Neutral Internet eXchange
NIX.CZ allows you to connect, interact and share in whole new ways.

Network topology

Topology NIX

Using our extensive experience with the Prague Peering Exchange, we have selected Cisco Nexus 7010 for the launch of the NIX.SK node. With this hardware profile the switches are located in two detached locations and are configured to be fully redundant. The interconnection of the switches is ensured by two independent routes each with double the desired capacity, so that failure of one of the routes does not affect the operation of the connection node.

Topology of NIX.CZ

In 2010, the topology of NIX.CZ Internet Exchange migrated from a “ring” topology to a virtualized “dual star”. This change provides greater stability to the whole peering platform and has also simplified further development. The change in the topology made it possible to carry out this transfer while only affecting one company at a time (i.e., the one being transferred). The change of technology also allowed for continued growth in the volume of transmitted data, as it was expected in the following years. At the same time, the new topology allowed the Association to open the new connection point NIX5, more than 8 years after the opening of the previous one. Because of a high demand for 100GE connections after 2014, NIX.CZ began to implement the Cisco Nexus 7710 access switches. In 2016, NIX.CZ also began running so-called passive points, which have no active technology but make it possible to connect to NIX.CZ using dark fibres.

NIX.CZ is currently using Cisco Nexus 7710 as its backbone switches in fully redundant configuration in the Czech Republic and Cisco Nexus 7710 as access switches in Slovakia. The interconnection technology of NIX.CZ continues to run on the link layer (Ethernet).

In mid 2019, NIX.CZ interconnected its nodes in Prague and Bratislava, enabling easy networking between the Czech Republic and Slovakia. This connection is realized by two independent 100GE circuits that are topologically separated.

Due to a lack of dark fibres, the backbone connections use the passive DWDM technology (Dense Wavelength Division Multiplexing) running over the existing fibres, to allow an increase in the capacity of backbone interconnections to up to 32 channels per pair of optical fibres, without the need for additional dark fibres, however, this technology will be gradually replaced by 100GE circuits.

The Association runs its autonomous 6881 system on a platform of Cisco ASR1001 routers, enabling access to the internal services of the Association and connection of hosted TLD DNS servers, both via the IPv4 and IPv6 protocols. The smooth operation of the infrastructure is ensured by support servers installed at individual locations.

The Association continues to enable its members and customers to be interconnected using the IPv6 protocol. This connection is implemented within the same VLAN as IPv4 a fact which enables members and customers to operate routers and interfaces in a double protocol buffer mode. As of the end of 2011, 88 IPv6 networks have been interconnected representing approximately 80% of all participating networks.


SITELPOP1, Kopčianska 20, Bratislava
1GE, 10GE, 25GE, 40GE, 100GE

SITELPOP2, Kopčianska 18, Bratislava
1GE, 10GE, 25GE, 40GE, 100GE

NIX.SK3 - DC Digitalis
DIGITALIS, Trnavská cesta 110/B, Bratislava
1GE, 10GE, 25GE, 40GE, 100GE

Vienna Interxion, Louis-Häfliger-Gasse 10, 1210 Wien, Austria
1GE, 10GE, 25GE, 100GE


Route Servers

Two route servers at NIX.SK1 and NIX.SK2 simplify the system of peering sessions between members and customers, where there is no need of individual BGP sessions. These route servers also simplify the process of new member’s and customer’s admission to the Association. The service is provided free of charge to members and customers and peering is supported in both IPv4 and IPv6 protocols. Due to the requirements for higher stability, the route servers were divided into two servers with different operating system and both are running at routing daemon Bird. Promoting peer prefixes over route servers can be managed through BGP communities.


IPv6 protocol

IPv6 protocol is a successor to the IPv4 protocol, which is currently widely used on the Internet. Its specifications were developed in the 1990's. One of the reasons for the development of IPv6 was a thread of IPv4 address space exhaustion. Besides the bigger address pool, the new IPv6 protocol comes with support for QoS, implementation of security mechanisms and support for mobile devices or automatic device configuration.

With the IPv6 protocol the bit- length of the address is increased fourfold; instead of the original 2^32 (4294967296), there are now 2^128 (340282366920938463463374607431768211456) addresses available. IPv6 address space is divided into several groups and global addresses are assigned from the 2000::/3 pool. According to the published hypotheses, the IPv4 address space exhaustion was expected in July 2012 but NIX.CZ has allowed IPv6 peerings for its members and customers since 2003. At the moment, 86 of 117 connected networks to NIX.CZ are using this option.

Please note that this page allows you to check whether your computer is connected via IPv6 protocol, or not. It is also possible to use IPv6 protocol parallel to the IPv4 protocol.

32bit AS numbers

Beside the IPv4 address pool exhaustion there was also a need to solve a similar problem for numbers of autonomous systems used for BGP router identification, as autonomous system numbers are being exhausted too. The original pool contained 216 (65536) autonomous system numbers, and by the new specification it was boosted to 232 (4294967296). This enlargement is still compatible with routers, which do not support 32bit autonomous system numbers. The exhaustion of the current 16bit pool was expected in June 2011 but the RIPE registry has been providing 32bit autonomous system numbers for new requests since the beginning of 2009.