Re: TCP_HIT vs TCP_MEM_HIT speed.

On Tue, 8 Feb 2000, Henrik Nordstrom wrote:

> Ok, then your question does make some sense.
>
> However, it is to few requests to say anything useful from. The
> statistical variation is too large. It may well be that there are a few
> large TCP_HIT messages which causes the kb/second to be high.
>
> My experience is that TCP_MEM_HIT is faster than TCP_HIT, but I have
> never measured it as average kb/s, only as request latency.

I assumed that the MEM_HIT was slow due to differnt sized requests, and
luck of the draw.. I just wanted to make sure it wasn't something stupid
on my part.

Thanks.
Received on Tue Feb 08 2000 - 18:49:18 MST