Transcoder Campaign for @dob

This is a campaign thread for a Livepeer transcoder node:

0xFb9849b0B53F66B747Bfa47396964a3FA22400A0

Edit on 5/30/18 - After 28 rounds of running with a 0% reward cut, I intend to move to a reward cut parameter of approximately 10-15%, to encourage decentralization and increase transcoder competition. See below update. As a delegator, expect to keep 85-90% of the rewards in proportion to your stake.

Initially I intend to run a fee share of 0% or near 0%, so I will keep the small fees that broadcasters pay to use the infrastructure I’m running.

Who Am I?

I’m Doug Petkanics, founder of Livepeer. You can find me:

Why do I want to run a transcoder on Livepeer?

I am interested in running a node in the early days of the Livepeer network for two reasons:

  1. To keep a pulse on the needs of transcoders as a critical role in the Livepeer network. By playing this role for a little while it will better inform our engineering roadmap, and product decisions around user needs.
  2. To ensure reliable service for early developers and broadcasters looking to test the network during the alpha.

I am not interested in running this node for short term financial gain, nor to make a business out of scalable transcoding. I much prefer to resign and delegate towards other, more qualified nodes who are competing to provide this service in a cost effective and scalable way, as soon as they emerge from the ecosystem.

Why you may want to delegate towards me

I pay very very close attention to Livepeer in all regards. I am intimately familiar with the protocol having designed it. I care a lot about the success of the network. During the early days I want to ensure that broadcasters and developers have a good experience using Livepeer, so I am happy to spend quite a bit of time running and monitoring this node, fixing things that break, and updating the node software to run more reliably.

Why you may not want to delegate towards me

I suggest you delegate towards other nodes if you believe they will provide reliable service that can be cheaper or more scalable than what I am providing. I also suggest you delegate towards other nodes if they are bringing some alternative benefit to the project such as development or community building.

I intend to resign this node as soon as more focused, reliable, and scalable node operators emerge and prove that they can run reliably.

What hardware will I run?

I will not be spending significant time, money, or effort on dedicated hardware for this purpose, as I intend to resign this node as soon as reliable transcoders emerge who are doing so. I’ll be running cloud instances that are CPU optimized in order to handle the FFMPEG-based CPU encoding workflow that the Livepeer node currently supports.

As I experiment with redundancy solutions and monitoring setups, I will update this post to report what works.

My goals are to call reward() every day to distribute newly generated token, and to have > 96% availability for transcoding jobs during the first month of the network.


I hope this gives transparency into my motivations for running a node and how I intend to operate it. I’m happy to answer any questions that people have as they’re deciding who to bond towards.

And I look forward to more inspirational and competitive campaign threads than this one, from all the great people building amazing things in the open video ecosystem!

3 Likes

Reward Cut Update (5/30/18)

After 28 rounds with a 0% rewardCut, I will begin to gradually update the rewardCut on this node from 0% towards approximately 10-15%. This change will take place slowly over a 2 week period, in order to give delegators a chance to react, and is starting with a 2.5% reward cut in round 990. There are two reasons for this change:

  1. Encourage decentralization - because of the 0% reward cut, this node has accumulated over 30,000 LPT worth of delegated stake, or approximately 40% of the total amount. It was the best deal for delegators, who would maximize their LPT each round by choosing this node. I would like to see a more decentralized network in which many competitive transcoders are staked more proportionately. Leaving room for other transcoders to both cover their costs with a higher rewardCut, yet still remain competitive, will help to strengthen the network.
  2. Increase competition - There are a number of transcoders who are running with sub 10% rewardCuts who have been doing a great job of calling reward() every day, and participating in the protocol discussion in the forum and chat. This change should give them a chance to compete for delegation and build up a stake of LPT that is commensurate with the effort they’ve been putting in. Likewise, there are also a number of great community driven efforts keeping a higher rewardCut of 25-50%, who are being transparent about their use of the excess token they’re keeping in order to grow the community or help the Livepeer ecosystem. Removing a big attractive 0% rewardCut from the top of the transcoder list will hopefully shine more attention on all of these nodes, and encourage delegators to do their diligence prior to just selecting the first node with the lowest cut upon arriving to delegate.

Node performance update

Through the first 28 rounds, this node has had near 100% uptime, and has called reward() every day, never missing a call - resulting in maximum LPT for its delegators. I will continue to run the node responsibly, and work for the network on your behalf!

2 Likes

Three and a half years ago when I started running this node, I shared that I was doing so in order to be in close touch with the orchestrator/transcoder experience, but that I wouldn’t be running a robust transcoding setup as I didn’t have the hardware. Since then, 99 other node operators have stepped up to provide far more reliable work on the network, and I think it is safe to say that they all help the network far more than this node does. I have been moving much of the stake away from this node recently towards high performers, and I encourage others to do the same - in line with our community goal of activating 66 high performing nodes..

I will be deactivating this node temporarily in order to encourage stake to move towards activating high performing nodes. While I may re-activate it in the coming weeks in order to continue debugging and staying in touch with the node experience, I likely will be doing so with low stake, high reward cuts to discourage delegation, and no guarantees of calling reward. If you have been staked towards this node and need help with identifying a high performing node to move towards, let me know. Happy to help.

-Doug

1 Like