Q2/Q3 RPC Service Provider Treasury Proposals

RPC services are essential to the ecosystem as they provide a standardized and convenient way for developers and external applications to interact with the Moonriver, Moonbeam and Moonbase Alpha networks.

Given the nascent nature of the Moonbeam and Moonriver networks, the availability of free, public API endpoints removes a key barrier to entry for developers wishing to experiment with and deploy to the ecosystem.

While having a diverse set of RPC providers ensures a healthy degree of decentralization, RPC infrastructure has real costs associated with it. Thus, it is important to achieve a high degree of service quality while balancing the number of service providers with costs of these services to the treasury.

In April, treasury proposals for two RPC service providers were approved by the Treasury Council (Onfinality, Blast). These proposals were to cover the costs of running RPC infrastructure during Q1 of 2023.

Now that another quarter has passed, it is expected that we will hear from service providers that have historically been funded from the treasury to seek payment for Q2. NOTE: their proposals have already been published and are being voted on

However, the Treasury Council is asking any RPC service providers wishing to provide treasury funded RPC services for Q3 & Q4 to put forward a proposal outlining core services to be provided along with estimates of volumes and costs. Proposals should be submitted PRIOR to August 11th, 11:59 PM UTC .

The purpose of this is to ensure a more open and equitable process for all interested parties by creating a period of time when submissions will be considered. The idea is to repeat this consideration period on a quarterly or bi-annual basis. This in turn will help set expectations on all sides - selected service providers can be reasonably confident they will be reimbursed for costs related to providing the service. Meanwhile, the Treasury Council will have more predictability and a better view into longer term treasury spend.

Following the submission period, the Treasury Council will review all proposals and circle back with any questions or clarifications needed before finally indicating support or non-support for each proposal.

Generally speaking, RPC service providers will be evaluated based on:

  • Cost to delivering the service
  • Quality of Service
  • Decentralization / Redundancy / Location
  • Track Record providing High Quality RPC services to blockchains in a variety of ecosystems
  • Consistent track record of cost optimization while maintaining a high quality of service
  • Contributions to the Ecosystem to Date
  • Adherence to the Moonbeam Code of Conduct
  • For Incumbents, disruption costs if service were to be curtailed
5 Likes

Looking forward for the intentions :slight_smile:

Abstract - Proposal for running public infrastructure for the Moonbeam, Moonriver, and Moonbase Alpha networks during Q3 and Q4. Dwellirs public nodes have been publicly available since June.

Motivation - It requires a level of DevOps skills that many do not have. A Moonbeam node for example, requires knowledge of SSL certificate handling, websocket management, https management, firewall configuration, security measures, monitoring, reporting, upgrades etc. It’s extremely costly to run a node. Since we own and operate our infrastructure our cost structures are fair more sustainable than cloud providers where one can expect huge increases in price.

Project Overview and Team Experience - Dwellir is an infrastructure provider in the Substrate community. Our goal is to provide robust decentralized infrastructure that teams can leverage in building out Web3.

Overall Cost - Our proposal is that we continue to operate the public endpoints for Moonbeam, Moonriver and Moonbase Alpha for $3000 per month. A total cost of $9,000 per quarter. We suggest that we split the cost 50/50 between the Moonbeam and Moonriver treasuries, but are open to suggestions by the community.

Use of Treasury Funds - The funds will be used to host multiple instances of the Moonbeam, Moonriver and Moonbase Alpha nodes, keep up with 24/7 monitoring of core functions including scaling mechanism, do upgrades and optimisations when required. Currently we operate two clusters in Northern Europe and one in North Africa.

Specifications - We operate our own machines at smaller independent co-locations sites, our nodes will add robustness to the ecosystem’s mix of RPC providers. In addition, if this proposal gets approved we will:

  • Add tracing support for the three different networks
  • We will build a Moonbeam snap package that the community can use. A snap bundles an application and all its dependents into one compressed file, which makes deployments very easy for the user.

Steps to Implement - Our nodes are already active and are used by the community. If this proposal gets accepted we will enable tracing on the nodes in Q3 and in Q4 we will deliver the Snap package.

The full proposals can be found here.

3 Likes

UnitedBloc - RPC Service Provider Treasury Proposal for Q3/Q4 2023

Abstract - UnitedBloc, soon to become a DAO, is a group of community collators who have been actively involved with Moonbeam since the early testnet phase. In early 2023, we introduced our public RPC service for Moonbeam, Moonriver, and Moonbase Alpha. This initiative rapidly gained traction within the user and developer community, with the RPC now serving ~1.5bln requests per month across all three chains. By submitting this proposal to the Moonbeam treasury, we anticipate the opportunity to secure a grant that will enable us to continue delivering exceptional RPC services to the Moonbeam ecosystem.

