Seven QoS Best Practices for Monitoring Cloud Traffic

The recent maturity of technologies such as high-performance multi-core CPUs, virtualization, and fast, ubiquitous Internet access has accelerated the adoption of cloud computing services. In fact, according to the latest research, 50% of software applications being developed today are cloud-based. This continued shift to cloud computing is affecting enterprises' network infrastructure as they rely more on cloud computing systems for critical applications. Specifically, when network traffic transitions to a cloud consumption model, the characteristics of network traffic change. And how these changes affect the quality of the user experience (or quality of service) should be a top priority for the networking team. In this article, I want to take a quick look at the different types of cloud services that are being adopted and provide some Quality of Service (QoS) best practices that network operations teams should consider.



The industry focuses on three main cloud service models: software as a service (SaaS), platform as a service (PaaS), and infrastructure as a service (IaaS). SaaS eliminates the need for end users to purchase and manage computing infrastructure and software. PaaS eliminates the need to purchase and manage computing infrastructure and licenses. It also manages additional platform services such as database and billing. In addition, PaaS can lock end users into a specific platform environment. Whereas IaaS eliminates the need to purchase and manage computing infrastructure and provides the flexibility to move to other platforms.

Most cloud services are delivered over the Internet using standard Internet client applications, and Internet access gives users the flexibility to use these services almost anywhere. If QoS policies are not adjusted, standard Internet protocols, such as HTTP and HTTPS, are treated as normal web browsing. Whether the expected load on the network will have an impact on end users is best determined based on the current network design and the services running on the network. However, any standard QoS policy developed by NetOps would treat these applications as leisure network traffic and assign it to the best effort class. To ensure you have an appropriate QoS policy, here are seven key steps to consider.

1) Communicate with other IT teams. In a large enterprise, there may be multiple teams involved in launching a cloud service. For example, if a NetOps team is putting unified communications in the cloud, they may need to work with a different group than implementing Google Apps for Business. The best way is to get the team involved in the discussion about the release as early as possible, rather than solving problems after the release. It is helpful to ask the following questions: how to best identify the traffic used for prioritization, and whether there are any known issues with applications being prone to loss, latency, or jitter.


2) Understand traffic. Network operations teams should have a good understanding of their current conditions by performing a baseline analysis of network traffic. This is because cloud services will cause a shift in traffic and usage patterns by increasing traffic to and from cloud service providers. For example, what used to be strictly internal office and backup traffic may now greatly increase the bandwidth demands on the network perimeter, requiring circuit upgrades. Teams can determine network traffic utilization by assessing specific application consumption using the Netflow reporting solution.

3) Develop an initial design. It is important to design how cloud services fit into current architectures and corresponding QoS considerations (and policies). Like traditional WAN environments, cloud service traffic is also best suited for a QoS model with four or more classes. At a minimum, real-time voice/video, transactional data, bulk data, and best-effort data traffic should be separated into distinct categories for bandwidth management. In some networks, more further subdivisions may be required. If the baseline analysis indicates a need for more bandwidth or changes to the network structure model, this should be included in the plan.



4) Review the plan. QoS policies should be presented to other IT teams and key stakeholders within the organization. They will want to know about the changes and how those changes might impact current services. If the existing network architecture needs to be updated (upgraded WAN services, new routers, or even a different network design), it also needs to be reviewed.

5) Design details. Most cloud service traffic will use existing internet protocols. Be sure to know the IP addresses of the services you're connecting to in order to distinguish them from random internet traffic. Utilizing a NetFlow reporting solution can greatly assist this process. Some NetFlow solutions can also help identify traffic beyond a basic list of IP addresses.