Ensuring data redundancy and integrity has become more important than ever, especially as far as video surveillance is concerned. Besides hardening each individual component in the system, the resilience and robustness of the whole system should also be considered.
Ensuring data redundancy and integrity has become more important than ever, especially as far as video surveillance is concerned as the footage generated is important not only for post-even investigations but also for proactive, real-time decision-making processes. Besides hardening each individual component in the system, the resilience and robustness of the whole system should also be considered.
Server and database replication
According to
Synectics in its whitepaper titled “Surveillance Redundancy and Resilience: Preventing Data Loss and Downtime,” server replication or the real-time replication of primary server data and functionality over multiple physical/virtualized servers is perhaps the most crucial measure available. “Replication at this level guarantees full system control and data access should the primary server fail and should be configured as an automated process on detection of failure, removing the need for operator intervention which would delay action,” it said.
Of note, the backup servers, especially those used in mission critical settings, should always be located at separate sites in the event of a fire or disaster at the main site.
Cloud
In failover situations, storage space becomes an even more important issue – particularly for large scale solutions that routinely manage huge volumes of data and footage from thousands of cameras. According to the paper, organizations are increasingly looking at cloud-based storage, which includes a number of advantages including a reduced need for additional physical infrastructure.
However, there are pros and cons with storage hosted in the cloud. A major issue involves bandwidth capacity. “Video surveillance redundancy necessitates far greater levels of bandwidth than any other application. An image captured by an HD IP camera is about four times the size of one captured via an analog unit. In the context of a mega-casino, major oil and gas processing plant, or a critical infrastructure deployment – where hundreds, often thousands of cameras are relied upon every minute of every day – the challenge becomes clear,” the paper said. “Even if bandwidth would allow for this size and volume of data to be stored, speed would remain an issue. For sites where surveillance is mission critical and real-time capabilities are essential, the downside of cloud-based redundancy is substantial.”
But that is not to say the benefits of cloud should all be negated. “Take, for example, an organization that may need to routinely share information securely with third parties – perhaps law enforcement – that do not need or want to have access to all surveillance footage. In these situations, it may make sense to implement evidence locker redundancy to replicate data physically stored on the LAN/WAN within a secure cloud-based ‘locker,’” the paper noted.
Health monitoring
As prevention is better than cure, many organizations are now focusing on systems with real-time health monitoring functionality, which means the system can identify indicators of potential critical/catastrophic failure. This functionality can alert operators of these issues and, if necessary, take its own remedial action. It may, for example, hot swap digital recording systems if it detects indications of a potential fault, the paper noted.
“There are several significant advantages to this approach. It can help avoid complete data replication and storage scenarios, which in turn has positive cost implications,” it said. “Importantly, the combination of health checking and automated, individual device hot standby, means faults can be identified, repaired or replaced without any system downtime of impairment to everyday operations.”
As a side note, the paper suggested that while there are several data protection mechanisms available that can be implemented either from the unit level or system level, there’s not a one-size-fits-all solution as every end user organization is different, and requirements vary to regulatory demands, existing network infrastructure, solution size, camera types, sector needs, security vulnerabilities and many more factors. However, it did point out the following considerations that will always be relevant.
“Whatever the measures deployed, the end result should always be that there is no potential single point of failure,” it said. “Fully IP and hybrid surveillance solutions are designed to offer scalability and flexibility – enabling systems to grow organically as need dictates. The same applies to redundancy/resiliency measures. Focus on the absolute essentials first – further measures can always be implemented later.”