The problem was the client server. It was too small (1 core, 4G of memory).
On Monday, July 18, 2016 at 4:03:26 PM UTC+3, Rayland wrote: > > I have an HTTP server in AWS on an instance with 4 cores and 8g of memory. > The server is not processing anything, just replying with 500. > > I have another machine in AWS issuing POST requests with Vegeta. > > > > These are my results: > > *10k req/s * > Requests [total, rate] 10000, 9628.13 > Duration [total, attack, wait] 5.672831255s, 1.038623689s, > 4.634207566s > Latencies [mean, 50, 95, 99, max] 4.184377168s, 4.372545912s, > 4.970730896s, 5.101392199s, 5.378851954s > Bytes In [total, mean] 610000, 61.00 > Bytes Out [total, mean] 900000, 90.00 > Success [ratio] 0.00% > Status Codes [code:count] 500:10000 > > *5k req/s * > Requests [total, rate] 5000, 4927.80 > Duration [total, attack, wait] 2.400928423s, 1.01465221s, > 1.386276213s > Latencies [mean, 50, 95, 99, max] 1.161943215s, 1.248041144s, > 1.860136828s, 2.047804227s, 2.088889511s > Bytes In [total, mean] 305000, 61.00 > Bytes Out [total, mean] 450000, 90.00 > Success [ratio] 0.00% > Status Codes [code:count] 500:5000 > > *1k req/s* > Requests [total, rate] 1000, 1000.88 > Duration [total, attack, wait] 999.611806ms, 999.124863ms, > 486.943µs > Latencies [mean, 50, 95, 99, max] 647.43µs, 531.921µs, 1.380096ms, > 2.26174ms, 4.933884ms > Bytes In [total, mean] 61000, 61.00 > Bytes Out [total, mean] 90000, 90.00 > Success [ratio] 0.00% > Status Codes [code:count] 500:1000 > > > My questions is if it's normal to have those wait times for over 1k req/s > considering the server is not even reaching 50% of it's processing power? > -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.