Web Hosting Talk : Web Hosting Main Forums : Dedicated Server : 100tb terremark POP

[elvis1]

According to http://www.100tb.com/why-10tb/softlayer.php they will someday have a pop at the NAP of the Americas, right?

Well my question is: what benefits would this bring? Having a pop does not mean they will actually have servers in place, right?

In any case, would this help lower the latency of servers ( to latam) ?

Thx

[woods01]

Good luck with anything terremark related.

I can't name a company of the same size as them that's running on auto-pilot like they are.

Your question is going to be really hard to answer by anyone.

We host in LA with level3, cogent, savvis, and also any2 (crg) and the routes really do vary.

Sometimes the routes will go through crg however sometimes they would just go through level3.

That makes a big difference since level3 might redirect back to los angeles to connect to savvis to complete a route where-as crg would probably already have a connection to the endpoint.

We have no say in how the route is going to go.

Looking back at Softlayer they had a history of manipulating their routing so if they still do manipulate their routes perhaps you could benefit in some situations since they won't be using BGP.

[IGobyTerry]

It's a POP that's already active, as part of the SoftLayer network that the 100TB USA offerings use. Miami typically does provide for great access times to Latin America.

Keep in mind that a POP is not a datacenter.

[cresci]

Yes, their pop there is just por peering/transit exchange, not for colocation/having servers.

[elvis1]

so in which way can I make use of having a server with 100tb and them having a POP in terremark to have low latencies to latam?

thx

[bizness]

It really depends on what you are trying to accomplish. What cliental are you going after? ... etc ...

we have a single cabinet at terremark which we use for peering while our 1.0 datacenter is a few blocks south of terremark and our upcoming 2.0 is 40 miles north of terremark. we inter-connect between the facilities using DWDM over dark-fiber we rent. We took this route due to the extremely high cost of terremark prices.

Although extremely expensive, terremark is a great place to inter-connect and peer.

[elvis1]

Originally Posted by bizness

It really depends on what you are trying to accomplish. What cliental are you going after? ... etc ...

we have a single cabinet at terremark which we use for peering while our 1.0 datacenter is a few blocks south of terremark and our upcoming 2.0 is 40 miles north of terremark. we inter-connect between the facilities using DWDM over dark-fiber we rent. We took this route due to the extremely high cost of terremark prices.

Although extremely expensive, terremark is a great place to inter-connect and peer.

I am seeking to provide a low latency service to clients in Argentina ( just below Brazil). Most clients would ask for a shared hosting environment with static content .

Could you please provide something I could ping, something I could test the load speed, do some traceroutes and something to download please?

Thanks in advance

[dotHostel]

Originally Posted by inogenius

Miami typically does provide for great access times to Latin America.

Unfortunately this is not 100% true as there is just one Internet exchange point in Miami(http://www.datacentermap.com/ixp/tmrk-mia.html). If your backbone provider and the local data center's backbone provider don't peer at Terremark, then the traffic will go to another private or public US IXP before returning to Miami via a data center's backone provider.

As a side note, in the past I used for a long time Atlantic.net (Orlando,FL), a very good data center. Take a look at the traceroute from Sao Paulo (Speedy, Telefonica):

Tracing route to www.atlantic.net [209.208.9.38]
over a maximum of 30 hops:

1 3 ms 1 ms 1 ms 192.168.1.1
2 * * * Request timed out.
3 65 ms 34 ms 35 ms 201-0-93-9.dsl.telesp.net.br [201.0.93.9]
4 36 ms 32 ms 32 ms 200-100-3-229.dsl.telesp.net.br [200.100.3.229]
5 73 ms 31 ms 31 ms 200-100-98-185.dial-up.telesp.net.br [200.100.98.185]
6 33 ms 32 ms 40 ms xe0-0-6-0-grtsanem1.red.telefonica-wholesale.net[213.140.51.17]
7 144 ms 158 ms 139 ms xe2-0-3-0-grtmiabr3.red.telefonica-wholesale.net[94.142.123.254]
8 173 ms 183 ms 170 ms xe7-1-0-0-grtwaseq4.red.telefonica-wholesale.net[94.142.122.157]
Washington peering point Telefonica <-> Level3
9 183 ms 209 ms 200 ms xe-11-2-0.edge3.washington4.level3.net [4.53.112.17]
10 194 ms 196 ms 199 ms vlan69.csw1.washington1.level3.net [4.68.17.62]
11 186 ms 187 ms 178 ms ae-61-61.ebr1.washington1.level3.net [4.69.134.129]
12 163 ms 180 ms 175 ms ae-8-8.ebr1.washington12.level3.net [4.69.143.218]
13 183 ms 194 ms 183 ms ae-6-6.ebr1.atlanta2.level3.net [4.69.148.105]
14 196 ms 193 ms 200 ms ae-1-8.bar1.orlando1.level3.net [4.69.137.149]
15 181 ms 179 ms 195 ms ae-6-6.car1.orlando1.level3.net [4.69.133.77]
16 196 ms 224 ms 198 ms atlanticnet.car1.orlando1.level3.net [63.209.98.66]
17 194 ms 193 ms 194 ms andc-abr-2-g3-13.atlantic.net [209.208.6.54]
18 198 ms 198 ms 188 ms andc-abr-2-g3-13.atlantic.net [209.208.6.54]
19 187 ms 166 ms 184 ms www.atlantic.net [209.208.9.38]

Trace complete.

