andget back to him like two minutes, because this is Hor. I thought of thatquestion. ' Will just mot yeah yeah. I know. To be honest, the reason I'mgoing to say yes is because it's considered invalid and that means yourblock would be voted against on right from an fattestation perspective. SoI'm gonna say yes, but you know what ot me brb in that in that time period I thinkit's reasonable to start talking about slashable events in network and I'm a proprepropote gipen. Theproposing an invalid block is not a slashable offense. So all you do, isyou lose your block reward now? Your block reward is is a decent reward. Imean it's worth as much as I'm like eight ATI stations, but it's it's not not a huge chunk. Give you if you missit, so it's probably worth a bit more actually, but it's not a massive lossif you miss it. So if you have invalid...

...data, your block is considered invalid by the network andyou lose your blockheworld, but there's no punishment for for that, andcertainly not slashing. If you propose to conflicting blocks at the sameheight, then you can be slashe. That's what I wanted to get into you. It's been ten days, there's alreadybeen quite a few ashes more than I expected, and I think, based on the pre, likeBesina information that I saw the last time I checked. All of them are almost all of them aredue to people running the same note multiple times right, yeah, so there've been fiveevents unless there's been another one today I e four of them are individuals who had single nosand in in each casethey were running to validators with the same set of keys or the say, thesame private key on two Vali, Datas and and of you know, confessed to doingthis, and basically, as the backup they wanted to have high up time. If one oftheir validators went down, they wanted to have the other on. You know a hotstand by so that it would fell over, but actually the the result is that thethese eventually these validators are contradeed bcause, see each other andthat's a slashable offense, and it's not so bad. Currently, you get finedimmediately point to Fivee. You have them Exra, penaltiees, notvery large. You probably lose about point three ou point four of eneath,but the big consequences Youare then kicked out as a validator so and youreat is the remaining state. Thirty one point sixtee or whatever is then lockedup until we mergee uner neath two, which could be a year of year and ahalf two years away. So it's then unproductive. So that's the mainconsequence. You don't lose a huge amount by getting slash right now wereduce the penalties for a few months, but you are then locked into the systemwith unproductively full of foreobal future. I didn't realize that gettingkicked out of the network currently kicks you out for two years. I thoughtlikthat the that it was just for a period of time, but that's Wen were live. THAT'S NOTRIGHT! So Y need one one soi right S, yeah Netd is that it's just Pasero Lika Chadoing much other than what it's supposed to do, which is which is thedependent function upon the other phases, and so we have to make surethat this works appropriately and works well and provides incentive for peopleto do it, but that the east that lives on eath to is a one wiy function ascurrently, and it's going to be a while until we finishe the other phases ofthese. Two before you could you can have any utility whatsoever with theeat associated with it, and so right, like you say, yeah, goahead. Sorry, yeah just finish off on the slashing that there were ten nodesflashed which belonged to a single staking service who had done a homebrewand te slashing solution and one of their notes a got out to think and theymanaged to miss the slashing protection theydidn't, detect the condition ND and ended up with ten nodes being slash.They pretty quickly turned off everything else and fixed it beforerestarting yeah and to just after, like Ino,thespeck Yeah Been Ben Zaccurate in the you know, as long as you're proposingwhat you think is accurate right, like you're, going to be fine from a sashingcondition. Lashing conditions are the intent around you're trying to bemalitious. I think that's probably the better way to like really describe it.It's like you're not going to get because something just didn't work right. Youknow from like you collecting data you're going to get because you triedto do something very. You tried to be malicious to the actual PROTOCO TiausIti overwhelming approvable, that you've done something. That's negative,towardor thatwork Ahbutit's interesting...

