Absolutely fantastic video - and jammed packed with both scaling considerations, recommended tooling for measuring the capacity/performance, and some great tuning Besu for achieving the peak for tests and production. Keep these videos coming, please! I'm curious if Kaleido has considered other metrics beyond TPS? The main benefit of digital assets continues to be the ability to flexibly program the assets among increasing institutions or even end-users on public chains. Otherwise, if just measuring a simple transaction "Mary sends Bob 1 token," or similar limited use cases, most institutions might just continue to use a non-DLT architecture, and skip Besu altogether. An interesting alternative performance metric for Ethereum might be "gas per second" as that can help measure more complicated programmatic execution scenarios. It appears the vast majority of TPS performance is lost due to time in consensus, it would be interesting to consider metrics which can also exclude the consensus time (which was appropriately shown to be highly dependent on number of participating validator nodes) and instead focus solely on execution layer speed. Perhaps in the future we may see more enterprise chains consider to use EigenLayer's EigenDA or similar techniques as a means to achieve even higher TPS and gas per second scale with decentralized security?
That was an excellent watch, thanks!
Absolutely fantastic video - and jammed packed with both scaling considerations, recommended tooling for measuring the capacity/performance, and some great tuning Besu for achieving the peak for tests and production. Keep these videos coming, please!
I'm curious if Kaleido has considered other metrics beyond TPS? The main benefit of digital assets continues to be the ability to flexibly program the assets among increasing institutions or even end-users on public chains. Otherwise, if just measuring a simple transaction "Mary sends Bob 1 token," or similar limited use cases, most institutions might just continue to use a non-DLT architecture, and skip Besu altogether.
An interesting alternative performance metric for Ethereum might be "gas per second" as that can help measure more complicated programmatic execution scenarios. It appears the vast majority of TPS performance is lost due to time in consensus, it would be interesting to consider metrics which can also exclude the consensus time (which was appropriately shown to be highly dependent on number of participating validator nodes) and instead focus solely on execution layer speed. Perhaps in the future we may see more enterprise chains consider to use EigenLayer's EigenDA or similar techniques as a means to achieve even higher TPS and gas per second scale with decentralized security?
Thanks for taking the time to watch!
hello, thanks for the wonderful sharing, Is there a link to download the slides? thanks
Hi there! Send your email to marc.lewis@kaleido.io and I'll send follow up materials.