I haven't had the time to do that research yet. I made available the spreadsheet if someone /does/ have that time.
Just as an opener, I /do/ know that 12.xx belongs to ATT, 65.xx, 68.xx and 69.xx are broken into multiple class B's owned by various datacenter groups (Atlanta's 12 marrieta owns multiple of the 68.xx and PEER1 owns multiple 65 and 69.xx class b's -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Ooks Server Sent: Monday, September 20, 2004 11:28 AM To: [EMAIL PROTECTED] Subject: Re: Re[2]: [hlds] RE: ServerList Destroyed::::Please take time to read What is the nature of the octest that get the traffic? Are they servers hosted in datacenters? And the ones that don't get traffic, are they IPs normally used to serve home consumers? I would like to know more about what the octets represent - location, ISP, usage, etc. ----- Original Message ----- From: "Shane Robinett" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Sunday, September 19, 2004 10:04 PM Subject: RE: Re[2]: [hlds] RE: ServerList Destroyed::::Please take time to read > Well, instead of just whining about it with no conclusive proof of > anything, I demonstrate how apparently I have nothing to do at > midnight.. > > This excel spread sheet found at http://www.phxx.net/dl/analysis.xls > will demonstrate what, at first glance (that's all I've had), there is > some definite outliers or preferential IP octet (first) that score the > bulk of play. At the same time, some IP's simply don't appear to get > /any/ traffic. > > Anything in baby blue is a EMPTY ip range. 41.xxx, > 62.xxx,81.xxx,82.xxx,83.xxx,213.xxx,217.xxx for example are horrible > empty. With 3200+ servers (using default 27015 port), they have a 1% > fill rate! > > Yet, ranges 41.xxx, 61.xxx, 128.xxx, 210.xx x,211.xxx with 200 servers > between these octets have upwards to 50% fill rate! > > Even my own 300 servers have a fill rate of 17-24% based on our various > octets (I guess we are lucky). > > Anyone is willing to do more review of this data. - It's not at all > inclusive in it's research ( I haven't traced geo data, etc) but it does > give us some numbers to work with. > > > > > > Shane > http://www.PHXX.NET - CIO > http://www.simple-share.net share something > http://www.w00t.net - server/player/map/sex search > > > > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Shane Robinett > Sent: Sunday, September 19, 2004 7:24 PM > To: [EMAIL PROTECTED] > Subject: RE: Re[2]: [hlds] RE: ServerList Destroyed::::Please take time > to read > > <chuckle> I think we are going off topic, my fault - I apologize. My > point wasn't so much redundancy as it was peering and level of peering > with other providers. > > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Jim Norton > Sent: Sunday, September 19, 2004 7:04 PM > To: [EMAIL PROTECTED] > Subject: Re: Re[2]: [hlds] RE: ServerList Destroyed::::Please take time > to read > > On Sun, Sep 19, 2004 at 06:46:58PM -0400, Shane Robinett wrote: > > I doubt that your single-hone bandwidth provider for your DSL provides > > the same quality of pipe people get at some quality carrier hotels and > > datacenters. > > Well the "last mile" of my connectivity has a single point of failure, > but > the reality is that in the past 2 years, I've not had any network > outages. > > My DSL is provided by Megapath Networks. One of the best quality DSL > providers > around. > > I'd love to have triple redundant OC48 connections, but I don't have > that kind > of money, and I am not enamored with co-location. > > > > > However. > > > > It sounds like in your one instance you have a quality box and you > > probably manage your server to the point it performs very well for the > > people that do get good pings to your DSL provider. > > > > In argument for the people who buy/manage servers off data center > > bandwidth, I think I can prove that 70%+ of all people /playing/ on a > > server are playing on datacenter bandwidth. This is NOT the same > > statement as most servers are run off of datacenter bandwidth (I think > > the opposite is true, I'll try to prove the point. - maybe that will > > convince Valve to change their listings alg.) > > Are you saying you'd have Valve bias the serverlist in favor of > datacenter > run servers? > > -- > Jim Norton - http://www.jimnorton.org > > "The art of listening is indispensable for the right use of the mind. > It is also the most gracious, the most open and the most generous of > human habits." > (Attributed to R. Barr, St. John's College, Annapolis, MD) > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds > _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