Motivation - Our first objective is to foster Moonbeam’s growth, it’s crucial for developers to have seamless access to top-tier RPC services. Our free public RPC service serves as a stepping stone, supporting new developers as they launch and expand their projects. No signup. No accounts or accounting, just point the app to a UB endpoint and go. If at any point the project RPC requirements exceed our capacity, we’ll guide them toward commercial service providers. Our second objective is to facilitate the RPC needs of non-profit endeavors. If community focused non-profits ever need more RPC than our general public service provides, we will work with them to meet the need.

Project Overview and Team Experience - UnitedBloc members excel in Linux administration, allowing us to deploy robust servers. Each member organization takes responsibility for deploying and maintaining their individual RPC servers. This approach leverages our collective expertise for a decentralized network architecture.

Overall Cost - This proposal encompasses combined operational costs for Moonbeam, Moonriver, and Moonbase Alpha. The total requested amount per quarter is $10,800, equivalent to currently GLMR 44,535.7152.

Use of Treasury Funds - As community collators reliant on community delegations, we have launched the RPC service to give back, operating with a non-profit focus. This grant covers our infrastructure costs for the RPC service across Moonbeam, Moonriver, and Moonbase Alpha. It excludes any margin or profit for UnitedBloc.

Specifications - The UnitedBloc RPC service encompasses four regions: North and South America, Europe, and Asia-Pacific. The service has grown to nearly 1.5bln requests per month across all three networks. We employ global DNS load balancing to efficiently route users to UnitedBloc servers in their respective regions. Within each region, OpenResty enhances RPC requests with additional load balancing and optimization. When demand surges, we promptly adjust by adding more servers or specific services. Our design is thus scalable, flexible, and cost effective.

We provide our endpoints at the following addresses:

https://moonbase.unitedbloc.com
wss://moonbase.unitedbloc.com

https://moonriver.unitedbloc.com
wss://moonriver.unitedbloc.com

https://moonbeam.unitedbloc.com
wss://moonbeam.unitedbloc.com

Steps to Implement - Our public RPC service has been fully operational, serving the community since early 2023.

The full proposal can be found here.
UnitedBloc Linktree

4 Likes

Hello,

We’ve submitted our proposal: [Proposal: XX] Blast support for Moonbeam, Moonriver, and Moonbase alpha (Q3 & Q4 2023)

2 Likes

Proposal: OnFinality High Performance Public Infrastructure

Abstract - Proposal for the continuation of running of high performance, scalable, and reliable public infrastructure for the Moonbeam, Moonbase Alpha and Moonriver networks in Q3 and Q4.

Overall Cost -

We are requesting a fixed price of $8,000 per calendar month for Moonbeam & Moonbase Alpha (combined) and $5,000 per month for Moonriver. It is no surprised that our fixed price would be the highest as we are the largest RPC Service provider in the Moonbeam ecosystem and have been for years - generally serving hundreds of millions of requests every single day. Cost optimisation is a huge driving factor for us, which is why our pricing is actually one of the lowest based on cost per million requests.

Competitive Advantages -

  • OnFinality is the only provider with trace API, this is required for scanners and other data teams (read our recent case study here)
  • OnFinality is the only provider that provides API insights to consumers (read more here)
  • OnFinality is the largest RPC provider in the world in terms of networks supported
  • We are one of only two that allow private API keys
  • We have run Moonbeam and Moonriver bootnodes for the community for years
  • We have proven reliability of 99.9% which is always available to check here: https://status.onfinality.io/, meanwhile other RPC providers to not share apublic uptime status page

Full Proposal can be found here: https://docs.google.com/document/d/1xwEOaNQkOaHFsx619UfHZZfeN1xVdHJD7jCBvEWrsoA/edit?usp=sharing

2 Likes

Moonbeam councilors, teams and ecosystem participants would be doing themselves a disservice if they don’t opt to help bootstrap and support the UnitedBloc. These members are easily your most active, responsive and loyal daily ecosystem members. Their passion and support for the network is unparalleled. The differences in how services are deployed and the redundancy provided (not just technical redundancy, but jurisdictional redundancy), should be considered.

6 Likes

hey sir, can we know your pub rate limit?

i would say UnitedBloc’s commitment to providing RPC services with a non-profit mindset is commendable and demonstrates a strong dedication to the Moonbeam community.

But the little concern i have is that how do you plan sustaining the RPC service over the long term, especially as the ecosystem and usage continue to expand?

Hi @jose.crypto,

We allow 40,000 per 10 minute, which equals to 4000 requests per minute. This allow surges in requests that can happen during certain spikes.

Thank you for the kind words – we do indeed have a strong dedication to the moonbeam community.

Regarding sustaining the public service long term, there are three factors which will contribute to our success:

First, the ability of UB to remain active collators. As active collators, UB members have the resources to “give back” RPC services to the community.

