Owr own netowork linking all data centers
Our network connects all locations managed by us, allowing for free communication between clouds and redundancy of connections with the public network.
Thanks to this design, users get a point of contact close to their geographical location, and lower access time to the infrastructure. All connections allow for achieving high bandwidth and are constantly expanded, depending on the demand.
Connection with hundreds of internet service providers
In each of our data centers, we have contacts with hundreds of internet service providers, through internet exchange points (ACX, PLIX, DECX), dedicated points of contact with large providers, points of contact with other clouds as well as links to scientific and academic networks. Connections are constantly modified and adapted to current needs.
Bandwidth that corresponds to the amount of data stored
The public data repository occupies dozens of PB. Our connections must match the amount of data which are transferred to our clouds and results of processing performed by customers every day. Our monitoring system constantly watches over the network capacity and guarantees high-quality access to our clouds.
Full network redundancy
Every element in our network is redundant. There are many routes from each internet service provider to reach our network. Within our infrastructure, we use dynamic routing protocols that ensure the correct and very efficient operation of the network. Our network does not have a single central point, it is spread among many devices and locations.
Access to GEANT
We are directly connected to the GEANT scientific network which covers the whole world. This allows institutions, universities and research centers having a similar connection to optimize costs and improve quality. In our infrastructures, we enable clients to authenticate and confirm their identity using the available mechanisms.
Physical separation of the network
In our data centers, we use a physical separation between the public network, private clouds, private customer networks and the infrastructure management network. Thanks to this, we minimize the attack vectors, increase redundancy and resistance to human errors as well as interference-free operation.
See also: technologies we use
Ceph is a distributed storage system that is scalable to thousands of nodes. Learn how it helps us in our work.