[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: request for data

Hi Joe, 
After much work and deliberation, we think that maintaining your feed from
"thelma" is the best option at this time. This keeps you at the top level
for data distribution and takes advantage of abilene and vbns connections.

Enjoy your top level connection,

-Jeff
__________
Jeff Weber
Unidata/NWS-COMET Case Study Library
University Corp for Atmospheric Research

address@hidden
PH:303-497-8676   
URL--http://www.unidata.ucar.edu/staff/jweber
________________________________________________

On Wed, 9 Feb 2000, Joe St Sauver wrote:

> >Date: Wed, 09 Feb 2000 15:40:51 -0700 (MST)
> >From: Jeff Weber <address@hidden>
> >Subject: Re: request for data
> >To: address@hidden
> >Cc: address@hidden
> 
> Hi Jeff!
> 
> >I am helping Robb with the IDD support now and can serve as your contact
> >point.
> >
> >I am thinking that the University of Washinton may be a good site for you
> >to link to.
> >
> >Perhaps you could do a traceroute and ping to University of Washingtons
> >machine:
> >
> >grayskies.atmos.washington.edu
> >
> >If quick it may be our best choice.
> 
> Our topology actually *least* favors UW relative to other regions -- our two 
> Abilene OC3's I2 go south to Sacramento and east to Denver, which means
> that traffic going to Washington will first go all the way to Calren North
> and then up to Seattle, e.g.:
> 
> traceroute to grayskies.atmos.washington.edu (128.95.89.2), 30 hops max, 40 
> byte packets
>  1  cisco3-gw (128.223.214.1)  2 ms  2 ms  1 ms
>  2  cisco7-gw (128.223.2.7)  1 ms  2 ms  2 ms
>  3  ogig-uo.oregon-gigapop.net (198.32.163.1)  2 ms  1 ms  1 ms
>  4  sac.oregon-gigapop.net (198.32.163.10)  11 ms  11 ms  11 ms
>  5  sttl-scrm.abilene.ucaid.edu (198.32.8.9)  25 ms  25 ms  25 ms
>  6  abilene-gsr-POS1-0.pnw-gigapop.net (198.48.91.5)  25 ms  25 ms  25 ms
>  7  uwbr1-GE0-0.cac.washington.edu (198.32.170.12)  25 ms  26 ms  25 ms
>  8  dustbuster-V50.cac.washington.edu (140.142.155.2)  26 ms  26 ms  26 ms
>  9  grayskies.atmos.washington.edu (128.95.89.2)  27 ms (ttl=56!)  26 ms 
> (ttl=56!)  26 ms (ttl=56!)
> 
> Compare that to:
> 
> traceroute to amber.Berkeley.EDU (128.32.25.12), 30 hops max, 40 byte packets
>  1  cisco3-gw (128.223.214.1)  3 ms  2 ms  1 ms
>  2  cisco7-gw (128.223.2.7)  2 ms  1 ms  1 ms
>  3  ogig-uo.oregon-gigapop.net (198.32.163.1)  2 ms  2 ms  2 ms
>  4  sac.oregon-gigapop.net (198.32.163.10)  16 ms  11 ms  11 ms
>  5  BERK--abilene.POS.calren2.net (198.32.249.41)  16 ms  14 ms  14 ms
>  6  pos1-0.inr-000-eva.Berkeley.EDU (128.32.0.89)  14 ms  14 ms  14 ms
>  7  pos5-0-0.inr-001-eva.Berkeley.EDU (128.32.0.66)  15 ms  15 ms  15 ms
>  8  fast1-0-0.inr-007-eva.Berkeley.EDU (128.32.0.7)  15 ms  16 ms  16 ms
>  9  f8-0.inr-100-eva.Berkeley.EDU (128.32.235.100)  16 ms  16 ms  16 ms
> 10  amber.Berkeley.EDU (128.32.25.12)  16 ms (ttl=55!)  16 ms (ttl=55!)  15 
> ms (ttl=55!)
> 
> with half the latency... Any chance that you might have a feed available from 
> a Calren North school? [The Michigan machine-in-California you mention sounds
> great, for example.]
> 
> >I am also investigating a machine in California operated by the University
> >of Michigan. 
> 
> That would be far better for us topologically (and politically, although I'd
> rather not get into that). 
> 
> Another option for us would be a Verio connected site (we peer with Verio),
> and soon we'll have additional "tier 1" ISPs in at the Oregon Internet 
> Exchange here in Eugene. 
> 
> On a completely different note.... do you have MRTG-style profiles for 
> typical product feeds? I'm seeing a fair degree of variability in whitestar's 
> network load, and I want to make sure that this is normal for the  
products 
> I've got running for testing, that I'm getting everything I should be 
> getting, and that the pattern I'm seeing is not a sign of a problem that 
> needs to be debugged.
> 
> Whitestar's MRTG pages are at 
> http://goop.uoregon.edu/mrtg/joe/128.223.220.18/128.223.220.18.3.html
> (if you can't get at that one due to .htaccess issues, I've got a snap 
> of the relevant gif over on http://darkwing.uoregon.edu/~joe/whitestar.gif)
> 
> This is for an ldmd.conf request line of:
> 
> request WMO ".*" thelma.ucar.edu
> 
> Are comparable MRTG profiles available for the other UNIDATA products? 
> 
> Thanks, 
> 
> Joe
>