Data center infrastructure efficiency is becoming an important component for business success. These days hyper-scale operators (biggest players of on-line business hosting hundreds of thousands of servers) invest in self-developed solutions by leveraging on combination of open equipment and open software. Such leverage requires huge efforts and may sound infeasible and focus-loosing for majority of medium to large companies.
Our out-of-the-box automated solution provides a single intuitive GUI to manage on-premise networks of any size effortlessly with efficiency and technology leverage previously available to hyper-scale operators only.
Reducing CapEx/OpEx up to 10X. Saving time up to 100X.
Our 24/7/365 TAC is available by phone/e-mail/messenger/ticketing and all inquiries are directly addressed to experienced specialists acting as a single point of contact for the whole system.
We guide our customers through the whole open networking transformation journey:
XCloud Networks’ platform provides functionality for engineering team to easily manage security policies taking care about possible approval procedures, logging requirements and change history. XCloud agent running on the hardware side spins configuration in accordance to approved ACL entries or object oriented service groups which engineers define from the GUI.
Stateless access control enforcement in combination with easy-to-use GUI in many cases eliminates need of expensive firewall boxes and it is completely decentralized.
By attaching our “TCAM off-loader” to the switch which is a good example of combining hardware and software it becomes possible to handle full routing table and do NAT. Adding BGP peers by few clicks in the GUI. 40Gbps towards high prefix neighbors + 720Gbps towards low number prefix neighbors.
48 X SFP+ ports under 20K$.
Zero extra cost hardware load balancer is scalable with no throughput constrains, basically switch forwards packets in load balancing mode at the full line-rate. This most efficient load balancer hashes source/destination ip/protocol/port to ensure integrity of transport layer such as TCP.
Need an application layer load balancer or SSL offloading then HAProxy integration helps. Connect some servers to the network fabric, running HAProxy+Linux+XCloud agent, then it becomes part of whole system including monitoring, analytics, logging and management is fully available from the GUI handling all the necessary details automatically.
Our integrated fault prediction and detection system constantly follow telemetry information from every hardware unit.
In case of maintenance has to be done, device can be safely powered off after enabling maintenance mode which tells every neighboring device to re-route the traffic.
In the event of switch replacement it should not necessarily be replaced by the same model or even same vendor, just any of validated switches with amount and types of ports you need, then network operating system can be installed automatically and configuration of previously removed switch can be automatically restored.
VPN & Encryption
Seamless integration with OpenVPN enables easy-to-configure elegant site-to-site or remote access VPN without need to buy expensive hardware.
By simply attaching Linux server with OpenVPN and XCloud agent to the fabric the whole system becomes capable of automated VPN and encryption setup.
You simply enable site-to-site and/or remote-access VPN by push of a button then configuration is rolled out automatically including setup and future updates of OpenVPN endpoints, dynamic routing between sites, redundancy, encryption keys and monitoring.
Seamless integration with open source platforms including Proxmox and Openstack helps to apply same network principles and to use one single GUI to control the whole network including virtual network of the KVM based server virtualization.Fo
L3 LEAF/SPINE DESIGN
L3 leaf/spine topology solves problem of west-east traffic. Fabric allows utilization of all links, avoids usage of legacy protocols like spanning tree, with almost no limits in scalability.
Servers can be connected to any number of switches (not just 2) fully utilizing every connected port and in this case, there is no stacking or any kind of state replication between switches so top of the rack switches is acting totally independent which saves from failures commonly happening in the software part of stack-designed models.
Switches of diverse modes, vendor and technology can be used in this design.
CDN & Distributed Networks
Deploying CDN basically means running a distributed network spread across the world. Our GUI helps to structure and see all statistics, monitor the health and proactively prevent failures as well as to manage a big number of remote devices on the scale and cost-efficiently.
You can download the solution brief for Data Center Networks here.
If you need more information please contact us.