Update on Guardian Nodes and announcement of the Guardian Node onboarding program
With the Theta mainnet launch fast approaching, we’d like to share with the community some important updates on Guardian Nodes, a critical piece of Theta’s Multi-Level BFT consensus mechanism. This includes the introduction of the Guardian Node onboarding program, which will let users start earning TFUEL and will help select the initial operators of Guardian Nodes. Here’s a quick TL:DR for those who just want the takeaways:
- Guardian Nodes finalize blocks in Theta’s Multi-BFT consensus protocol, and are meant to be run by members of the Theta community. Users that run Guardian Nodes will be rewarded with TFUEL.
- The plan is for a minimum requirement of 10,000 Theta Tokens staked to run a Guardian Node (details on why below)
- As a transition to Guardian Node full functionality, we are launching the Guardian Node onboarding program, including a downloadable client (to be released by end of March) that will allow you to earn TFUEL based on the amount of Theta Tokens you hold, your “node” uptime, and your system specs. For the Guardian Node onboarding client, there is no minimum number of Theta Tokens to hold, and all users will be able to earn TFUEL.
- Evaluation of the top participants in the Guardian Node onboarding program will influence which users are first to participate in operating Guardian Nodes upon their release
Background — What are Guardian Nodes?
In Theta’s Multi-Level BFT consensus, 10–20 Validator Nodes make up a smaller committee to produce and validate new blocks to be added to the Theta blockchain. Guardian Nodes are a second node type that make up a “second line of defense” of up to 1,000s of community-run nodes that download and examine the Theta blockchain and try to reach consensus on finalized checkpoints. The name “Multi-Level BFT consensus mechanism” reflects the fact that the Validator/Guardian division provides multiple levels of security guarantee. With thousands of nodes, it is substantially more difficult for attackers to compromise blockchain integrity, and thus the blockchain has a much higher level of security. We believe this mechanism achieves a good balance among transaction throughput, consistency, and level of decentralization, the three corners of the so-called “impossible triangle”.
(If you’re interested in learning more, check out Theta CTO Jieyi Long’s paper on Multi-Level BFT; it’s getting some academic recognition and will be featured at an upcoming cryptography conference! We’ll share more on that when the news goes public.)
In our first implementation of guardian nodes, there is a theoretical limit of ~1,000 Guardian Nodes that can participate in block finalization before communications latency starts becoming a problem. However, even to reach that expected number will significant require testing and optimization. To that end, our plan is to begin with a Guardian Node pool of 100 nodes, and as we confirm that performance is consistent and satisfactory, that limit will be consistently increased to further diversify the Guardian Node pool.
Guardian Nodes will be run by the Theta community to perform block finalization, and later participate directly in block production by acting as Validator Nodes on a rotating basis. For running a Guardian Node and staking their Theta Tokens, user will be rewarded with a proportional share of all new Theta Fuel (TFUEL) generated. The Guardian Node functionality is currently in development and is expected to launch to the mainnet in Q1 2020.
User requirements to run a Guardian Node
Our expectation is that the minimum number of staked Theta Tokens to run a Guardian Node will be 10,000 Theta Tokens. The main reason for the minimum requirement is security — if at first the network can only support a limited number of Guardian Nodes, we don’t want it to be possible for a malicious actor to attack the network by running a majority of those nodes while holding only a small amount of Theta. By requiring a minimum number of tokens to be staked, the cost of an attack becomes much higher. While we’d like to allow everyone holding Theta Tokens to participate in the consensus process, security of the protocol itself is the most important consideration for the long-term value of the Theta Network.
Expected hardware requirements to run a Guardian Node are:
- Internet speed: 5Mbps+ up and down
- CPU: 4 cores or more
- Memory: 8 GBytes or more
- Disk size: 1TB or more, SSD hard drive preferred
We expect the initial limited set of 100 nodes will be far fewer than the number of Theta tokenholders interested in running a node. It would be neither healthy for the Theta protocol nor fair to let the largest tokenholders run all the limited nodes, so we propose a way for all participants to prove out their ability to contribute to the network:
The Guardian Node onboarding program
The Guardian Node onboarding program begins with a downloadable client that verifies the number of Theta Tokens you hold, your hardware specs, and your uptime/consistency running the client. By scoring well in all those categories, you will earn TFUEL rewards in your Theta wallet just as you will when the actual Guardian Node functionality launches. Your TFUEL earnings will be proportional to the Theta Tokens you hold in your wallet — the more Theta you hold, the more TFUEL you’ll earn. This TFUEL will come from the TFUEL reserves that Theta Labs holds, and until Guardian Nodes formally launch the supply of TFUEL will remain fixed at 5 billion. The Guardian Node onboarding client will be released no later than end of March 2019.
For the GN onboarding client, there is no minimum requirement of Theta Tokens to hold. You can hold as few or as many tokens as you want, but your TFUEL earnings will be proportional to how many Theta Tokens you are holding. While we expect the minimum to actually run a Guardian Node to be 10,000 tokens, we will be using this onboarding program as a way to gauge the landscape of Theta tokenholders, and that minimum token requirement may change based on testing and observation.
Your uptime running the onboarding client (how consistently your computer is online) is critical; the consensus protocol requires nodes maintain consistent contact with the network. If nodes are significantly asynchronous, block production could be delayed, or the protocol could become vulnerable to an attack.
While the onboarding client is light and does NOT require specs as high as the actual Guardian Node to run, the client will take into account your hardware specs, and users that meet the actual requirements of running a Guardian Node will be noted if there is excess demand to run Guardian Nodes at launch.
More broadly, one of the benefits of running the GN onboarding program is that we can evaluate which users are performing to the standards required to run a Guardian Node, which will allow those users to take priority in the first Guardian Node set upon launch. Ideally this can be done at the protocol level prior to GN launch and make the process as unbiased as possible.
Stay tuned for the release of the Guardian Node onboarding client later this month so you can start earning TFUEL! As always, please join us in the Theta Telegram channel with any feedback or additional questions.
Cheers,
The Theta Team
