Master The Performance Management Data Tsunami

Welcome to October 2012 issue of Creanord EchoNEWS, keeping you informed of developments at Creanord and in the fields of SLA assurance, automated Ethernet service delivery and OAM (Operations, Administration and Maintenance).

EchoVault is the leading Ethernet SLA management platform, helping Service Providers focus on their SLA to drive service provisioning and reporting, to enable differentiation and improve customer satisfaction. This article focuses on managing and exploiting the huge and ever growing onslaught of performance management data from the network.

Locating The Grains of Relevant Data

Ethernet provides the most cost efficient way to provide connectivity solutions such as Mobile Backhaul and Business Ethernet. The requirement for differentiated services and service assurance brings the need to collect data on how well the services are performing. As packet based traffic is bursty by nature, granular data is required to detect for instance micro-bursts and also to be able to display data that is up-to-date. With a setup collecting data at 1 second intervals for instance, the amount of data being collected is 300 times higher than with a typical SNMP collection set at 5 minutes. Assuming 20 KPIs at one second intervals and 1,000 nodes, the amount of performance data coming from the network is 1.2 million data points per minute. The question is, how do you find the relevant data from all this?

Performance management data is used for several different purposes and each of these pose different challenges. The user needs to find their data relevant, sorted and viewable from the appropriate perspective in order for it to be useful. And further, it needs to be up to date - who has time to wait for 5 or more minutes to see what the status is? So it is not only a question of finding the relevant grains of data, you need to go further and present it in the right context and in a timely manner.

Each of the relying processes have differing needs, but what they do have in common is that the more up-to-date the information it is and the more relevant it is, the higher the value of the data and the more efficient your organization will be in completing their tasks. Additionally, with more timely, to-the-point data, the quality of decisions will be better and more well informed.

Surf the Waves of Big Data

In order to be able to collect and analyze enough data from thousands of nodes, the solution needs to be built from the ground up to be able to take in the performance data and do the heavy analytics and SLA compliance calculations. The requirements do not stop there, however. Additionally the data needs to be appealingly and dynamically presented to different user groups, naturally in near real time, even extending visibility securely to customers. This means that there needs to be a service perspective, to ensure that the information is displayed in a way that is relevant to the user.

Figure 1. Finding the relevant data points. If the SLA defines monthly performance targets, and a warning threshold is set, performance management may send events when the target is breached for e.g. 5 or 15 minutes, even if the monthly target is not at risk. Using an SLA-target based monthly trigger will ensure that you find the grain of relevant data from the onslaught of warnings.

Creanord EchoVault provides interactive views to current status as well as scheduled reporting of performance management data with analytics and status against monthly SLA. These views can be customized for different users, internal and external through the secure URL proxy-based portal.

For Fault Management, the EchoVault helps to prioritize, so that the most important circuits nearing SLA breach can be remedied first. For Capacity Management, data about network resource utilization helps to identify bottlenecks and underutilized components. For Account Management it provides input to identify current status and upsell opportunities. For SLA management it provides the link from business targets to service performance as well as unparallelled visibility and transparency by showing the information through a portal to customers. All this is enabled by the powerful underlying engine performing analytics, crunching the numbers and using millions of data points and operations an hour to calculate performance against thresholds while adding SLA relevant information such as maintenance windows and presenting the data in context and in an appealing way.


Let's stay in touch!


Stay tuned with us!