2020-12-22 Meeting notes

Date

Attendees

Goals

Discussion items

WhoNotes
  • Seems that I found a reason for a single problem
    • In the case of bad network and races between peers, there is a chance that the flow of transactions cannot be handled correctly. I have described all the concerns to Andrei Lebedev and proposed the simplified solution.
    • Andrey agreed that the proposed solution may help.
  • Regarding the timer, I just increased the value to check the influence of it with Konrad.
  • There is still a single issue left, so I will work on it after I finish the current improvements. 
    • Seems like there is a problem in consensus. I will start with a deeper code review to find the reason and behaviour issues.
  • We will be ready to pass the custom build for Konrad today. Will contact him and send the build.
  • Sara is trying to understand the prepared "AddPeer" script.
  • Working on the fix for one of the metrics for performance testing. Seems like Influx DB selected wrong data format because of our approach (int instead of float). We need to create a mechanism which will explicitly set all types in the InfluxDB.
  • We also need to find a reason why Locust is crashing after 10 minutes of load. We have a workaround, but it will be better to fix it in an elegant way.

Action items

  •