back to all posts
When PBXs first hit the market, manufacturers told companies to buy one trunk for every four phones. Obviously too simplistic to apply across all vertical markets and companies, so the age of traffic studies was born. It was all about busy hours and trying to pick the right timeframe to run the studies. Since those days, the introduction of technologies such as cellphones, texting, VoIP, SIP and web collaboration have rewritten the world of network performance management. Today, you have the option to be proactive and make true engineering decisions based on your company’s unique collaboration network. No longer do you have to sit back and wait for periodic traffic studies or let your telecom provider tell you how much capacity you need. Empower yourself and your organization with the right methodologies and tools! This blog is not meant to have all the answers, but hopefully it will get you moving and thinking to transform your business processes for network performance monitoring and capacity planning.
Trend analysis and data decomposition can expose the root cause of current and impending service issues along with providing the basis for accurate capacity planning and network design. Usage & Trend Analysis
Start with a high-level picture to capture cyclical patterns in calls and usage carried over all your network facilities. View trends broken out by incoming, outgoing and total traffic as the Year over Year monthly example below highlights. You can quickly understand the months that drive higher or lower traffic volumes. Viewing trends by enterprise and drilling down to individual voice servers or trunk groups or routes along with weekly and daily trends will be insightful too.
Data Decomposition
But there is so much more than just call counts and usage hours to consider! Data decomposition depicts how well your network resources are supporting desired call activity versus how much capacity is being consumed servicing failed or fraudulent activity. Views like example below highlight these trends and percentages. A Normal Call Resolution indicates the call completed as expected with one of the parties hanging up. All other conditions may indicate some type of failure in the call process. In this example, over 20% of monthly call capacity in July and August was some type of Equipment Failures. And, dropped calls started to spike in August. User can quickly drill into detail to troubleshoot. Ideally, if you are proactively monitoring, these types of failed calls would not continue for months.
Capacity Planning Peak Percent Utilization This high-level view has value, but additional insight is gained by drilling into detail. You will want to be able to see the utilization by hour trending. Quickly note any trends or aberrations in activity that may account for elevated peak utilization. Or confirm if your current capacity is too much or not enough. In the example below, the peak capacity consistently remained around 50-60% with the exception of the 100% spike on December 6. Was that spike driven by the business (a marketing campaign or service outage to generate calls) or maybe an inbound hacking attempt? If not “real”, this could indicate a potential opportunity for savings by decreasing channel capacity. Concurrent Call Paths
The second example is much less consistent. In many cases, it is inbound hacking call attempts with hundreds or thousands of very short calls that drive the spikes you may see in CCP like the August 16 spike below. Again, you would want to drill down to the details to determine what is behind the abnormal behavior. Networks are dynamic. To transform your network management and create optimum resourcing designs and strategies, you need visibility into your current and historical usage, performance and capacity metrics. Check back again soon for the next post on troubleshooting some of those abnormal call spikes.
Clearly See Your Voice Network Activity and Performance with Optic
Related Posts
|