Semi-dedicated vs Reverse Proxies OCR
Ran a test on semi-dedicated versus reverse proxies today as far as the number of appearances in a sample is concerned. The result:
Sample size: 451507 lines of log (just posting), 16 proxies in total, 150 threads
10 Semi-dedicated proxies:
--
excluded lowest fail count: 286
excluded highest fail count: 10455 (sic!)
--
average fail count excluding extremes: 1145 entries in log per proxy
5 Reverse proxy ports:
average fail (no extremes exluded possible): 998 entries in log per proxy
1 Dedicated proxy (accidentailly left on):
416 entries in log
Conclusion: It seems reverse proxy quality is almost similar to shared proxies. The 10k fail proxy is getting reported, this has to be worse even than public proxies :P. Couldn't do my private proxies due to a migration, but based on the previous results I would think they will get much better results if not every time, then at least in average.
What do you think?
Sample size: 451507 lines of log (just posting), 16 proxies in total, 150 threads
10 Semi-dedicated proxies:
--
excluded lowest fail count: 286
excluded highest fail count: 10455 (sic!)
--
average fail count excluding extremes: 1145 entries in log per proxy
5 Reverse proxy ports:
average fail (no extremes exluded possible): 998 entries in log per proxy
1 Dedicated proxy (accidentailly left on):
416 entries in log
Conclusion: It seems reverse proxy quality is almost similar to shared proxies. The 10k fail proxy is getting reported, this has to be worse even than public proxies :P. Couldn't do my private proxies due to a migration, but based on the previous results I would think they will get much better results if not every time, then at least in average.
What do you think?
Comments