Shoutouts 🙌

The top performing providers are: silent, STAVR, hereWeGo
‍These providers excelled in their quality of service, which drove high traffic to their nodes.

Lava & Evmos’s community MVPs were: Appieasahbie, LiveRaven, Stake Village
These providers were highly engaged in the initiative, helping fellow RPC providers with onboarding and technical issues, helping the core teams with providing support and improving the initiative overall.

We call all providers to engage with Lava & Evmos on Discord & Twitter, help others onboard and solve issues, write onboarding guides, help improve documentation, post valuable content and let us know. The teams will evaluate all contributions to decide who are November’s community MVPs.

October in numbers 🧮

Usage

• Total relays — 62M
• Total CU’s — 828M

Lava’s public endpoints have been added as the default for users on Evmos’s co-pilot, Chainlist, and developers viewing the Evmos docs. These placements helped bring traffic to the endpoint, which is now significantly higher than the usage of Evmos RPC on Lava pre-ipRPC.

Why we measure Compute Units (CU’s)?
Different RPC calls require various levels of computation depending on their complexity. Looking at the number of CU’s rather than number of relays allows for a deeper understanding of the work providers are doing to serve users on Lava, and the compensation they deserve.

Providers

• Number of providers — 71
• Geo location — The vast majority of providers are in EU.

‍

Payments

• Highest payment — 12,727 $Evmos
• Average payment — 567 $Evmos

The payments account for the volume of requests each provider served (in CUs), and the quality of their responses (availability + latency + sync). This means providers are incentivized to excel. We used the following formula:

Updates 📢

Rules & requirements

1. Providers participating in ipRPC are required to stake a defined amount of Lava testnet tokens to ensure fair competition between providers. That’s because higher stakes = higher chances to pair with users. Providers who had overstaked during October will not be panelized. However, in the following months they will be eligible for lower rewards in correlation to their overstaking rate. i.e. 3x overstaking → 1/3 rewards eligibility.
2. Providers are no longer required to support both Evmos Mainnet & Testnet. Having less providers on testnet will lower the costs for most participants, and allow the ones running testnet to increase their income eligibility.
3. Providers are required to run their own nodes. Providers who are routing relays to a 3rd party providers are not eligible for payment.

‍

Protocol optimizations

As the network experiences more usage, our team and community are able to analyze the data to optimize the protocol’s parameters to ensure both high standard of service for users, and even participation and opportunity for providers.

1. Lava’s pairing function is in charge of connecting RPC consumers with providers. By default, it returns 5 providers each time, meaning conumers get service from 5 providers every new epoch.
   After examining the data from October we decided to increase this parameter to 12, to have higher provider participation, and more even payment distribution in the following months.
2. We’ve also increased the Provider Optimizer’s “exploration” and “pertubrbation” rates by 25%. This means the optimizer will chose more providers which will increase diversity, with a potential risk of lower consistency — we’ll revisit this point in the November report.
3. The provider optimizer “remembers” bad providers for 24 hours. We’ve now changed this parameter to 1 hour to allow providers to return after being jailed or after returning flaky (inconsistent) responses. The optimizer also remembers good providers, so they will be selected more frequently.

Conclusion 🌋

October was an exciting month for ipRPC, with major launches and remarkable growth! We look forward to reviewing next month's data including the first month of Axelar ipRPC. Until then, be sure to stay updated in our Discord 🌋