> Jon Kay wrote:
> > You telling us that iMimic caches don't support ICP, Cache Digest, or
> > even cache hierarchies? Really??? Yet another reason to go Squid!
>
> This argument is getting tiring. I responded to Joe because he
> (possibly unwittingly) misrepresented my previous statements.
Ahem. This particular part of the argument descends directly through
your response to my own posting. Surely you don't expect me to not
defend myself in public :-).
"He did it first!!!" :-)
> Now you seem to be intentionally doing so. Nowhere did I make
> the sorts of statements that you suggest above. I see no point in
> continuing.
I plead guilty to "spinning."
That said, there is a legitimate point that I was getting at
concerning cache hierarchies. You guys spent time implementing ICP.
Why is that, if you feel coordinated caching is a bust?
> > There you go. Your users are 1/8 faster if you go Hint Cache. And
> > it's all far more scalable.
> As for your strawman comparison, it would be a lot more believable
> if the numbers were from an actual large deployment,
On the contrary, that's what benchmarks are for. Just which point
during the day should one take deployment numbers for? How typical is
that deployment? Is it atall comparable to this Squid (or pushcache)
deployment?
> or if you'd
> used the best DataReactor numbers (17ms hit, 2652ms miss).
I used the numbers from your own box! Are you disowning them?!?
Plus, the iMimic throughput numbers are slightly less than 10x the
Swell/Squid numbers, which, since I was talking about 10 Squids vs. 1
iMimic, happens to model the relative load factors nicely.
One point I am trying to make is that the central box is a box with
heavy load on it, and thus it has trouble keeping the latency down.
> One tech report or paper at ICDCS doesn't constitute reality. Our OEMs
> are seeing numbers in the field in-line with the throughputs they've
> shown at the cacheoffs, and in many cases with even higher hit
> rates.
An implementation is reality.
> I'm not against hint caches or cooperative deployments per se -
> in fact, if they become popular, one of our OEMs would probably
> love to sell a few hundred units at a time to interested parties.
> This scenario would be ideal for the "microcache" segment of the
> most recent cacheoff, where both entries were sub-$1000 and showed
> capacities of 120-325 req/sec.
We'll talk to just about anybody. Y'all come on down!
-- Jon Kay pushcache.com jkay@pushcache.com http://www.pushcache.com/ (512) 420-9025 Squid consulting 'push done right.'Received on Sat Dec 22 2001 - 23:05:12 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:05:28 MST