I'm opening a new PR since I switched branches, but the old one is here for comment reference. I fixed the bugs...looks like it works now.
A new branch is a good idea; I just got lazy since I was working in my fork. The branch has been changed and I rebased it against main. Do you want me to target another branch with my PR as well?
I think I've got docker-compose correct, but it got slightly complicated when rebasing, so if you could verify, that'd be great. I just commented out the CERN container for now.
A couple other comments:
The default update time is set as 300sec, but it takes a while to actually get through all of the GEANT routers. Is there any merit in extending that time?
Not every query returns a peer ("learned-from") for some reason. The code handles it without failing, but I'm not sure how that might affect what's built on top of this.
Many (~10?) of the routers time out each update() cycle, but it's not the same 10 every cycle. It might be beneficial to limit the time given for each request to minimize the delays from these timeouts.
I'm opening a new PR since I switched branches, but the old one is here for comment reference. I fixed the bugs...looks like it works now.
main
. Do you want me to target another branch with my PR as well?A couple other comments: