Agriculturist tends to his harvest
A hopeful perspective
The meadows are ablaze
summary
- Medalla progressing effortlessly
- Diversity in clients is essential
- eth1+eth2 (Phase 1.5 also known as The Merge) comprehensive demo
- Evaluations and tests continue as we near Phase 0 launch
Medalla appears promising (after some excitement)
A calm testnet raises suspicions.
If you’ve been monitoring Medalla at all in recent weeks, you are likely aware of the significant 5-day event that took place on Friday, August 14th. Refer to Prysm’s detailed analysis for specifics on the technical aspects and timeline, along with Ben’s latest blog posts ([1][2]) for an overview. Client teams put in long hours over the weekend following the incident, deploying synchronization and peering updates to mend the severely scattered network.
Although the incident introduced immense pressures on the testnet, it offered all clients an opportunity to strengthen their defenses against the most unpredictable scenarios. I can genuinely state that client software is far more resilient following this event. I can now rest a little easier leading up to the eth2 mainnet launch.
Since that event, Medalla has progressed quite seamlessly: presently featuring 39k active validators with an additional 12k in the activation queue (that’s equivalent to 12 days)!
Diversity of clients is vital
Despite the availability of numerous [outstanding, feasible, sturdy, user-friendly, etc] eth2 clients currently being developed, the network is predominantly governed by a single client — Prysm.
Historically, there is sound reasoning behind this — Prysm has concentrated on early testnets, community involvement, and usability for over a year. Hats off to the Prysmatic team. Fostering a community is both incredibly challenging and essential for our sector and open-source endeavors overall.
However, the episode on Medalla was substantially magnified by the failure of the leading Prysm client. As we advance towards mainnet, we, as a community, must intentionally strive to address this issue. Having explored all the eth2 clients on Medalla, I can assure you that most clients are sturdy and well-supported, and all client teams are actively present on Discord and GitHub to assist you with any challenges you may encounter.
Safeguard yourself
Client diversity not only enhances the eth2 consensus but also shields you in extreme circumstances: due to the anti-correlation incentives present in eth2, the greater your negative behavior aligns with that of others, the higher the risk you face.
For instance, consider that 60% of the network loses connectivity for several days due to a failure in client-A, while clients B and C remain stable and operational. Even though the chain will continue to be constructed by B and C, it won’t finalize because of the >33% downtime. If you’re using client-A, you risk losing an escalating amount with each epoch that the finality outage lasts (referred to as an “inactivity leak”). Conversely, if you utilize clients B or C, your balance remains secure since you stay online. [Note — an inactivity leak is significantly worse than standard offline penalties.]
Now suppose a minority client-B (comprising 20% of the network) encounters a critical failure causing a client-wide outage. In this instance, the chain can still finalize (since 80% of the network continues to participate). There are no “inactivity leaks” incurred by the offline validators, only regular penalties. Hence, those using client-B will only face minor penalties compared to the scenario mentioned earlier.
Clients facilitating easy transitions
In addition to community efforts in exploring new clients, client teams are diligently working to ensure that transitioning between clients is both straightforward and secure. With the introduction of several cross-client standards, you will soon be able to switch from one client to another with minimal downtime and no risk of unintentional slashing.
Such standards, which avoid client lock-in, are a crucial aspect of a sturdy eth2 network. The ability to easily modify software will empower the community to address issues (similar to the Medalla incident) more swiftly if/when a singular client encounters failures.
eth1+eth2 complete demo
A principal aim of eth2 is to achieve Phase 1.5 (also called The Merge), at which moment the consensus of the existing Ethereum chain will be integrated into eth2. From that point onward, the chain we cherish will be constructed by proof-of-stake validators instead of the current energy-intensive proof-of-work consensus.
The transition to Phase 1.5 is intended to be as smooth as possible for current users and applications. Eth1 clients remain the workhorses for state, transactions, and execution. By keeping the vast majority of this user layer unchanged, Ethereum can harness existing tools and APIs to power transactions and dapps, just as they do now.
To this end, Mikhail (TXRX) and Guillaume (geth) recently unveiled an end-to-end demo of a multi-sharded beacon chain (with an eth1 shard included). In the released video of the demo, Mikhail transmits several transactions to the eth1 shard chain utilizing an unaltered Metamask wallet.
You can explore and engage with a dockerized version of the eth1+eth2 demo, or if you prefer a deeper dive, you may build and run from source.
Ongoing tests and audits, focusing on Phase 0 mainnet
Operations continue as usual on this front.
Client teams are putting in tremendous effort, auditors are examining every detail, and plans are being laid for mainnet launch 🚀