[nycbug-talk] network protocols: lab time?

[G.Rosamond]

On May 12, 2004, at 2:14 PM, Pete Wright wrote:
>>
>> We should either use multiple protocols on one OS/platform, or one 
>> protocol with multiple plaforms/OS's.
>>
>> It seems  that the main contention is about protocols, so maybe we 
>> stick to one platform/OS and test multiple protocols.
>>
> yea that a good point, maybe we should send an email to the apple 
> folks for an Xserve/OSX server lisc. for this.
> -p

We are still waitiing on them to give us a time and place to test the 
Xserves.  Almost 30 people signed up for that.  It will probably be on 
a weeknight in June, I'd assume.

I do think we should be a bit more methodical . . .

Yes, we could contact Apple, but we'll have to be a bit more clear.

If we are testing networking protocols, then we could try a *BSD box 
and an XServe, and compare how they are with NFS, SMB, etc.  We'd be 
comparing Xserve to a BSD, but the specs on each wouldn't be 
comparable.

The better bet, IMO, is to try using i386 boxes with the same specs, 
with each of the relevant OSs (OpenDarwin on i386), then test each 
protocol that way.

It would be much easier to organize, and I think I could find a vendor 
that would loan us the equipment. . .

Of course, we'd have to determine a criteria for tweaks. . .or even 
basic config.

Anybody got a good location?

g