[Dougy]

I saw traces from down in that area come up to NY then go down to miami.

[bizness]

Originally Posted by elvis1

I am seeking to provide a low latency service to clients in Argentina ( just below Brazil). Most clients would ask for a shared hosting environment with static content .

Could you please provide something I could ping, something I could test the load speed, do some traceroutes and something to download please?

Thanks in advance

sure, you can try the following two ips.

208.68.235.1
208.68.233.1

[funkywizard]

Having a pop there may help the overall network, which is better than not having one there, but if you're specifically targeting to host people in south america, you'll want your servers there as well, or you won't get the same benefits.

[elvis1]

funkywizard : thanks man!
bizness: I enclose my traceroute. Both IPs are the same.

C:\Users\elvis>tracert 208.68.235.1

Traza a la dirección host-208-68-235-1.biznesshosting.net [208.68.235.1]
sobre un máximo de 30 saltos:

1 <1 ms 1 ms <1 ms 192.168.1.1
2 1 ms 1 ms 1 ms 10.0.0.2
3 22 ms 18 ms 16 ms host14.200-3-60.telecom.net.ar [200.3.60.14]
4 153 ms 154 ms 156 ms host82.200-3-81.telecom.net.ar [200.3.81.82]
5 151 ms 153 ms 155 ms te7-3.ccr01.mia03.atlas.cogentco.com [154.54.10.
101]
6 157 ms 159 ms 159 ms te7-2.ccr01.mia01.atlas.cogentco.com [154.54.3.1
25]
7 159 ms 161 ms 162 ms host-208-68-235-1.biznesshosting.net [208.68.235
.1]

Traza completa.

[dotHostel]

Tracing route to volico.com [208.68.233.200]
over a maximum of 30 hops:

1 2 ms 7 ms 2 ms 192.168.1.1
2 * * * Request timed out.
3 33 ms * 31 ms 201-0-93-9.dsl.telesp.net.br [201.0.93.9]
4 30 ms 33 ms 31 ms 200-100-3-233.dsl.telesp.net.br [200.100.3.233]

5 31 ms 31 ms 31 ms 187-100-58-69.dsl.telesp.net.br [187.100.58.69]

6 33 ms 53 ms 34 ms 213.140.51.33
7 145 ms 146 ms 147 ms Xe2-0-3-0-grtmiabr3.red.telefonica-wholesale.net
.123.142.94.in-addr.arpa [94.142.123.254]
8 190 ms 189 ms 179 ms Xe-2-1-0-0-grtwaseq4.red.telefonica-wholesale.ne
t [94.142.124.117]
9 180 ms 190 ms 179 ms xe-11-2-0.edge3.Washington4.Level3.net [4.53.112
.17]
10 183 ms 190 ms 190 ms vlan99.csw4.Washington1.Level3.net [4.68.17.254]

11 178 ms 176 ms 186 ms ae-81-81.ebr1.Washington1.Level3.net [4.69.134.1
37]
12 188 ms 192 ms 181 ms ae-8-8.ebr1.Washington12.Level3.net [4.69.143.21
8]
13 201 ms 197 ms 197 ms ae-6-6.ebr1.Atlanta2.Level3.net [4.69.148.105]
14 194 ms 179 ms 193 ms ae-72-70.ebr2.Atlanta2.Level3.net [4.69.138.19]

15 202 ms 180 ms 179 ms ae-24-52.car4.Miami1.Level3.net [4.69.138.101]
16 170 ms 215 ms 180 ms ae-24-52.car4.Miami1.Level3.net [4.69.138.101]
17 177 ms 183 ms 169 ms host-208-88-120-74.biznesshosting.net [208.88.12
0.74]
18 178 ms 168 ms 168 ms host-208-88-120-74.biznesshosting.net [208.88.12
0.74]
19 200 ms 177 ms 178 ms host-208-68-233-200.biznesshosting.net [208.68.2
33.200]

Trace complete.

[elvis1]

We ve discussed over and over that brazil and Miami are not the best of the best. Cresci said it landed in NYC ( with one of your providers IIRC).

Could you please test :

enotchnetworks.com

regards

J

[funkywizard]

Originally Posted by elvis1

We ve discussed over and over that brazil and Miami are not the best of the best. Cresci said it landed in NYC ( with one of your providers IIRC).

Could you please test :

enotchnetworks.com

regards

J

Miami is physically a lot closer to Brazil than NYC, so if the routing is done well, then you will get much better performance out of Miami.

Sure, you can get carriers that make bad routing decisions, and send the traffic to NYC before sending it to Miami where your servers are. Similarly, you could have servers in NYC, and the traffic goes to Miami before coming to your servers in NYC.

In the end, it all depends on the carriers. Get the right carriers, making the right routing decisions, and Miami is the best place to host this kind of thing. Get carriers making less optimal routing decisions, and all bets are off as to the best place to host this stuff.