[bind10-dev] more about cache (Re: cache effectiveness)

[Michal 'vorner' Vaner]

Hello

On Fri, Mar 01, 2013 at 10:59:06AM -0800, JINMEI Tatuya / 神明達哉 wrote:
> > I noticed you kind of implicitly assume threads. I guess nothing would change
> 
> Not necessarily, but admittedly I have a feeling that for some large
> scale deployments using threads is quite likely to be the only
> feasible choice to achieve highest response performance without
> bloating memory and/or network.

I'm not sure about it. If we want to be fast, we need the threads to be mostly
independent (no locking, and waiting for the other threads). Such things have
quite high overhead and they don't scale well. And, as I said, the threads are
hard to work with.

So, if we make the threads independent (not communicate with each other too
much), they might as well be separate processes.

> That might actually be an option if we can assume the "layer-1 cache"
> covers the vast majority queries.  But it would depend on many things
> whether this design is really effective, so I suspect we cannot be
> sure at least without some actual experiments.

I guess it's interesting enough idea that we should do some kind of experiment
to confirm or reject it. Do you have an idea how to make such experiment, other
than the analysis of the data you did before?

With regards

-- 
Hello, I'm an extension to the infamous signature virus, called spymail.
Could you please copy me into your signature and send back what you were
doing last night between 10pm and 3am?

Michal 'vorner' Vaner
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <https://lists.isc.org/pipermail/bind10-dev/attachments/20130304/6d3d30c5/attachment.bin>