Quote from: r0ach on October 06, 2015, 12:34:11 AM

Quote from: Fuserleer on October 05, 2015, 02:27:40 PM

Now, after your testing over the weekend where the test net died at a mere 1-2% of that claim

Oh please, where does it say 100kTPS on a 1 core CPU? They mentioned it would need something like 32 cores to do that. Some guy named "clayop" in Korea ran a test while Larimers weren't even around that peaked at 3k TPS and some of the delegates running were single core VPS I believe. It's a dumb test because he doesn't even list hardware used or what the bottleneck seen was, etc. They need to make sure nobody is running some gimpy single core VPS before doing any more tests.

100k / 32 = 3.125k tps on a 1 core....it should be attainable using that math, yet it died.

Anyway the issue I believe wasn't CPU but networking, which I've pointed out time and time again either that or disc IO would be a problem....and it is.

Doesn't matter if the Larimers were around or not, its not a dumb test. If you are continuously claiming that your network can support 100k tps sustained "conservatively", it should be able to handle a few 1000 tps peaks without issue regardless of hardware in use, whether the overlords are present or not etc etc

What happens in the future if some delegates are voted in that have "gimpy hardware".....it croaks and falls flat on its back? :|