Thursday, March 4, 2010

System Redundancy

In my last post I started the argument that IP video surveillance, in general, has no real advantage over its older brother, analog CCTV systems. Pelco, Bosch, Panasonic, OnSSI, and most others are still using a server centric system, which leads me to the purpose of this post… system redundancy.

With more and more systems integrating with video surveillance and more and more facilities and security managers relying heavily on their video surveillance it is becoming essential to have some fail safes in place to keep the system up in case of failure. One of the critical components of this is the DVR/NVR. Most systems (named above) require the camera to talk the NVR. The video is recorded, analytics (motion recording is an analytic) is performed at the NVR, compression, user access, masking, audio integration, text inserting etc, is all performed at the NVR. But what happens when the NVR fails…? The entire system dies. The cameras cannot be viewed, nothing is being recorded, you’re dead in the water until some tech decides to make some time for you and charge you $250 per hour to replace a $60 power supply.

So, this is where a properly designed and implemented IP based video surveillance system comes into play. There are a couple of systems out there that have complete system redundancy. This is done on several levels and several different ways.

1.       Not server centric – a system that does not rely on the NVR to tell it how to work is the first step to reducing a single point of failure.

2.       Hardware Redundancy – most servers available today have redundant NIC and power supply options. NIC and power supplies are the most common part to fail on a server, other than the hard drive.

3.       Software Redundancy – this is a good feature, some video management systems allow a NVR to fail over to another NVR automatically. When a NVR fails it automatically sends the video feeds to a designated NVR to continue to record with no down time.

4.       Hard Drives – get a server with a video surveillance rated hard drive. Server grade hard drives are designed for heavy read activity, in a video surveillance application there is a whole lot more writing to the hard drive, this is what causes them to fail. There is now, video surveillance rated hard drives designed to have huge amounts of data written to them. Also, have a system with a RAID 5 (or some level of RAID, RAID 5 is preferred) with hot swappable hard drives. This allows for a hard drive to fail without losing the video and you will be able to replace the hard drive without  having to turn the system off.

These are just a couple of ways to build in redundancy into your video surveillance system. Again, a properly designed system is key to providing a system that will be operating when you need it.