Closed Cyrix126 closed 3 weeks ago
fixed in https://github.com/Cyrix126/gupaxx/commit/5a4759109f90971baf3c6ee43a0facf4d3486547 and https://github.com/Cyrix126/gupaxx/commit/279bdcc678bcf5b116587a1a3369fee727fd1bc6
Issue will be closed at next release.
Bug
Since node does not implement ICMP, HTTP requests are made to find the latency of remote nodes (including the ones of XvB). Since only one request per node is made, the latency found can be a bit random. Currently, this means the list of remote nodes and the automatic chosen XvB node can sometimes be inaccurate.
Steps
Steps to reproduce the behavior:
Go to P2pool tab
Click on "Ping remote nodes"
See that the results is not always accurate.
Start XvB process
See that sometimes the output console says that the wrong one has been chosen.
Solution
When testing the latency, make multiples request and take average.