The EOS Node Operator Roundtable meets every Wednesday at 14:00 UTC to prepare for the next EOS consensus upgrade and share feedback on opportunities to improve the Antelope protocol for node operators.
This week Brian Hazzard (ENF) lead the discussion to gather feedback on the challenges we're looking to address with potential changes to the Antelope resource model. As can be seen with the current congestion issues on WAX, there may be opportunities to improve the resource model on Antelope chains. In preparation for a future discussion on potential solutions with Areg (ENF), this discussion was focussed on documenting the pain points observed by node operators that we might want to address with changes to the resource model. For security reasons, this call was not recorded - but it's been summarized below.
The next discussion (January 11) will focus on the implications of the Resource Model to Service Node (like history and api).
Resource Model Pain Points
⁃ RAM usage
⁃ Will eventually run into physical hardware limitations cap for lower grade hardware OR economic viability and accessibility concerns at higher grade hardware.
⁃ Any new resource model should better align incentives/disincentives for storage utilization relative to the available storage, especially with regards to RAM
⁃ Any new resource model should provide a means to increase the available storage on the network in a cost effective way
⁃ Ease of use concerns
⁃ Goal is to balance network congestion, abuse prevention, usability, and costs.
⁃ A user-paid Gas model would be more user friendly than raw resources, rex, or powerup, but has it’s own usability concerns due to cognitive load of micro-transactions and user psychology related to loss aversion.
⁃ Dapp / Contract payer solutions would be the easiest for a user, but would need to be economically viable for the contracts and solve for abuse prevention.
⁃ We could evaluate what users, developers, wallets, and node operators are already doing to work around limitations and shortcomings of the resource model, and implement a native solution that codifies and improves on those solutions.
⁃ A NET resource should probably not be part of the future resource model solution
⁃ Powerup
⁃ Even daily expert users can’t figure out how the ratios work
⁃ Daily rental is a lot of mental overhead (WAX has 30 day PowerUp rentals)
⁃ Some people just bake in PowerUp on each transaction (adhoc gas)
⁃ Might be able to improve, rather than replace PowerUp
Plans for the Next Discussion
⁃ Discuss the implications of the Resource Model to Service Node (like history and api)
The EOS Node Operator Roundtable meets every Wednesday at 14:00 UTC to prepare for the next EOS consensus upgrade and share feedback on opportunities to improve the Antelope protocol for node operators.
💬 Join us in Telegram 🗂 See all meeting summaries
🗓 January 4 - 14:00 UTC
🎥 No recording this week
Technical roundtable for EOS Node Operators
This week Brian Hazzard (ENF) lead the discussion to gather feedback on the challenges we're looking to address with potential changes to the Antelope resource model. As can be seen with the current congestion issues on WAX, there may be opportunities to improve the resource model on Antelope chains. In preparation for a future discussion on potential solutions with Areg (ENF), this discussion was focussed on documenting the pain points observed by node operators that we might want to address with changes to the resource model. For security reasons, this call was not recorded - but it's been summarized below.
The next discussion (January 11) will focus on the implications of the Resource Model to Service Node (like history and api).
Resource Model Pain Points
⁃ RAM usage ⁃ Will eventually run into physical hardware limitations cap for lower grade hardware OR economic viability and accessibility concerns at higher grade hardware. ⁃ Any new resource model should better align incentives/disincentives for storage utilization relative to the available storage, especially with regards to RAM ⁃ Any new resource model should provide a means to increase the available storage on the network in a cost effective way ⁃ Ease of use concerns ⁃ Goal is to balance network congestion, abuse prevention, usability, and costs. ⁃ A user-paid Gas model would be more user friendly than raw resources, rex, or powerup, but has it’s own usability concerns due to cognitive load of micro-transactions and user psychology related to loss aversion. ⁃ Dapp / Contract payer solutions would be the easiest for a user, but would need to be economically viable for the contracts and solve for abuse prevention. ⁃ We could evaluate what users, developers, wallets, and node operators are already doing to work around limitations and shortcomings of the resource model, and implement a native solution that codifies and improves on those solutions. ⁃ A NET resource should probably not be part of the future resource model solution ⁃ Powerup ⁃ Even daily expert users can’t figure out how the ratios work ⁃ Daily rental is a lot of mental overhead (WAX has 30 day PowerUp rentals) ⁃ Some people just bake in PowerUp on each transaction (adhoc gas) ⁃ Might be able to improve, rather than replace PowerUp
Plans for the Next Discussion
⁃ Discuss the implications of the Resource Model to Service Node (like history and api)
Participants (12)