I have a customer who has the following issue
We're seeing some really strange results as we benchmark our new dedicated servers before putting them into full production.
In essence, we have conducted a series of siege tests from Server3 , through the public network to a Gateway to Server2.
Notice that the ability of Server1 to generate requests seems to fluctuate DRAMATICALLY, which flows through the system as expected.
is anything we're missing that might erratically influence the amount of requests we can generate in our siege test? As near as we can tell, this should be a straight forward test that produces consistent results, and could use a little insight .
So basically every server has a private and a public interface.
Any input would be appreciated.