Second, the token price of glmr/movr. If the token drops precipitously, then our ability to “give back” is lessened relative to hardware costs.

Third, the future performance of the moonbeam/moonriver client. Purestake is developing RPC performance improvements (SQLite) in the upcoming client v33. We are optimistic that performance improvements will keep pace with increased usage going forward.

We cannot predict any of those three factors, but we have optimism that all 3 factors will contribute to sustaining our public RPC service going forward.

Thank you for your comment and I hope this helps address your concern.

–Daniel

5 Likes

Kudos to Dwellir team for giving this proposal.
Given your experience with other projects, could you provide examples of how you’ve managed sudden spikes in demand for RPC services?

1 Like

Since we own and control the physical machines that run the nodes, we have a very high capacity to start with.

Our model is different compared to if you run a node in a public cloud, where there is almost a 1:1 mapping between cost and usage.

The hosts that have the nodes are powerful machines, they run multiple nodes and can share the same memory and storage pools. If a node requires more resources, then it will eat more of the shared resources.

We always ensure we have excess capacity as our customer base and usage is constantly growing. We do this by forecasting at least one quarter in advance. We have excellent relationships with hardware providers which is a bottleneck that is often challenging for small companies.

This is all part of our business model and roadmap. Given our knowledge of the growth of the polkadot system we are not concerned about our ability to scale to meet the demand of the Moonbeam community nor the wider polkadot community.

2 Likes

Thanks for sharing this proposal. from what i see, OnFinality have an impressive track record so far, Kudos to that.

With your extensive experience in the Polkadot/Substrate ecosystem, how do you anticipate scaling your infrastructure to support the growing needs of multiple networks, including Moonbeam?

hey guys, I greatly appreciate all the proposals and the effort you’ve put into structuring and writing them. your ongoing support for our beloved chains is commendable. however, I have some concerns regarding why our community needs to pay for so many RPC providers. I understand the importance of decentralization and having backup options in case of RPC downtime. nonetheless, I believe that paying for more than 2 providers might not be very practical, as it could potentially lead to resource wastage.

please understand that this is my humble and limited perspective. I would welcome input from both community members and RPC providers, highlighting the pros and cons. my intention is not to offend anyone, but rather to initiate a discussion to better comprehend the reasons behind having more than 1-2 RPC providers

2 Likes

Hey @turrizt ,

I understand your concern about resource usage with multiple RPC providers. To strike a balance, the community could consider a phased approach.
Initially, they could start with fewer providers and gradually scale up based on demand and usage.
This way, we maintain decentralization while avoiding unnecessary resource allocation.

3 Likes

Thanks for your question @db_defiedge!

We pride ourselves on scaling our infrastructure and have a long history with Moonbeam networks (as early as January 2021 when we started supporting Moonbase Alpha). This experience has allowed us to master the art of scaling up and down all the Moonbeam networks based on changing demand and pressure (there are graphs in our full report showing the fluctuations over time).

We have a dedicated team expanding our chain support (as it stands today, we are the largest RPC provider in Web3 in terms of supported networks) while ensuring there are no sacrifices to the quality and reliability of our service (maintaining uptime of 99.9%). With our proven track record, we are confident we can continue to scale alongside Moonbeam and Web3 as a whole while continuing to provide service excellence.

1 Like

hey sirs since you have a paid version could you tell us the following:

For paid plans whats the recovery time for a lost node and is it automatic?

and for all:
WIll the public RPC network persist without interruption if a region/origin drops?

@Marc_Blockshard @brittanyseales @Mihai_BwareLabs @gustavnipe

1 Like

Thanks for your question @jose.crypto! Can you please clarify what you mean by “if a region/origin drops”? If it drops due to a technical issue or if we decommission the region on purpose?

At Dwellir, we prioritize the continuous availability of our services and understand the importance of ensuring that our RPC network remains uninterrupted. Our infrastructure is meticulously designed to withstand regional disruptions. Here’s how we ensure that:

1. Global Distribution: We have strategically located physical machines across independent data centers, spanning multiple continents. This global distribution ensures that even if one region or data center faces an unexpected issue, the others can seamlessly take over, ensuring uninterrupted service.

2. High Redundancy: The likelihood of multiple data centers experiencing downtime simultaneously is extremely low. Our setup is designed such that even if multiple centers were to face issues, as long as one site remains operational, our services will remain accessible.

3. Proactive Monitoring: Beyond our robust infrastructure, we’ve equipped ourselves with a modern monitoring software stack. This software constantly oversees our network, scanning for potential upgrades, detecting issues, and identifying technical faults in real-time.

We are fully committed to ensuring that our RPC network remains resilient and dependable, regardless of regional disturbances.

/Gustav
Team Dwellir

1 Like