Telemetric network monitoring is a process by which network data points are sampled in real-time, collected, and then analysed for trends and patterns - that will expose faults and anomalies. It is a very crucial part of the broadcast process, more so, because broadcast operators rely on network architectures more than ever. Monitoring of the network is an absolute vital component to understand how the operation is functioning as a whole, which helps determine and rectify glitches in time.
“Traditional monitoring can be commonly be found probing dozens of specialised broadcast processors with little and sometimes no root cause correlation between alarming devices. With products such as Lawo’s SMART leveraging telemetry, analysis can be performed across multiple processing planes in parallel, namely the broadcast processors and the network they traverse. Long term historical data retention allows for precision trending and therefore exposing root cause correlation much faster and which segment of the operation these faults originate,” explained Tony Zare, senior director, Product Management, Lawo.
By leveraging software defined architectures, the technology can be effectively applied physically and virtually, in turn, allowing deeper penetration into the media network. It also ensures heightened accessibility, irrespective of the location and time. Costs and executions are constantly being refined in favour of the end user. Telemetry is evolving to a larger number of data points, more retention and deeper analysis, which will ultimately result in faster analytics to enable better and faster decision processing. By leveraging COTS compute platform, the need for expensive specialised hardware processing is rapidly becoming less of a requirement. “Consumers will not tolerate disruption in the user experience, especially in the event of an outage, advanced telemetry and monitoring will provide the operator with the right data needed for faster corrective action, therefore, minimising downtime,” added Zare.
Wide adoption of an advanced telemetry and monitoring solution is always a challenge to overcome, according to Zare. The user will usually second guess themselves on the investment as there is a cost component to it. However in the midst of a crisis, having the right toolset can mean the difference between customer retention and loss. He explained this with an analogy, “In some aspects, there is a parallel between the technology and the seatbelts and airbags of a car. We don’t expect to use these safety measures, but are grateful they are in place when they are really needed. Awareness and education is a necessary responsibility of our firm to overcome such challenges both on the potential pitfalls of an IP broadcast but also the solutions that are available to mitigate such scenarios.”
There have been several key instances, in which the technology has been used to expose issues buried deep in the facility. PTP is a common bottleneck that any modern day IP broadcast facility faces. The technology discovers the PTP network, analyses the PTP distribution for compliance, and immediately exposes the weaknesses or faults within the PTP network. Another example where telemetry comes in handy, is for packet compliance from transmitters. These have been determined to fall in a pass/fail model very quickly. Network monitoring helps this by determining which devices are compliant with the governing standards. It also gives an end-to-end network visibility including routing states, bandwidth consumed, and connectivity between network-attached appliances. “The technology immediately exposes where a fault in the chain occurs, and provides a clear demarcation for corrective action between network engineering and media engineering,” said Zare.
The future will demand quicker consequences with utmost precision. This means, it will require more data points of telemetry and will rely on more computational models to process the harvested data. The evolution of these models will naturally integrate Machine Learning (ML) and Artificial Intelligence (AI), in order to reach the ultimate solution; predictive analytics.