[Proposal: XX] Nodefleet - customized Infrastructure

Governance Treasury Proposals
, ,
1

  • Title - [Proposal: XX] [Status: Idea] Moonbeam/Moonriver/Moonbase Alpha customized Infrastructure proposal

  • Author - Nodefleet

  • Abstract

Ongoing costs for the running of customized bare metal public infrastructure for the Moonbeam, Moonbase Alpha and Moonriver from 1st June 2023 until 31st August 2023 (3 months).

  • Motivation

As a node provider company that already runs several blockchains and already supports Moonbeam/Moonriver networks, we understand the advantage of having various providers. Especially a dedicated hosting/baremetal with higher throughput (convenient for spike usages) and the opportunity to bring a cost effective infrastructure solution to Moonbeam so it can keep offering RPC access to 3 of their main networks to his community/developer.

This is done by providing a quality and battle tested provider improving decentralization and increasing availability while delivering a solid platform for its builder community.

  • Project Overview and Team Experience

Nodefleet is a Web3 blockchain and node running company focused on delivering value for investors/builders on the Pocket Network ecosystem providing top quality engineering and quality infrastructure around all of its products. Currently our main product is POKT nodes staking, but also staking in other blockchains and datanodes services.

Working at Pocket Network since the initial years, we also started, since 2019, running nodes. Then, in May 2022, Nodefleet was born, and ever since, the company has been among the top providers in the space.

Nodefleet is the third largest node provider on pocket.network with all of our Pocket Network domains combined, our main domain which is nodefleet.org is processing around 117-150m requests per day served through a decentralized blockchain network with around 23 different blockchains nodes across 3 regions; Southeast Asia, North America and Central Europe. The last 30 days we’ve had a total of 3.6 billion request processed (attached in the links section)

From these chains we also run Moonbeam/Moonriver and process around 200,000-500,000 requests daily.

We are highly interested in collaborating with Moonbeam Foundation by delivering quality service/products in order to add more value to the network and community

Links

https://twitter.com/nodefleet

https://nodefleet.org/

(Pocket real time statistics relay process per chain)
https://www.poktscan.com/node_runner?domain=nodefleet.org&tab=chains

Alternate domains we manage:
https://www.poktscan.com/node_runner?domain=blackholestake.io&tab=chains
https://www.poktscan.com/node_runner?tab=summary&domain=whiteholestake.io
https://www.poktscan.com/node_runner?tab=chains&domain=thundernetwork.xyz

https://www.poktscan.com/node_runner?domain=lightspeedstake.io&tab=chains

Nodefleet total chain report last 30 days

Public Endpoints

RPC
https://moonriver-mainnet.nodefleet.net/rpc/
https://moonbeam-mainnet.nodefleet.net/rpc/

Websocket
wss://moonbeam-mainnet.nodefleet.net/ws/
wss://moonbeam-mainnet.nodefleet.net/ws/

  • Team -

Lowell Abbott - Senior DevOps/ Infrastructure Specialist
Steve Martinez - Devops / Backend Engineer

  • Rationale -

The Moonbeam infrastructure grant adds value to the Moonbeam network in the following ways:

  • Providing multi region access via http/websocket to 3 of their main networks (Moonbase Alpha/Moonriver/Moonbeam)

  • Provides a quality based and cost effective solution globally supported in 3 different regions

  • Provides a dedicated bare metal infrastructure with higher throughput and limit API limit on API calls for the Moonbeam community

  • Provides up to date maintenance with the nodes

  • Provides dashboard with traffic usage shared for the Moonbase Foundation team

  • Provide open source deployment tool of our software stack

    • Overall cost -

$6,000 USD monthly for the support of the service on a trimestral basis on the first day of every month. The amount can be payable in GLMR calculated at an avg USD price over the last 7 days.

  • Use of Treasury Funds -

Infrastructure

The costs shown below are assuming there’s 2 nodes for every region
(US-east/Asia-Singapore/EU-Frankfurt)

  • Moonbeam Mainnet 6 nodes - $1,000 USD
  • Moonriver Mainnet 6 nodes - $1,000 USD
  • Moonbase Alpha Mainnet 6 nodes- $1,000 USD

Support and maintenance

  • $3,000 USD

Total

  • $6,000 USD monthly

  • Specifications

We’ll be offering private HTTPS and WSS (Websocket) multi-region access to bare metal nodes with Moonbeam, Moonriver and Moonbase Alpha networks through nodefleet.org infrastructure, in order to bring availability, infrastructure costs savings and a real time dashboard with traffic/usage for analytics purpose for Moonbeam foundation.

We’ll be delivering the following list of features and specs for this project:

  • Service Availability 99.9%.
  • Multiprovider deployment - the nodes used for this project will exist in 2 different providers to avoid downtime in case of provider failure.
  • 24/7 Engineer/Customer support.
  • Blockchain friendly providers - The providers we use are blockchain friendly in order to avoid any sudden service cancellation or sudden change in provider policy regarding running blockchain nodes.
  • Real time monitoring and alerting via a shared dashboard with the internal team for visualizing traffic overview/incidents/downtime.
  • Developer support and alarms monitoring on dedicated Discord channels.
  • Support for 3 regions and 2 nodes each on every region for every blockchain (Moonbeam/Moonriver/MoonbaseAlpha):
    • North America (US-east)
    • South East Asia (Singapore)
    • EU-Central (Frankfurt)
  • Geo-Balanced traffic and geo-balanced redirection - meaning that we’ll be redirecting the request to the closest region and if one region is down we’ll be redirected to the closest region available.
  • Unlimited calls/second (rate limited just for 2k Request per second to avoid DDoS among other security rules).
  • Baseline of 300 Millions API calls monthly.
    • Everything over 300 millions, it’s charged $1.00 USD per 1m.