...though, because Bris of Curza currentlystands hi. This is mo main. I don't want to call it in competence, but likebad devo, PRACDC practices over optimization lack of understanding ofconditions. Things like that, like no ones actively trying to Ticke over thenetwork, but you kno not going to get slashed unless the network can provethat you've done something like bad like at that can lead to a Mo likea movicious act as opposed to to yea complete information yeah- and this isa point the network can't tell if you're being evil or incompetent. Ithas no idea, so we do have a concept of correlatedactivity. If you see a lot of people simultaneously breaking the rules, thenthe penalties are much higher, because that looks like a coordinative attackwhich is going to be much more dangerous, but odd. One off events arevery lightly peenalized. For that reason, yeah a I'd like to put thisinto perspective for some of the listeners I', actually ask this Tou Cory with so for those who are like thinkingof of running a validator, you know understanding why you shouldn't be malicious. If wecan put that into sort of monetary terms. For those you know unaware right, like you need to take thirty twoeee, whichis you know, one eath is sitting at today. As of today's value s like what six iave no in I don, I never checkFori, I fiveity UST, so you know to put that in perspective.You're not gang like if you get drop from the network, the thirty two es isgoing to sit them at deposit contract for the next two years, roughly whoor longer, depending on howlong phasl takes, and you will not be receiving any apyon that. So the the thirty two eath would not be working for you at all and it would be simply loft, so itisn't so II guess I guess maybe it'll be good to talk a little bitabout. It sounds like some of these. These people were trying to dooptimizations or midication, but they weren't, but they were, they were thinking it. You know selfish termsrather than in network terms, and I think it would be interest. Itwould be good forpano dine into like how should someone be thinking aboutthis? If the think in about optimizations it sounds like theyshould be. Thinking like more simpler, you know just make sure thet stulf andnot trying to overcomplicate it. My Yeah so sorry coy co on what is currently the ultimate way torun. A validator and beacon note is to havededicated hardware that has a few times the computitional resourcisneeded to run a beacon note, which is relatively like right life. You, likethink about a Nembus hemisnode, which is like the lighter of resources of thenetwork like running on a pie. Foor you can, I don't recommend it. I recommendsomething: That's Foor Cores Akicke Bite, O memory andbefore theywork itvicpeves an equality quality mata hard drise face. I thinksomeone said and our note program, Er running it like I, the amount of consumption of harddrive space is not trivial. So like running on, like araspberry Pi, four is not good right, because those are usually run off offlash memory, which is is at good...

...memory. It's not rebus. T E main thingis you want to have something: that's going to be online for a long period oftime and stable, and it's dedicated. It always has a good Internet connectionand, like you said earlier, there's a good source for eath one which, ifyou're, going to run yourself because a significant amount of more signivantabount, more e computational resources to do and what you're really trying tooptimize ust. You run the beacon, goof software. You make sure that you have agood practice on how to update it and that's going to be dependent upon theclient software that you have and it's usually going to be. Stop the node like update the software in thebackground on it and a separate folder, not the same. fother move the binary towhere the youyour scripts keep it. And if you don't understand these words,that we need better software practicis to make it easier for more people to do,which is us kind of stands to like push the idea of like who should be runningthese things is the people who understand what I'm saying right now.Eventually, when we get better at this, we can broaden that. So you stopped you upgrade yoursoftware base on whatever updates you need to do that are critical, stot thenote and restarted you' never, and you need very solid devops practices interms of how you stop your note, how you upgraded how you started and that'sjust like that's just running it cool. We have hed a craded software, that'srunning a piece of software that we know how to upgrade and it stays onlineall the time now what you need to optimize for- and this is what I thinkthe current most of the curtent work- is being done in a client teams.Obviousl it this for us an embus. Is How do you monitor this thing? How doyou know when you need to do software crads, but those software of SoftwareCersan tail? How do you know when you miss that asestation? How do you knowwhen your number of peers isn't sufficient, like all of thesegot kindof metrics around what your note is doing and that you're doing the rightthing in the right amount of time? Whath your inclusion distance over aperiod of time right so like how fast did your actestfision get get acceptedby the network? All of these things, you need to have ar very quick way ofunderstanding, and if things aren't meeting acceptable criteria, you needto have a way to be alerted, so you can adress it appropriately. It kind of wesaid this earlier was like it's not immediately obvious like whenI you consider that festation, but it didn't get accepted or something'swrong with it. There's there'st kind of like subtle ways in which at thestations aren't getting ar being included optimally, and so, like themonitoring part of all, this needs to be what's being optbized and that's that'snot only a client side thing, but it's also a networks. Type thing in terms oflike, like you said earlier, I think it was you correct, a like most of thepeople who are checking on whether or not they're doing this approprately isthey're checking beakon chain, which is beacon, chain Fille out but dot. I Minstead of the full word which I want had, that is ether scans version ofnetwork history yeah. So that's you know everyone has. Everyone hasdifferent viewpoints of the network if anyone's been around long enough tobe doing trying to get into into sales back in two thousand, and sixteenyou know would know that you want to spen different nones, because it'sdifferent views of the network, so hin that same in that, in that same vein,like, I think I think what t what people are getting at is is there is not one single source oftruth? It is a collection of sources of truthand pivs right. It depends on where you'relooking right, because if you look at the, if you look at the aggragatedsource, the BLO chain itself, that's Ustin, he source of truth. TAKESI takesa lot of individual sources of truth. To get to that point, and that's where,like finality comes in that's the whole...

