Data centers and connection

Last change on 2022-10-28 • Created on 2020-03-25 • ID: GE-81F59

General

With "data center" we mean the building itself, the space inside the building where we house the servers of one or several companies or organizations, and the organization itself who takes care of the servers. The common abbreviation is DC.

You can find more information on our website:
https://www.hetzner.com/unternehmen/rechenzentrum/

Which data centers does Hetzner operate?

Data Center Park NBG1
Location Data Center Park Nuremberg, Germany
Number of data center 5
Data Center Park FSN1
Location Data Center Park Falkenstein, Germany
Number of data center 18
Data Center Park HEL1
Location Data Center Park Helsinki, Finland
Number of data center 6

Facilities in each of our data centers:

  • Uninterruptible power supplies (N+1 redundant UPS)
  • 2.5 MVA diesel generator
  • Power supply via two separate power paths from the substation to the low-voltage distribution

Which backbones are the data centers connected to?

See: https://www.hetzner.com/unternehmen/rechenzentrum/

How are the data centers connected to each other?

All Hetzner data center parks are connected to our backbone via redundant dark fiber connections. This ensures the availability of a data center if one of the connections fails. The n*100 Gbit/s connections provide ample bandwidth between the data centers.

The bandwidth of the connections between Nuremberg-Frankfurt, Nuremberg-Falkenstein and Falkenstein-Frankfurt are at least 120 Gbit/s. Our Frankfurt location transports data to our peering partners at DE-CIX and also to the following uplinks: Noris, GLBX, Aixit, AMS-IX, Init7 and Level3. At the Nuremberg location, there are connections to Noris, KPN, Init7, Level3 and N-IX.

In each data center, we operate several Juniper EX Core switches and bundle the streams of the data center to the backbone and then over the various uplinks.

In which data center is my server located?

Make a trace to your server's IP address. On Windows, the command is:

tracert <your\_IP>

On Linux, the command is slightly different:

traceroute <your\_IP>

The initials in the name of the hops in front of your server shows you the data center:

2   217.0.117.200 (217.0.117.200)			18.149 ms  17.080 ms  16.750 ms
3   87.190.176.130 (87.190.176.130)			17.602 ms  17.786 ms  17.778 ms
4   217.239.47.14 (217.239.47.14)			21.653 ms  21.880 ms  22.086 ms
5   ae8-0.fra20.core-backbone.com (62.157.251.158)	31.661 ms  22.087 ms  22.106 ms
6   ae1-2014.nbg40.core-backbone.com (81.95.15.206)	26.164 ms  23.759 ms  24.207 ms
7   core-backbone-100g-nbg.hetzner.de (81.95.15.6)	25.032 ms  25.771 ms  25.485 ms
8   core11.nbg1.hetzner.com (213.239.229.161)		26.463 ms  26.234 ms  26.000 ms
9   core22.fsn1.hetzner.com (213.239.245.213)		27.311 ms  25.535 ms  25.825 ms
10  ex9k2.**dc3.fsn1**.hetzner.com (213.239.229.246)	26.027 ms  26.072 ms  26.317 ms
11  eigenerServer (vvv.xxx.yyy.zzz)			26.959 ms  26.807 ms  26.958 ms

In this example, you can see that the server is located in FSN1-DC3. If the second-to-last hop shows a name in the form of xxx-yyy-zzz-xxx.clients.your-server.de, then a correct rDNS name has not (yet) been assigned to the corresponding router.

Peering Policy

Our primary goal is to exchange traffic, as far as possible, at central exchange nodes, such as DE-CIX, AMS-IX, DATA-IX and V-IX. In addition, we follow an open peering policy, so that everyone who wants to can organize a peering with us there. We are also open for the private interconnection of routers, should it make sense from a technical or economic point of view.

For all targets that cannot be reached via a private peering or via a peering point, we route traffic through 2 of the leading transit providers. If another route is desired, we are always open to setting up a direct peering.

Table of Contents