IP VMS Scalability through Aggregation
Today’s IP video management systems (VMS) must deliver new services and features to a large and ever-growing class of new users, whose roles go beyond the traditional security mission. Scaling a VMS to provide these new services becomes increasingly complicated across multiple independent systems. System aggregation offers a path to scalability with improved ease of administration and better overall system performance.
Savvy organizations have recognized that video system data can inform all kinds of decisions. Growing businesses are hungry to consolidate and put this information to good use, providing oversight and driving optimizations across their organizations. VMS camera and data feeds are now being used to manage day-to-day operations, ensure policies are being followed, and ultimately make organizations run more efficiently. Consider these examples:
- The regional manager of a retail store wants to ensure a promotion is set up correctly without driving to each store.
- A facilities representative at home wants to validate that an off-hours alarm is a credible threat before dispatching local law enforcement, avoiding potential fees for nuisance alarm responses.
- A factory manager utilizes security camera feeds to ensure conveyor belts are running as expected, potentially even detecting belt failures prior to breakages, minimizing risk to employees and improving factory up-time.
- An airport uses analytics built into security cameras to calculate the wait time at security check-in locations and updates signs directing passengers to shorter lines, thus improving customer satisfaction.
The Challenge of New Users
Demands are being made from marketing, operations, facilities and more to leverage the information captured by the security deployment, requiring more effective ways to manage and distribute access to video data.
Additionally, an increased expectation of connectivity, driven largely by expanding networks and the proliferation of mobile technology, is placing new demands on systems that previously could have been close-looped. Users need access to pertinent information, often from multiple systems available whenever and wherever they happen to be.
Administrators need to ensure that the right people have access to the right information at the right time across these global distributions. An influx of non-security and untrained operators is placing new demands on user interfaces, driving the need for customization and simplification to ensure that the data a given user needs is what is made available to them.
Traditional video management systems were optimized around the security and surveillance user – supporting CCTV-style keyboards and calling up cameras by camera number or selecting a camera from a lengthy list of cameras, sophisticated search queries for finding and retrieving recorded video, and even more complicated export processes to ensure the authenticity of evidence.
While these systems continue to evolve to further incorporate the needs of security operators, the resulting interfaces and architectures are becoming unbearable for non-security operators. For example, finding and displaying a megapixel camera on a monitor in the security operations room is one problem. Serving up a view of that camera and all other related cameras from twenty different stores for the regional produce manager of a grocery chain working on a laptop from his hotel room is quite a different problem indeed.
To meet these diverse needs, Pelco by Schneider Electric is pursuing an approach using proven, standards-based technologies to provide the necessary structure to manage geographically and logically distributed systems.
By treating the VMS as an information system and grouping disparate VMS system information together, employing an aggregated architecture similar to those used by news aggregators and web-hosting companies, critical information can be both centrally managed and accessed by the expanding user base.
An aggregation solution
An aggregated architecture approach decentralizes the actual information, by keeping it local, while centralizing its flow, or the access to it. The aggregated approach brings reduced overall system complexity, reduced cross-traffic due to synchronizations, less concern over link failures creating database synchronization problems, more efficient latency management, and improved load balancing opportunities. The introduction of the aggregation layer also provides an effective way to provide access to new feature enhancements for a legacy install base.
Traditional VMS systems have had the ability to connect to multiple, independent systems for several years. The Digital Sentry system from Pelco by Schneider Electric, for example, uses the ControlPoint client to connect to any number of systems and display the cameras and recorders from those systems. However, if a given user’s permissions change to where that user needs to now see a new camera or be prevented from seeing an existing camera, it is important to have a method of updating clients and servers with new credential information. Active Directory and LDAP interfaces help, but still require a considerable amount of synchronization information to be exchanged between different servers. This same problem is also present in traditional federated solutions, which attempt to coordinate access across systems by sharing, synchronizing and replicating database information.
An aggregation approach, on the other hand, treats each system as an independent entity, pulling from the relevant systems at run time, through the client interface. While permissions exist on each independent system, through the aggregation interface an administrator can manage permissions and push changes out from a central location; pertinent information can be captured into one database and shared with other enterprise applications instead of repeating the same function on each independent system, one at a time.
Aggregation architecture enables centralized access to and management of the broader organization by providing and facilitating access to the relevant local systems rather than storing a representation of the full organization of systems and attempting to keep that synchronized. This allows the structure to be separated from the user access/permissions management, pushing those decisions downstream; thus allowing it to scale more appropriately.
Each site can be effectively managed locally as an independent entity. When an organization determines that it is necessary to see across all their installations, the aggregation server software can be installed, creating the aggregation layer.
The aggregation layer is accessed via the same user experience interface being distributed with Pelco products today. There is no waiting period while the local databases synchronize with some central authority or with each other. The only difference the user sees is that they now log into the aggregation server.
By not having to store information from the local system databases, the aggregation server becomes very easy to scale when system load increases due to the expanding user base. Pelco’s aggregation server software can be placed behind any 3rd party load-balancer, similar to those employed in the standard web-hosting model.
This allows an administrator to monitor usage and plan to add additional hardware as needed. Furthermore, the use of a standard load-balancer implementation creates an implicit failover redundancy model.