Skip to main content

Cloud-based Monitoring Service for Software Defined Networks

Periodic Reporting for period 1 - SDN-Polygraph (Cloud-based Monitoring Service for Software Defined Networks)

Reporting period: 2016-05-01 to 2017-04-30

Enterprises increasingly depend on computer networks and access to the Internet. Costs associated to network malfunctions are huge: they have been estimated via market surveys from around 30,000€/hour on average, to 4,000€/minute in data centers.

The first requirement to avoid malfunctions and correctly operate a network is network visibility, that is, the ability for network engineers to learn what is happening in a network: what applications are being used, identifying the root cause behind performance issues, and detecting attacks.

However, current network visibility solutions are not widespread enough, because of:

1 - Deployment complexity. Deploying a monitoring product requires expert technicians and entails planned interventions and potential downtime

2 - Excessive upfront costs. Current products are overly expensive, based on appliances or specialized hardware – and therefore upfront costs are hard to justify with a dubious return over the investment.

On the other hand, a technical revolution called Software-Defined Networking (SDN) is currently undergoing. SDN allows software components to alter the network and make it “programmable”.

Leveraging SDN, this project aims to create a zero-hardware network management solution that not only can be deployed with one click, seamlessly collect network traffic statistics for network visibility, and send them to the (our) cloud for analysis, but can also act on the network to block attacks or throttle applications that consume excessive bandwidth, impacting the quality of experience for other users.
We are following an iterative approach to developing SDN-Polygraph and preparing it for massive-scale commercialization. The starting point for our project is a minimum viable product (MVP), which is actively being commercialized in the hunt for early adopters.

So far, we have been able to create a solid technological foundation by combining big data and stream processing technologies with machine learning algorithms. This technical foundation powers SDN-Polygraph to ingest massive amounts of network metadata from the cloud. The main advantages of this approach are (1) it enables zero-hardware network visibility, and (2) it extracts rich insights from network activity, thus achieving a product that solves the problems of deployment complexity and excessive upfront costs of incumbent solutions.
Current network visibility solutions are not widespread because they are excessively hard to deploy and expensive, which creates uncertainty about their financial viability. This is because they are based on specialized hardware and/or appliances that are meant to be deployed within the network, gathering a copy of all communications, which is computationally expensive. Additionally, the network visibility industry has missed the transition to the cloud and the software-as-a-service (SaaS) model that other industries have undergone long ago.

With SDN-Polygraph, network visibility no longer requires deploying any hardware nor appliances (physical nor virtual). Instead, network activity metadata is collected from the cloud, thus simplifying deployment to the extreme. Additionally, this is a solution that is offered as a service from the cloud (SaaS), which removes upfront costs. As a consequence, companies can measure return over investment without commitment, and by de-risking investments in network visibility, more companies can now deploy it. Since nowadays companies increasingly depend on networks, we envision SDN-Polygraph will contribute to a more efficient productive fabric.

In the following months, we will focus on leveraging SDN technology to not only gather visibility, but also to act on the network, with the end-goal of fully removing the need for deploying hardware appliances in the network for both visibility and control.
Main view of TALAIA tool