Deliverables

  • Geo-Load Balancer RPC dedicated endpoint with http and websocket for Moonbeam/Moonriver/Moonbase Alpha.

  • Shared load balancer traffic/usage dashboard with Moonbase team for monitoring/analytics purpose.

  • Discord channels

    • Monitoring Discord channels with proper status/high peak usage alerts notifications.
    • Maintenance/status update channel regarding maintenance windows on the infrastructure.

  • Steps to implement

Below is a set of next steps that we’ll take after the grant it’s approved by the foundation.

We have most of the steps done like the public geo-balanced endpoint but since we are going to provide dedicated bare metal nodes we’ll be doing some infrastructure setup for it:

Milestone 1

  • Deliver geo-balanced dedicated endpoints for moonbeam foundation with support for http/wss for the 3 node mainnet networks:
    • Moonriver
    • Moonbeam
    • Moonbeam Alpha
  • Perform load testing and deliver endpoints to the Moonbase Foundation

Milestone 2

  • Create and configure the grafana dashboards shared with the Moonbase team with proper uptime charts/ traffic usage metrics

  • Create and configure alarms with discord channels shared with Moonbase team to keep infrastructure updates communication and automated alarms from the infrastructure

The steps mentioned would take around 2-3 weeks or less to complete since we need to clone existing nodes to provide availability, execute performance tests and customize the dashboards mentioned above.

2

Hey, @commander_nodefleet, thank you for your proposal.

I wanted to bring to your attention the need for you to submit a Treasury proposal as it aligns perfectly with the specified needs and requested amount. to obtain detailed information on how to submit the Treasury proposal, kindly refer to the provided link: Using the Moonbeam Community Forum | Moonbeam Network – feel free to make adjustments to this post based on the information provided there

1 Like
3

:-1: No thanks, i think it is something not necessary atm.

4

Hey, Lordglmr, when providing feedback on a proposal, it would be helpful if you could increase the clarity of your thoughts, ask counter-questions and give arguments as to why, in your opinion, this is not necessary at the moment. this approach will contribute to a more constructive discussion and will allow the proposer to get a clearer idea of what is worth paying attention to

1 Like
5

Hey! @turrizt Thanks for the quick response!

Based on the guidelines, since i need to send this over the Moonbeam Community Forum I’ll be deleting this proposal until all the others steps are being completed

Thanks !

6

hey again, I think you don’t need to delete this proposal as I have already changed the category, I suppose you can just make some adjustments to this proposal based on the guidelines I sent you. therefore, feel free to take your time and make the necessary adjustments. if you really want to delete the proposal and make a new one, then answer it in the affirmative and I will be able to close this one

1 Like
7

Got it! @turrizt I just didn’t realized you modified the category

I just did the adjustments requested:

Updates:
→ Added proper notation on title of the proposal
→ Reordered/renamed sections according guidelines
→ Deleted unnecessary section on the proposal

Let me know if this is what’s required, and again, thanks for your time!

1 Like
8

awesome, LGTM! now, you’ll need to gather feedback within the next five days from both the community and the treasury council. it is expected that during this time they will express their thoughts and feedback. good luck to you👍

1 Like
9

sure, i think this service is brining nothing new therefore investing 6 k x month is unnecessary.

10

Hi @Lordglmr our service has different features that will open more use cases needed for the development of more robust products, for example:

  • No limit API calls per seconds

    • Useful for dapps/platform for doing benchmarking, loadtesting and others more complex kind of usages
    • Useful to avoid interruption of service in case of spikes of usages

  • Baseline limit of 300m API calls

    • Helps dapps/systems in no worrying in incurring charges during product growth phase

  • Smart cost usage

    • Very useful now in current market conditions

  • Using additional providers is always a good choice

    • If the Arbitrum foundation had more diverse providers, maybe their downtime would have been less than expected. It is always good to have various providers, specially the ones with less limits in their API and more access to raw compute power (bare metal machines). Those tiny issues can totally destroy a reputation/coin price in case of important launches/high spike usages.

We are also open to offer archival nodes with same setups/limits for the networks included in this proposal, which will add more value to the proposal. The only downside is, this can increase the monthly price.

Hope the reasons and options I listed above would help to answer your questions.

Thanks!

11

hey how are you, thank you very much for your interest, my personal opinion on this topic is that we already have enough rpc providers can see the list: Moonbeam API Providers and Endpoints | Moonbeam Docs

and as for treasury funding, for now it is done through a post-payment, according to the stats and as for me criteria I like to see if it has a large number of stats accordingly, that can benefit the entire ecosystem - public good

So for now it would be a NAY, in this proposal, but some new modalities are being considered on how to handle this type of proposal, so I would recommend waiting for an update on that topic. :raised_hands: Hope you can understand

1 Like
12

btw, matey, just to clarify and ensure accuracy, does it appear to be a typo?

what specific case of downtime are you referring to?

13

I’m taking as an example the situation that happened in Arbitrum when they released their airdrop. They had scalability issues that could be addressed if more providers would have been available at that moment.

1 Like
14

Hey!
Thanks for your response!

Understood! In any case we can also offer archival nodes which we also run on our infra along with the other mainnet nodes which is something I spotted that isn’t available as a public good and would be good to have since is useful for indexing/analytics and other heavy use cases.

I’ll be waiting on the new updates and new modalities guidelines related to this topic.
Thanks for your time!

15

This topic was automatically closed after 5 days. New replies are no longer allowed.