1. Development plans about the Roadmap, Victor and Attila talked about their thoughts.
    • Support multi-platform(not only hyperledger project). We can start from Corda and then Quorum. Of course current supported Adapters should continue to be updated.
    • GUI : the other fabric performance tool gived a GUI which makes beginners start easy. Caliper is lack of this. We have already published a internship program about this.
    • Monitoring：now only use some original tools to monitor, it’s better to integrate with Grafana.
    • Reporting: well formatted file is needed.
    • Installation process should be considered again. We might consider about providing Npm package, Docker Image and Vm.  
      • Nick will be responsible for Npm package. 
      • Kent said someone from PSWG would like to discuss with Caliper about hooking with fabric.  We need to attend next PSWG meeting.
    • Efficiency: some companies used caliper to test TPS with 10,000.  Caliper’s client is not able to support such large sending rate. And the backend blockchain will crash. So we need a  Backend Simulator and increase sending rate.
      • Nick: it's better first to upgrade Caliper to Node10, and use multiple threads to send txs.  
      • Attila: Fabric's paper gives a mock tool. We can take a look at it.
    • Stability: Caliper is not able to run 24h/ 10k TPS.
    • Documentation: Tutorials Slides(more details, all use this slide ), detailed configuration explanation with examples. The code of the example should be generated from the a tool.
      • Kent would like to contribute this.
    • Version:  after an API is fixed, a release will be published. 
      • First publish API documentation, make an API list
      • Qinghui will make an API list, take into count that these types of API might be necessary: external user interface, internal, adapters. 
      • Tokens are necessary for a release.
  2. Findings on the bootcamp: Hrfd , functionality is same as caliper. It has integrated with grafana and also has distributed clients, which are deployed on different dockers. However, it depends on kubernetes, which might be a problem for usage. We need to look into it  and see if there is any other things we can study from . 
    • Nick can get documentations to share.
  3. Attending an event in May: Nick will check the schedule and try to attend.
  • No labels