Knut Anders Hatlen wrote:
Kristian Waagan <[email protected]> writes:
Do you think the numbers look acceptable?
Is there anything we should investigate further?
Hi Kristian,
I think the numbers look more than acceptable. Most of them are a lot
better than 10.3 and 10.4, some even orders of magnitude better. Those
few test that don't show an improvement have more or less the same
performance as before, so overall the situation seems to be greatly
improved.
Hi Knut,
That is true.
Nevertheless, if people suspect performance degradation due to the recent work, they are more than welcome to write tests and add them to [BC]lobAccessTest.
Measuring something is a requirement to make sure regressions aren't introduced, and often the first step to improve performance :)
Actually running the tests is required too, of course!
--
Kristian