...like beauty of Rotan isess, is you havean emergent, an emergent finality from a lot of individual perspective? Right I mean finality okay, there's there's finalitysure, but you still have. You still have to consider like it as a clock right if I'm running a,if I'm running a let's say, I'm just running in thes,one node work cong compering, you know, like my node, compared to what will agree with what is being said oneither scam but there'se still going to be slike timing variances there is, itdoesn't matter much for this definition, probably not but obstensively. I stillwould believe that there 's still a slight view of slight difference inview of the network. Yeah Wat Couri said about finality is key here. Sofinality is a point of which a whole network agrees that it has a consistent single point of truth, vewof the network. No, nobody deviates from that. If you're in one provor work doesn'thave finality, you've got farm assurance that nobody will ever comealong with he chain, that's different from the one that you see, but it'snever. You can never say never in Ethere Im to and IMPROVA stake ingeneral. You can say before this point in time we will never change theHisregen. Everybody has the same view of the history of the chain and it willnever change and we achieve finality an neath two in about twelve minutes,which is not very ambitious. I mean you can things are effectively final, muchquicker than that under most circumstances, but absolute finalitiesis achieved in about twelve minutes and anything before that you can guaranteeit's. You have the source of truth, so a definition as to what finality is,and an implementation of that is actually new for you to, because wedefined I finality before, based on what Bitcoin was doing and now there's a different idea ofwhat finality is, because we've learned that it's not necessarily fun final,that's that's intthis is, is a longer subtle conversation on the differenceesof what Consistos agorithms ar running, whether it be traditional consensus orclassal consensus, macamodil consensus or variations of traditional consensusthat include Crypto econotics and whatu like s nocamotic consensus, whichis stanardly preferred to as like proof of work is probablistic has propoliticfinality, which means that over time you have a larger a d larger. Iprobability that a something won't be changed and that's why you need to wait.Six blocks in deckoin or whatever it is worth is to like to say like okay, thisthing is good, but if you're buying a cup of coffee with Dick Coiin, youdon't really care if it, its lickit e likelihood that it's overturned. Isn'tthat big of a deal taditional consensus is when thenetwork or t the people participating in consensus validators in the point ofat two Combe Decision? It's done, but with Casper FFG. It's a mixture of thetwo kind of you have hut the concept of an epoch which is somwhat probabalisticfinality until Li th EPOC is over and then it's snapped over. It's final forthe rest of the further ofthe time, and so, like that's, that's a, I think,a larger conversation in terms of like what it means to be final and theeconomic differences of probablitic finality throughout an epoch and andthe security around like as you move through an EPOC and thenwhen it's a done, which is a a very settleand compigatedconversation, interesting thread for sure. But yeah. Maybe we can kind of deviatea little from that onfinality, but it sounds like overall, the network is going well...

