Replies: 2 comments 2 replies
-
I agree with your assessment here and this is definitely a valid concern. Some variation is probably a good call to avoid any bias on transits. I've been following this project R0GGER/public-iperf3-servers pretty closely lately and like what I see so I'm thinking about making the iperf locations that YABS selects somewhat dynamic or random using locations pulled from that project. Just an idea I've been sitting on for a bit, but I'm open to new ideas too. |
Beta Was this translation helpful? Give feedback.
-
Hello, i'm pretty new to this benchmark script, as @hohl pointed out, i also want to suggest if you can add some Asia locations (Singapore, Jakarta, Australia or Hongkong if possible). Thanks. |
Beta Was this translation helpful? Give feedback.
-
The current selection of
iperf
-servers has quite a bias to a small number of global transit networks. Especially Arelion (fka Telia) and Lumen, which is in part due to the strong presence of Clouvider (and an overweight of European providers in general).I know the available choices of public
iperf
nodes is very limited—especially outside of Europe—, but I thought it might still be interesting to point this out. This could be a point to consider when adding/removing locations, so that we get some more diversity in the transit networks tested against.Btw. I like the recent addition of Uztelecom, it gives at least some variation!
Beta Was this translation helpful? Give feedback.
All reactions