2019 Q4 Hyperledger Besu

Created by Danno Ferrin, last modified by Swetha Repakula on Dec 12, 2019

Project Health

Besu is strong and growing. This quarter Besu has prioritized switching to the Hyperledger tools and growing its community. Besu has been successful on both fronts while also delivering its 1.3 Release in October 2019.

Questions/Issues for the TSC

  • RocketChat presents an unneeded point of friction bring in members from the larger Ethereum community.  With Gitter we could send a link directly to our project room and anyone could read it without logging in.  Similar links for rocket chat (example: https://chat.hyperledger.org/channel/besu ) present a never ending loading screen unless the user has already logged in.  

    • We ask the Hyperledger operations employees to enable a configuration option that would permit read-only access to ch”at rooms without authentication (“ Allow Anonymous Read ”).
  • The Besu team is still looking for clear guidance on active status from the TSC. Per the latest communication from the TSC, we believe they want 3 maintainers from different organizations (Note that this is on track to be achieved on 10 Dec). Can we gain confirmation th”at is the requirement that will enable the TSC to vote “yes” on active status?

Releases

Besu has released one minor version update (1.3, 3 Oct) and six bi-weekly patch releases (1.2.4 on 23 Sep and 1.3.1 through 1.3.5 on 15 Oct, 22 Oct, 5 Nov, 5 Nov, and 20 Nov respectively).

Major functional improvements include

  • Support for Ethereum Classic and Classic testnets (from Edward Mack and Greg Markou of Chainsafe Labs)
  • Support external mining for PoW networks (from Antoine Toulme of Whiteblock.io )
  • Ethereum Mainnet Istanbul support
  • Added state pruning of blockchain data

Overall Activity in the Past Quarter

Hyperledger Besu was fully transitioned from a PegaSys project to a Hyperledger project this quarter.

  • Github repos moved to Hyperledger org
  • Bug tracking moved to the Hyperledger JIRA
  • Chat support room moved from Gitter to Hyperledger’s Rocket Ch”at - CI/CD build infrastructure moved from Jenkins to Hyperledger’s CircleCI
  • Established a bi-weekly contributor call

As a part of joining the Hyperledger community, the Besu team has participated in the community by:

  • Added support for Besu in Caliper (Oct. 20)
  • Participated in the CI/CD working group conversations and sharing our project’s research and best practices (Sep.- Oct.)
  • Attended the Maintainer summit in Minneapolis (Oct. 8-9)
  • Presented a talk at the Hyperledger booth at Sibos (Sep. 23-25) 
  • Submitted several talks and were approved to present at the Hyperledger Global Forum (Sep. - Nov.)
  • Attended the Hyperledger meet-up at Devcon in Japan (Oct. 9)
  • Presented privacy solution at Architecture WG - Privacy Confidentiality track (Sep. 20)
  • Presented at Hyperledger meet-up in Japan (Dec. 3)

Current Plans

The Besu project team has several upcoming priorities:

  1. The project team is currently working towards its 1.4 Release, scheduled for February 2019. The 1.4 Release is expected to include the following features:
  2. Multi-tenancy
  3. End to end TLS on network protocols
  4. Public network hard forks (Muir Glacier and Aztalan potentially)
  5. The Besu team is also focusing on finalizing and publishing performance metrics with its work on Hyperledger Caliper. Besu was successfully integrated with Caliper in Q4 and the team is now in the process of testing various scenarios for Caliper. We think publishing these performance metrics will help provide the community additional context on the project’s useability for a variety of use cases.
  6. Besu is also continuing to engage with its community and grow the diversity of its maintainer and contributor base. The Besu team is hoping to continue to grow attendance for its bi-weekly contributor calls to grow engagement across contributors.

Maintainer Diversity

Our maintainer diversity is growing.  Early next week we should have maintainers representing three different organizations.

  • At the beginning of the quarter Besu was still a PegaSys project, so all the maintainers at the beginning of the quarter are PegaSys employees.
  • Ivaylo Kirilov of Web3 Labs was added because of strong contributions in EEA Privacy
  • Edward Mack of Chainsafe labs is in the process of being added because of strong contributions in Ethereum Classic support.

Our process for adding Maintainers was featured on the Hyperledger Blog as a best practice for a maintainer governance process - https://www.hyperledger.org/blog/2019/11/07/best-practices-and-lessons-learned-from-hyperledger-besu-establishing-a-maintainer-process

100% of current maintainers made a contribution this quarter.

Contributor Diversity

Since the Besu project was approved by the TSC into the Hyperledger community at the end of August 2019, there have been a total of 39 code contributors.

  • 23 contributors from PegaSys (58% of contributors)
  • 16 contributors from outside of PegaSys (42% of contributors)
  • 7 different organizations that can be identified
  • 7 contributors whose organization were not determined.

Additionally, there have been content contributions in the community, including technical content enablement materials.

Additional Information

None at this time.

Reviewed by

Comments:

  • The Besu team is still looking for clear guidance on active status from the TSC. Per the latest communication from the TSC, we believe they want 3 maintainers from different organizations (Note th"at this is on track to be achieved on 10 Dec). Can we gain confirmation that is the requirement that will enable the TSC to vote “yes” on active status?

From Project Incubation Exit Criteria :

  • The project is not highly dependent on any single contributor (there are at least 3 legally independent committers and there is no single company or entity that is vital to the success of the project)


The key criteria that most were commenting on was “there is no single company or entity that is vital to the success of the project”. With 19 of 20 maintainers coming from one company, the concern is that if th"at one company pulls out their support then there is only 1 maintainer from the project.

Posted by tkuhrt at Dec 03, 2019 08:40

As noted in the maintainer diversity, we have a maintainer from Chainsafe that is coming on board as a maintainer when the nomination window closes next Tuesday.  That will then have maintainers from 3 companies, PegaSys, Web3 Labs, and Chainsafe, filling the 3 maintainers criteria. 

As far as sponsoring companies (those who pay people to work) ETC Cooperative will be funding work, mostly through chainsafe.  So if ConsenSys were to pull out (we won't) ETC Co-Operative would be able to keep the project alive, like they already do for some of the other ETC clients.

Posted by shemnon at Dec 03, 2019 15:40

2/21 = 9.5% maintainers that are not part of PegaSys doesn’t meet the “there is no single company or entity that is vital to the success of the project” portion of the requirement though.

Posted by tkuhrt at Dec 03, 2019 16:34

If there were less PegaSys maintainers would that satisfy your concern?

Posted by shemnon at Dec 03, 2019 17:08

This isn't a serious question, is it? Obviously, artificially balancing the number of maintainers by removing PegaSys employees wouldn't address the crux of the problem, would it?

If anything, the fact that this question was raised shows the challenge of trying to define an objective measure in this case. Any hard number will evidently lend itself to possible gaming. This is very much the kind of manipulation Hart Montgomery has been warning us about when discussing election eligibility.

I honestly don't know what to make of this. :-/

Posted by lehors at Dec 04, 2019 13:03

When the objection is expressed in percentages this is the incentive it brings.  But on the other hand we would like some guidance as to when we should ask again, so we don't ask too early.  Without some measure it boils down to "when the TSC feels like it" and it's hard to measure when the proposal should be raised again with that form of a standard.  We have had Bob Somerwille of ETC cooperative state he would finance besu development if needed if ConsenSys were to pull out.  I have yet to hear any TSC member address this and instead every time this is mentioned the subject is changed to the number of maintainers.

There was also an objection expressed earlier in a different forum that too many maintainers may present a security challenge in coding reviews, so this isn't the first time we've heard the "Too many PegaSys maintainers" concern, and we've heard it in more than just a diversity context.

Posted by shemnon at Dec 04, 2019 15:04

As clarification, do you mean that "too many PegaSys maintainers" will lead to a security challenge or in general too many maintainers lead to that challenge? Specifically what is the security concern?

Posted by swetharepakula at Dec 12, 2019 00:14

I think the concern was too many maintainers in general.

Posted by shemnon at Dec 15, 2019 05:21