...apart from a few people got flashed and unfortunatelynowthey there's like what eleven people in ten days or eleven validators in tendays that have each at least thirty, two eet locked up for two plus yearsbefore they're allowed to potentially rejoin the network. Maybe that kind of leads into what thenetwork is designed for. Given you know, two thousand and twentyhas been kind of rough on people. The the design of the network is is meant to outlast cataclysmic events. What doesthat mean? It's Correat, that's you always with the tough ones. Er Yeah know I mean like. As we talkedabout, I mean for Li catastrophic standpoint. You know we could lose.Someone could like you know we could lose, and this gets into a largerconversation without, like the actual web, is technically wired up from likea sea cable perspective, but, like let's say, like the connection fromlike North America to Europe, it's cut rigand. We lose that ability tocommunicate with those people, and now we have to rud our thoughrselvesthrough, like Asia, to get to Europe right. That's how the North Americanconnection to Europe. It sends up happening like we discuss it's like noteven the penalties that are going to be that bad right and it's like we it. TheProtocol Self, is designed to be resilient to the idea of like massoutages to some degree. Obviously, unless our threshold drops were like,we literally don't have enough out ers to kind of push the chain Ford, which Ithink we saw in one of the testmets. I just can't remember which one ore abunch of the valdues just stop validating. So I think it's been designed quitewell for, like those type of like DDA scenario, like dayzero kind ofscenarios and whatnot yeah, I mean I don't know it's j Stlike to see that type of like to see like acatastrophic situation. wher we lose that much of the network would be,would be interesting, l, you know, and I think it's been designed well and it yeah it's referrd to as the world warlthree scenario in in some of the documents, and the idea is that all all the validatorsvotes are waited by he r their stake, and so, if validators go offline and so my viewsitting here on the UK, if I, if I can't see half the world's validateors,I suddenly disappear from you, I'm on a much smaller network and I have half ofValidatas and missing. In those circumstances I can't achieve finality.There are not enough validatas alive out of the whole set that I know shouldbe there to agree on the state of the chain. So I can progress. The chain I mean,have half the validators: Are there well make on overage blocks every everyother slought? So we can still make progress, but we can't achieve thisfinality, which is desirable, and so in tnose circumstances we have what wecall the quadratic leak whereby increasingly the validators whodon't show up their balances. In my view of the world are decreased anddecresing decrease and eventually they get down to around sixteen either fromthirty two and they kicked out of the of my local network, so that takes normally two to three weeks. Wi'vherelax that so itwuld be about six weeks plus now, just in case of any troubleand the idea is that it when enough have being kicked out, then I now wenow have enough local validate because the networks, small enough is, is justthe ones t at that. I can reach that. I have my own network. I've basically cut off therest of the world and have a separate network, and we can we can progress andstart finalizing again and after that, the different networks. You know,presumably there's one somewhere in China or whatever that's you know,separated from mine after three weeks and they'll. Never it's essentially ahard fork. Then they'll never be...

...reconciled. So there are, you know,three plus weeks to fix any problems, but after that time we prioritizeregaining finality and making he the network safe again. Can we talk aboutthat? Actually its prebably good point because it's been prougt up a few times.Let's talk about the difference between like a voluntary exid and gettingkicked off the network and like what that really means, because it's beenbrought up like Oh, your thirty twoth is like you know, you got to wait awhile you know so like. Let's, maybe then, ifyou want to touch on that, it's like the idea of like topping up and alsolike how much time you really have right before you forcefully get kickedoff versus like wanting to just like, withdraw yeah the I'm. The three ways break thatone is to be slashed in which CASS involuntary one is to neglect yourvalid ATA. So it's not running, and eventually you will get down to sixteeneve and if the rest of the networks functioning that's going to take threefour years and you'll eventually drop below thethreshold and then there's voluntary exit wherebyI say I no longer wish to be participating in the network and youcan exat. You join a que, but you usually would expect the C to be short.So four validators connect it every six minutes and then you've got a shortperiod of time. I can't remember exactly how much a couple of days andthen in principle you can withdraw your eath after that. However, we have nomechanism to do that until we've done this. This merger lev one a need to do, and but at that point you Coul exiiteand you could rejoin. We should talk about this. This wouldemerge and what comes next an we sort of had this two years before anyone canwith withdrawor anything. I hope it's going to be shorter thanthat, the so the initial Roadmat we started a couple years ago, the veryLinea Roadmap Wer you do plase zero. We do pase, one wich you Shardin whic ispace to which is a sort of abstract execution, engine sort of concept, andthen we implement eath one on top of one of these abstract execution engines,and it was sort of pure and clean and long and had this end goal of mergingof on o need to. But the pressure is on honestly to do the merge earlier andget to eat. One off proof of work and onto proof at stake and also o take advantage of the scalability sogradually that Merg as kind of moved forward in the roadmap Tso, it wasmoved to Thas one point: Five, which is sort of after sharding, but beforeexecution engines and now we're looking at a proposal called executoble beaconchain where we actually don't. Even we don't even have to wait for sharding tobring Iteran worninto with Irum to we put eth one strati into the beaconchain. We can do that before we even have shards we could. One of my colleagues is building a demoof this right now and it's not technically that complex ther a fewloose ends to tie up, so it may be much quicker to get eath one into eath to and off proofof work, and at that point people's validator balances when theyexit you'll be able to bring them back into thee, one that we'll know and loveand and can can freethem Muse them yeah. Thatbrings up a really good point too right, because for the last two three yearseverybody's been seeing e two development. It's ust the BEACA chainright, it's like we have all these other things and if you put, if you tryto put on a timescale, one's going to Gan chart and be like well, you tookthis long and howit's going to take this long. It's going to take this long,we're looking at nine years, raise like what you know if you were to put it infront of like a product, you know Maner. That was no idea. They're like this iscrazy, but I think the real key thing eis for also remember, like you, knowthe beacon chain and what we've done- and I think Danny said this really wellat CS KONAC. You know our ven we held...

