HackMDWhat's New in Eth2 - 12 August 2022
- The image file may be corrupted
- The server hosting the image is unavailable
- The image path is incorrect
- The image format is not supported
Edition 99 at eth2.news
:rocket: The LFG Edition :rocket:
:warning: There is no ETH2 coin. Anyone offering you ETH2 coins is scamming you. :warning:
Apparently I need to say this :man-shrugging:
Top picks
Finematics does The Merge. This is superb.
The Merge
The TTD has landed!
The number 58750000000000000000000 joins the list of the most important integers of our time. The moment Ethereum's proof of work chain accumulates that much difficulty (read, hashes done), the entire network will switch over to proof of stake[1].
That number was decided on yesterday's consensus devs call following the successful Goerli testnet merge. It targets September the 15th for the Mainnet Merge although, due to the vagaries of hashpower, it could drift either way by a few days. You can follow the up-to-the-minute Merge prediction, as ever, at bordel.wtf, or on this spreadsheet by Matthew Rabinowitz,
The choice of TTD is based on an analysis by Mario Havel. Please note that it is not yet absolutely set in stone, and might still be revised next week before final client releases are out.
So, the route to the Merge looks like this.
- August 18th: ratify the TTD number on the all core devs call.
- August 22nd: client releases containing the final TTD all done.
- August 23rd: Ethereum Foundation announcement via a blog post.
- September 6th, 11:34:47 UTC: beacon chain Bellatrix upgrade at epoch 144896, slot 4636672 (this preps the beacon chain for the Merge).
- Somewhere around the 15th of September: :champagne:
If you want to see how the numbers get inserted into the execution spec and the consensus spec.
[Insert ob. ItsHappening meme here. Other memes are available[2].]
Goerli testnet Merge
Ethereum's Goerli testnet reached its TTD and merged with the Prater beacon chain testnet at around 01:50 UTC on the 11th of August. I was sound asleep at the time, but the Aussies were awake, so that was fine.
It didn't go perfectly smoothly. Various issues meant that it took forty minutes to finalise the Merge rather than the fifteen or so we'd hope for in ideal circumstances. But it got there, and in the final analysis it was a success.
We talked through the issues the consensus devs call and I won't re-hash it all here. There is also an issue tracker that links out to some more detailed analyses of individual issues. Some issues stem from having two candidate blocks that met the TTD criterion. Some Besu instances incorrectly rejected one of them which caused some of our Teku nodes to end up on a wrong fork for a while. It is possible that we will have multiple terminal candidates on Mainnet, but it's much less likely than on any of these testnets. In any case, I expect this bug will be fixed. In addition, if one of the client teams had not (again) misconfigured their nodes then we might have finalised much earlier.
A few hours post-Merge, Goerli–Prater was back to performing at close to its usual levels.
So, all in all, it was ok, and we felt confident enough to proceed to the next step. Honestly, if Mainnet Merge goes like that, I'd be entirely happy. The Merge does not need to be perfect!
The Goerli Merge livestream with Bankless and EthStaker has had over 21,000 views.
CryptoGucci wrote a summary of the Goerli Merge as it unfolded. And there was press coverage in The Block, Decrypt, The Defiant, Blockworks, CNBC…
Goerli Shadow Fork 6
It's seems like a long time ago now, but GSF6 happened on August the 4th and went very well. Here's a summary for posterity based on Pari's notes.
- No client issues, all combinations are synced and at head, and processed the transition correctly.
- Two nodes ran out of disk space which explains the drop in participation rate (from 97% to 94%).
- Around 30% of the network ran MEV-Boost through the transition. We noticed no issues other than latency related problems with one node in India - potentially a defective machine.
Prepping for the Merge
Somer Esat has updated his staking guides to include the Goerli Merge configuration. Somer's guides pretty much became the standard (alongside CoinCashew's) for beacon chain staking - I used his Teku guide myself when setting up my own rig. These ones cover the post-Merge Goerli testnet only, but I'm sure will be updated for setting up Mainnet: Teku, Lighthouse, Nimbus, Prysm
Check out EthStaker's latest validator workshop, which also covered setting up MEV-Boost on your node. It was a good session with lots helpful advice and Q&A. These guys know their onions. The next is planned for the 24th.
Infura's excellent Merge workshops continue. Recent episodes:
- August 2nd: The Execution Client and the Consensus Client, featuring my estimable colleague Matt Nelson.
