James Mansion wrote:
> Ivan Voras wrote:
>> I wish that, when people got the idea to run a simplistic benchmark
>> like this, they would at least have the common sense to put the
>> database on a RAM drive to avoid problems with different cylinder
>> speeds of rotational media and fragmentation from multiple runs.
> Huh?
>> It's tough to benchmark anything involving rotational drives :)
> But - how the database organises its IO to maximise the available
> bandwidth, limit
> avaiodable seeks, and limit avoidable flushes is absolutely key to
> realistic performance,
> especially on modest everyday hardware. Not everyone has a usage that
> justifies
> 'enterprise' kit - but plenty of people can benefit from something a
> step up from
> SQLite.
>
> If you just want to benchmark query processor efficiency then that's one
> scenario
> where taking physical IO out of the picture might be justified, but I
> don't see a good
> reason to suggest that it is 'common sense' to do so for all testing,
> and while the
> hardware involved is pretty low end, its still a valid data point.
> .
You are right, of course, for common benchmarking to see what
performance can be expected from some setup in some circumstances, but
not where the intention is to compare products.
You don't have to go the memory drive / SSD route - just make sure the
databases always use the same (small) area of the disk drive.