...last week, which was like listen likewe spent this long because we have to get the consensus algorithm from eathone and recreate a new one, and basically where the beacon chaine Solpurpose is to prove that this consensus algorithm works and we can iteratereally rapidly at this point- and you know it's like get something intaproduction, and now we can start shipping quick. I think that's likesomething that you know classical web to development. Wilso like appreciateslike once you get that first base layer, it's like now. We can start iteratingon features and stuff right and like, for instance, in our case, one of ourpriorities is liht clients, because we want to be able to essentially say,like your metamass, no longer needs to leverage like an inferior like service,and now you can just run a light client, which is you know. Why went brouserfirst and our teams already we've been prototyping that you know like thesteps needed from the beacon chain and we're already like full steam ahead,getting going forward on like time to actually start working on that, becauseeverybody's already got like phase one, some phase, one components ready, youknow and they're. You know working on and that's where you know we're goingto start seing liht clients come out and phase one which is like here's awhole new useability layer right and then you know it's not just shards.It's like we start seeing a lot of the like ecosystem support things cominginto play and that's worn going to see a lot of like more rapid Vilm, and Icould see phase on happening really like parts of phase on happening likewell within the next six months, in which case we're really far ahead tolike see that eth one merger gets so much closer. I think that's a superundervalued appreciation of like why the beacon chain is really importantbecause, like we got the hard stuff done, you know it's like now. We canjust put features. It's like a transfer yeah. We can do transfers now, likemight not want it because of some edgecases, but at least like we havethe fundamentals to get it done. I did not realize that the first of all that each one mergerhad moved up h in the timeline. I was more aware of you know. Maybeit's going to be on a ghost chain with like a like aplasma chard or plasma a contract, or you know, maybe would be its own shardor whatever later, but this this early merger. That'sreally! That's actually really interesting, we're kind of coming up a little bit totime so I'll kind of give it back to coriefor. I do I just wanted to saythat I'm looking at beacon scan right now and there'scurrently nine hundred and twelve thousand plus e eligible ether to voteon attestations, which means that is, I think, that's the same about as howmuch is sticked so we're just under a million eter stake, which I think you know given thePrir city to right now is. I think I think we would. That is almost if notjust at a billion dollars USD worth of locked value. On the for this, like this, this phase zero.So that's so this is, you know for thes. You know o the fact that we'reiterating you, you know it sounds like everything's going to be able to iter.It really quick, because the difficult part was happening. That has alreadyhappened. That's a lot of confidence that that thate two is not only we, we knowwhat's happening, but it's it's happening and that's really exciting,especially for those of us who've been watching SINC for a long time to see to see that is, is really really cool,so yeah. Thank you, yeah and there's a two week,Q to get in there's another three hundred K, ef cued up to join and will joinover. So if you stake today, you andt actually become live as a validator forover two weeks now, because of all the stakes that are ahead of youin the QE...