- August 9th: Running Multiple Validators and Staking as a Service
Here's a merge configuration checklist.
eth-wizard will guide you through all the steps to becoming a fully functional validator. Release 0.8.5 Easy Panda is out and supports the Goerli Merged testnet.
The Great Explainers
Distributed validator technology is flavour of the week. Isaac Villalobos and Albert Garreta of Nethermind help us with Sorting out Distributed Validator Technology. And @markoinether comes with What is DVT and why do we need it?.
A good article from Phil Ngo of Lodestar on The Road Ahead for Ethereum Light Clients. There's been some genuine concern over the last few days about the state of decentralisation in Ethereum as a whole. Light clients are a key component in making all of our access more robust, and the good news is that the Eth2 protocol has been carefully designed to support a healthy light client ecosystem.
Also, Bankless: Everything You Need to Know About the Merge.
Media and stuff
Another terrific Know Your Client session from the Ethereum Cat Herders: Adrian Sutton presented Teku and The Merge. I cannot recommend this one highly enough.
Stephane Gosselin and Vasiliy Shapovalov joined the Epicenter podcast to discuss MEV-Boost – Preparing for the Merge. This turned out to be quite a deep dive into the topic.
Anthony Sassano joined the Show Me the Crypto podcast to talk all things Merge.
Busy as ever, Anthony talked to Mudit Gupta on SSV Network's latest Twitter spaces, who gave us a deep dive into the Merge.
Tim Beiko presented The Merge to Ethereum Vancouver. The sound quality is awful, and you can't read the slides. But, hey, it's Tim![3]
Research
If we are going to be doing this MEV-Boost thing then we need to account for it not being entirely trustless. In particular, in the current design, stakers must trust relays to behave correctly. That is, that relays are correctly reporting the value of their blocks, that they are offering valid blocks, and that they will not withhold blocks after the validator has signed them. (This is discussed starting here in the Epicenter podcast mentioned above. Hasu from Flashbots explores the third of these threats in Understanding liveness risks from mev-boost.)
One way to keep relays honest is to monitor them. Hence Alex Stokes is proposing a Relay Monitor Design that allows fault information to be gathered and reported as a scorecard for each relay. Stakers could then make informed decisions about which relays to access.
ethresear.ch is down as I'm writing this. But I had made a note of this breakthrough in constructing KZG multiproofs which will help improve the performance of Danksharding.
Regular Calls
All core devs
All Core Devs call #144 took place on the 4th of August.
- Agenda
- Video
- Twitter notes from Tim and write-up from Christine.
Lots of Merge planning. We spent the last 30 minutes discussing governance, and how we might start bringing the execution layer governance process (managed by EIPs and the ACD call) and the consensus layer governance process (largely managed via GitHub PRs to the spec) into better alignment.
Consensus devs
Call #93 took place on the 11th of August.
- Agenda
- Video
- My quick notes and a write up from Christine Kim.
Nothing interesting happened.
Merge Community Call #6
Lots of ad hoc Q&A about setting up clients for the Merge.
Upcoming events
- August 16th - Infura Merge workshop on DApp Developer Considerations.
- August 17th, 16:00 UTC - EthStaker Community Call #20, featuring the Swell team.
- August 24th - Next EthStaker validator Merge setup workshop. Info should appear on the EthStaker homepage.
- August 25th, 16:00 UTC - Blockworks' The Merge: Will You Survive, Thrive or Get Left Behind? with the legendary Taylor Monahan and a couple of guys. Registration required.
- September 1st - SBC MEV Workshop. Online attendance info TBC.
- September 9th 14:00 UTC - Merge Community Call #7.
- September 15th - who knows?
In other news
- Finalized no. 36 by Danny Ryan: Merge sequence engaged 🚀
- The latest from Stereum.
- Coinbase eth2 update 018. Special Merge edition. Some good background Merge info here.
And finally…
Not one of you complained about being Rick Rolled last time. Either (a) you all love Rick, (b) no-one clicks links, or most likely (c) you're all too embarrassed about being caught out. At SuperPhiz's suggestion I'm thinking about putting a little Easter Egg in edition 100 to keep everyone on their toes. Watch this space.
Learn More →
![[RSS]](https://benjaminion.xyz/newineth2/img/rss.svg){kind=small}
We also have an RSS feed.
Advertising on this newsletter.
