On 23.03.2011 10:42, Michael DELVA wrote:
Thanks!
What do you think would be relevant to bencmark? I was thinking at the
creation time of the proxy, and the access time, averaged on several
thousands of them, but I don't if it is enough.
What is your opinion?
There are two common kinds of benchmarks: synthetic and "real-world".
Synthetic benchmarks measure the cost of a specific operation isolated
from all external factors. For example, you could compare Find/Get
performance with or without your cache.
Real-world benchmarks on the other hand, use a (often simulated)
workload to measure how the system under test performs "end-to-end". For
example, you could setup an ASP.NET webpage using NHibernate and compare
load test results with and without your cache.
Of course, all benchmarks are heavily biased by the (often unconcious)
choice of data, performance measures, hardware, and other (often
invisible) external factors.
Best Regards, David
--
dasz.at OG Tel: +43 (0)664 2602670 Web: http://dasz.at
Klosterneuburg UID: ATU64260999
FB-Nr.: FN 309285 g FB-Gericht: LG Korneuburg