...and yeah. It's also I mean when thedeposit contract was announced, things were very slow for the first couple ofweeks and journalistms were contacting me saing. What are YOU GOINGTO? Do youGointo know the thresholds you're goingto delay whatever Iwas like no, noit'll be fine and we find e Gota. I gotto be tens to the weekend beforehand,but then, suddenly it just went vertical the deposits in the contractand it's is been the votof confidence has been awesome to see. Yeah yeah verydelighted, something Super offtop kind of offtopic. I just noticeis like if you actually look at the year epy and youlike spy Validator, you actually noticed that the person who's made themost right now has produced only five blocks. The most has been ten, andactually they actually have two sashing conditions that they've triggered,which is really interesting from an economic standpoint. Right theye pickedup, flashings other validators getting slashed and you get decent chunky reward for includingslashings reports in it in a block, so yeah yeah that'll, be where a good partof being comes from. Okay, I was looking at is a tha Godslashed, so we're running out of time and like to de respectful of Ja and ourtime, I'd like to try and rap up with a little bit here, Crat Ge, anythingquick to say, to wrap up the episode o anything. You want to tell anone tothink anyone shout outs to whoever you like, okay yeah, my biggest one ishones. I've been showing hus for like two years now, and you brought up it'slike make sure you're planning a hardware. Accordingly, you know paseone and face to are going to get scary. You know deat con. Last year we talkedabout you know tens of gigs of data. You know needing to be stored, O youknow a given week or two, and it's definitely something to consider- andyou know if you do want to run, that Raspberry Pie highly suggest,considering only the validator on and connecting it directly to where yourbeacan noticed it up, and you know, like you said, you know, we got ta planaccordingly and not only O, you have to know, but your software up grade past,but you have to also know about your hardbor ou great pass. You know how youit's something to start practicing. Probably in the new year, is hey. If Ineed to expand a thirty two ram for my Acizram, how am I going to do it and Wam I going to do it? So I don't get ash with a double. You know double valitorsituation, so definitely something to consider for a lot of people is knowinghow to do software up grades and also knowing how to do hard. Word up gradesbecause that's something you want to do probably in twenty four hours, you knowdon't want to leak too hard and check out load ster for some Ossi,an death tooling and we're being used everywhere, and you know if people arestaring to use it. We also have a what thejs has our EK to support coming out?Probably this week, our next week and so you'll be able to you know what sellup through jess and cory some stuff from the beacan note. If you want tohave that jabscript experience that you know awsome tin be closing top yeah. I mean I like to reflect on thejourney at times like this. It's been incredible. A lot of people, Sai Hi,couldn't be done. We've done this in perhaps the hardest way possible. We'vekept it permissional learth we kept it open. We have kept it decentralized.I'm talking about the protical development. All comers are welcome. Wetake ideas from everywhere. We don't have a dictator Ar life. You know Danny Ryn is close, but Y, and I think it has served as well isbeing hard and bumpy, and I've been a lot of R we've done everything in theopen right. A lot of protocols are retreated to a closed room, a lab andhave done their thing out of the public player and and haven't really produceda lot. We've done this all in the open and everyone' seen all the missteps andeverything, but we've delivered this speaking chainthing and I am deeplyproud. It's just the beginning. I like to call it proof a proof of stake atthis point, but with a lot long way to go, but I'm more confident than everthat we will deliver all of these two in a timely fashion. Ason. Thank youJAK. I mean it's been a phere, just you know,being there kind of on the sidelines...

...and then, of course, you know being with with consept beingable to work with a lot of great loggrat groups and, of course Tecu aswell and yeah. I mean it is crazy that we're here huge congratulations to the communitylike just so much work, huge, congratulations to all the teams and yeah the I mean this is this. Is this is phase of zero right so still long way to go, and I guess, asalways, Shille be filled right, othat's, a Tshirt, but iverive everheard one okay! So thank you all for joining us. Iappreciate you kind of helping me share kind of the wisdom that we'velearned in the process of trying to deliver this thing. What what we'vegained for running it building it. It's very clear that we're all veryproud of the work that has been done by the like Plessara of people that Ave thathave contributed to it, but there's still quite a bit of work to do and aquite a bit of education to give for those who'd like to participate, sowe'll keep trying to do this for those who are listening or watching for thefirst time on hashing it out. I hope you enjoyd the conversation theproduction will get better for the conversation will always be good and sexe guys.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (109)