gsgreen | Dark Hacker News

user:	gsgreen
created:	April 16, 2026
karma:	4
about:	Working on Webbynode — dataset-driven VPS benchmarking and network analysis.

1.	Everyone thinks AWS t3.large is calm. 3 benchmark runs may change things(webbynode.com) 1 points by gsgreen 21 hours ago \| 1 comment
2.	We're seeing distinct runtime lineages emerge inside the same cloud plan/region(webbynode.com) 1 points by gsgreen 1 day ago \| 0 comments
3.	We kept getting different "personalities" from the same cloud VM SKU(webbynode.com) 1 points by gsgreen 2 days ago \| 0 comments
4.	Three AWS VPS Runs Looked Identical – One Still Failed Under Load(webbynode.com) 1 points by gsgreen 7 days ago \| 3 comments
5.	Google Cloud and DigitalOcean Behaved Differently Under Repeated Deployments(webbynode.com) 1 points by gsgreen 9 days ago \| 0 comments
6.	DigitalOcean NYC looked fine – until we ran it repeatedly (Singapore held up)(webbynode.com) 1 points by gsgreen 15 days ago \| 0 comments
7.	DigitalOcean's NYC region looked fine – until we ran it again(webbynode.com) 3 points by gsgreen 16 days ago \| 2 comments
8.	Coding agents expose this: same VPS, 3 runs, ~65% drift(webbynode.com) 1 points by gsgreen 21 days ago \| 1 comment
9.	Fast VPS ≠ reliable: we tested 24/7 workloads – consistency beat peak speed(webbynode.com) 1 points by gsgreen 22 days ago \| 0 comments
10.	Most VPS benchmarks are wrong – performance drops after the first run(webbynode.com) 2 points by gsgreen 23 days ago \| 0 comments
11.	Same AWS plan, same continent – different behavior under load(webbynode.com) 2 points by gsgreen 28 days ago \| 1 comment