by Simen Frostad, Chairman of Bridge Technologies
Some of the great advances in human life have come about through material innovation – the invention of the wheel, or the discovery of radium – and some through changing the way that we think about the world. The theory of relativity or double-entry accounting, for example, can be seen as ‘thinking tools’ that allow people to organise and make sense of information in new ways, with profound and far-reaching effects. The more the sum of human knowledge increases, the more we need tools to explore, manipulate and mine it.
When a technology is complex, we need to know a lot about it to keep it working well. If it’s a particle accelerator, or a mission to Saturn’s rings, a vast amount of monitoring data is necessary, and there are very highly qualified expert scientists to interpret this data. But as complex technology filters further and further into everyday life, the amount of monitoring data typically overtakes the amount of expertise available to make sense of it; there are simply not enough highly knowledgable experts around to understand all the technology that powers our world today.
In digital media delivery, the technology exists to monitor in fine-grained detail every aspect of a network and its performance. The best systems are capable of producing huge volumes of data, which are invaluable for solving the often complicated and subtle problems that arise. But unlike a unique scientific enterprise like CERN, a media organisation exists in competition with many other similar organisations, so it can’t rely on being able to employ the ten or twenty best experts in the world to interpret its monitoring data and keep the systems running.
Under these circumstances, the most effective and valuable monitoring solution is the one that gathers all the data that is required, and then provides interpretation tools that make it easy to understand and act upon. This kind of system offers not only the means to acquire the data, but also the thinking tools that help technicians cope with the data and find answers quickly when they need them.
Or slowly. Sometimes the right answers only emerge when there is time to examine and think about the data at length, because when something goes wrong in a digital media service, the priority is to make it right again immediately, and not to spend time understanding the real root cause while angry subscribers flood the support lines with complaints.
To create a tool that would allow engineers to sit down after the event of a failure and examine in depth what happened, Bridge Technologies developed Timeline. This is a combination of server storage capacity for years of monitoring data, and a non-linear editor style interface for playing through the stored data. It allows engineers to see all the information gathered by the entire monitoring system over the days, weeks and months before and after the event, displayed like the tracks of an editing system for movies or music.
Each monitoring parameter is visible on its individual track, and users can scrub through the data at any point in the recorded archive, opening and collapsing groups of tracks, and zooming in to observe fine detail on all the visible tracks. The timeline shows content thumbnails, alarm markers and all the metrics and status displays familiar from the monitoring system’s real-time graphical displays, making visual navigation through the data simple and intuitive. The ability to observe and play through data recorded over a period of several months or years allows users to see recurring patterns that may be invisible over a shorter timescale. By zooming out of the timeline to view longer periods of time, unsuspected correlations can be identified and examined in progressively greater detail by zooming in to a short timeframe.
Engineers can search through the chain of events that led up to service failures, and generate reports for remedial action or the fulfilment of regulatory SLA obligations. The Archive Server and Timeline capability also simplifies detailed and comprehensive reporting for regulatory verification of closed caption conformance, loudness, SCTE35 signalling, RF trending and other parameters.
The data recording is provided by a specially-developed Archive Server in a simple connect-and-forget form, with each Archive Server module able to retain years of recorded data for multiple services. When connected to the Bridge Technologies VBC controller and configured through a minimal setup routine, the Archive Server automatically activates the VBC Timeline and its historical analysis capability.
One of the most important consequences of this technology is that when users can move back and forth through a graphical representation of the historical data, they can see all the connections between complex data types, and track the way different types of data interact and indicate behaviours that are unexpected. When engineers can do this, they find it much easier to examine an event and arrive at a clear understanding of what really happened and what the causes were. Examining the data retrospectively, they can observe patterns that begin to emerge only over a longer period. This is where root causes can sometimes be apparent that were not obvious before, revealing infrastructure weaknesses or provisioning issues.
With a thinking tool like this, engineers without a very high level of detailed knowledge can gain new insight into the trove of data gathered from the system, far beyond the immediate status feedback and diagnostic capability that the system provides for responding to errors and failures as they happen.
Published by FKT