Gamerail Stress Testing > Gamerail BLOWS

reply to twizlar

Re: Gamerail BLOWS

Tracing route to 208.75.25.70 over a maximum of 30 hops

  1    <1 ms    <1 ms    <1 ms  192.168.0.1
  2    17 ms     5 ms     5 ms  10.150.0.1
  3     8 ms     7 ms     8 ms  border.mountaincable.net [24.215.3.129]
  4    21 ms     8 ms    10 ms  ra2sh-ge3-4-14.mt.bigpipeinc.com [66.244.223.237]
  5     9 ms     7 ms     7 ms  rc1sh-ge9-0-6.mt.shawcable.net [66.163.66.49]
  6    20 ms    21 ms    19 ms  rc1hu-pos5-0-0.ny.shawcable.net [66.163.77.154]
  7    24 ms    22 ms    22 ms  64.243.225.142
  8     *        *        *     Request timed out.
  9    42 ms    47 ms    41 ms  208.75.25.70

Trace complete.

Tracing route to 64.34.240.161 over a maximum of 30 hops

  1    <1 ms    <1 ms    <1 ms  192.168.0.1
  2     9 ms     6 ms     6 ms  10.150.0.1
  3     9 ms     8 ms     9 ms  border.mountaincable.net [24.215.3.129]
  4    10 ms     9 ms     6 ms  VLN.Serialx-x-x.GWx.POPx.ALTER.NET [63.115.224.73]
  5    11 ms     7 ms     9 ms  0.so-6-0-2.XT2.TOR2.ALTER.NET [152.63.133.78]
  6    27 ms    27 ms    26 ms  0.so-6-0-0.XL2.CHI13.ALTER.NET [152.63.70.105]
  7    23 ms    27 ms    23 ms  POS7-0.GW1.CHI13.ALTER.NET [152.63.69.181]
  8    24 ms    30 ms    23 ms  bandwidthoptions.net-gw.customer.alter.net [157.130.173.238]
  9    23 ms    37 ms    23 ms  peer1-chi01-ge2-core1-equinix.ord.gowebman.com [69.90.119.10]
 10    24 ms    22 ms    24 ms  209.8.108.22
 11    21 ms    24 ms    23 ms  64.34.240.161

Trace complete.

I know however they are peered directly with the same upstream providers in dallas.
--
AMD Athlon64 4000+ @ 2723mhz - mountaincable.net wireless Intarweb |Ipods SUCK

I don't think you are understanding me. I'm not using the GameRail client. Those are servers located exclusively on their network which peers directly with the same providers in dallas and yet their routing is completely messed up. The packets should not be routed from dallas to LA to Denver and BACK to dallas when gamerail is peered directly to them. This has nothing to do with the client side and everything to do with their backbone routing and peering arrangements. I used traceroutes from here as an example. IF they were setup properly I should not have a higher ping and more hops to a server located on their network as opposed to one that isn't the packets SHOULD take the best route
--
AMD Athlon64 4000+ @ 2723mhz - mountaincable.net wireless Intarweb |Ipods SUCK

i'm not talking about the client side of gamerail, only their network and routing. ANY provider that doubles or triples latency is garbage, regardless of if they are using the client or not. There is no reason for pings to be higher to a server in the SAME building with the same available upstream providers. You are missing the entire point, and the possible future implications for server providers.
--
AMD Athlon64 4000+ @ 2723mhz - mountaincable.net wireless Intarweb |
Ipods SUCK

The peering arrangements should be done from one clear shot am I correct?

You are using two different hops at a peering point with one having more congestion than the other and you are stating that Gamerails network is the one with the higher latency if I am understanding you.

Either way if the routing was longer the pipes are most likely dedicated and used specifically for gaming which still provides lower latency even at longer distances.

While a more significant improvement will be if it where a clear shot, that does not seem to be the case here, but in theory if the pipes are not congested with regular traffic then you will still see some benefit but not as much.

I live in FL, and I am routed all the way to Atlanta, GA where my POP is located. My first hop is 30ms in latency due to that distance.
--
The only limits we have are the one's we set ourselves.

reply to twizlar