Understanding NVR Storage: Types, Capacity and Best Practices

If you watch police shows, you’ll frequently see detectives getting their hands on videos from security cameras that record the perp doing something bad or fleeing the scene of a crime. Have you ever wondered how that video gets captured and stored? In most cases, video from surveillance cameras gets processed and stored on a Network Video Recorder (NVR).

An NVR is a computer system with specialized software that captures video from network-connected cameras. It functions without dedicated video capture hardware and records video footage (an obsolete, but still useful term) to a hard disk drive, an SD card, a USB flash drive, or other storage device. Many security systems utilize an NVR as the main storage method. They have an important role to play in physical security because these systems generate enormous amounts of video. NVRs simplify the processing and storage of this large about of video data into a single, on-prem device.

NVRs role in IP video surveillance

An inevitable question that comes up in a discussion of NVR is, “What part does the NVR play in an IP video surveillance system?” An NVR is paired with an IP (Internet Protocol) camera, or cameras, that feed video to the NVR over a network. NVRs, or servers in some cases, provide the on-prem hardware housing both the video management software used to process and interpret the data as well the physical drives used to store the data.

Storage options for IP surveillance solutions

IP cameras generate video, a lot of it. To a security team, it’s valuable surveillance footage. To the storage system, it’s a lot of data to retain. Where does all that data get stored? Typically, there’s a storage solution either built-in, attached directly, or connected to the surveillance solution, including:

• Internal hard drive—built into the on-prem hardware, the internal hard drive usually offers some performance advantages, but its capacity is finite.

• External hard drive—connected to the NVR or other on-prem hardware, an external hard drive (HDD) offers potentially more flexibility than an internal hard drive, adding more external HDDs as data volume grows, but it can be costly to scale.

• Network Attached Storage (NAS)—connected to the NVR system by a network, NAS offers potentially highly scalable and high-performing video storage but requires a bit more technical sophistication or IT resources to support.

• Cloud storage—connected directly to the video management software through the Internet, cloud storage offers essentially infinite data storage for video without worrying about hardware upgrades or maintenance costs.

Each surveillance storage solution has its pros and cons. There is no correct solution, only what’s right for a particular customer’s use case. For example, storage performance may not be a big issue for an organization that simply wants to store surveillance video for future viewing. Indeed, such viewing may never occur. Cost is a factor, too. Hard drives and NAS solutions are a capital expense (CapEx) where current and future storage capacity is purchased upfront, versus cloud storage which has no CapEx and offers a pay-as-you-scale approach to capacity planning.

Capacity of NVR storage

NVR customers ask, “How long will my NVR record for?” The answer, as is so often the case in technology, is “It depends.” How much video can fit in a gigabyte of disk space? This depends on several factors, particularly:

• The resolution of the video image—High resolution video, such as 4K (8 megapixels) requires vastly more data per minute of video than low resolutions, like 1280x1024 pixels.

• Frames per second—The more frames per second, the more data a minute of video will be required for its storage.

• Video quality—Most cameras and NVRs enable low, medium, and high-quality video recording. The higher the quality, the more storage capacity the video will require on the NVR system.

• The type of video compression—Digital video gets written onto NVR storage using one of several formats, such as MJPEG, H.264 and so forth. Each format will generate a different amount of data for the same camera input.

• Scene activity—Digital video codecs have methods of minimizing the recording of repetitive visual data. For example, a video of a brick wall, with nothing moving in front of it, will be smaller in data storage terms than a video of a football game with people rushing around on screen.

• The number of cameras—The more cameras feeding into the NVR, the more video storage capacity that will be required.

Some hard drive makers offer drive calculators that enable prospective buyers to estimate the drive capacity they will need for their video parameters. For example, according to a calculator provided by Western Digital, a leading manufacturer of hard drives, to store 30 days' worth of video from 10 cameras, running at 30 frames per second, 24 hours per day, at 1920x1080 resolution, medium quality, medium scene activity and H.264 compression MJPEG format would require 19 terabytes (TBs) of storage capacity. Note that if you drop the frame rate to 18 frames per second, the storage requirement falls to 11 TBs. It’s hard to provide the exact storage capacity needed because there are so many different variables coming into play with cameras capturing video. Even with the lack of predictability for storage capacity, the number one guideline is “don’t run out of drive space.” Best practices when capacity planning is to over-provision your on-prem storage media and consider a hybrid cloud approach to video surveillance storage to offload capacity so you can scale economically (more on this later).

Best practices for surveillance storage

Aside from not running out of storage capacity, best surveillance storage practices include making sure that you are choosing the right storage type and capacity for your specific needs.

For example, if your NVR is going to be deployed in an unattended remote location, then installing a storage solution that needs hands-on attention is going to be a mistake. Or, if you don’t have an IT department, then it’s a wise practice to go with a storage solution that does not require much, if any, administration. It is a good practice to design for storage redundancy. If storage unit A goes down, it should fail over to storage unit B. On a related note, it’s smart to do regular backups, especially if video footage is important for regulatory compliance or defending against legal liabilities.

Other best practices run from monitoring storage usage and performance to conducting whatever maintenance and upgrades are necessary. These latter tasks imply the presence of a person or team who has responsibility and accountability for keeping the storage solution functioning as intended. How full should you let your NVR or other storage solution get?

Opinions vary, but UniFi Video, a maker of cameras, suggests that it is best to leave at least five percent of a video storage drive’s capacity free. And, from there, do not let the spare capacity go below two percent. There is some wisdom in over-provisioning NVR storage, mostly due to unforeseen changes in video parameters. For example, if your organization decides to add another camera, or decides to upgrade video quality, it’s useful to have some storage “headroom.”

Similarly, if video retention requirements grow, for example from one day to five days, that will create the need for more storage capacity. If a move like that prompts a hardware procurement cycle, which can take months or even more than a year, storage will get maxed out and the NVR system will be unable to meet the new requirements. It is also a best practice to make sure that the video data stored, whether in an NVR or other on-prem storage device, is properly secured against cyber threats.

Surveillance video may seem like a low-value target for hackers, but the footage can be useful to people intending to vandalize or burglarize your operations. Videos of your workplace can be valuable for social engineering attacks, as well. A cyber-attack can also disrupt the surveillance process, which exposes your organization to risk.

Advantages of hybrid cloud NVR storage

Some users of on-prem storage solutions are adding cloud storage to their video surveillance systems. Others are migrating video storage completely to the cloud, but most of the time the solutions are hybrid in nature. This might mean storing the most recent footage on-premises, while utilizing cloud storage for older material and for backup purposes.

The hybrid cloud approach for video surveillance storage offers several benefits over a pure on-premises storage solution. One is scalability. Cloud storage can scale on demand, and effectively without limit. If you need 100 TBs in the next hour, you can get it in the cloud. It would take weeks or even months to acquire that much capacity for on-premises deployment.

Cloud storage is also flexible. It can “burst” up in case of a temporary increase in video material, but then rapidly scale back down when the need for storage has passed. For example, if a store is open 24 hours a day during the holiday season, it will need more surveillance video storage than it does at other times of the year. With cloud storage, it’s possible to get the storage for the holidays, but then go back to a lower storage volume afterwards.

The cloud provides some economic advantages, too, though it’s necessary to work those out carefully in each case. The main benefit is the avoidance of CapEx, especially when contemplating over-provisioning. If video storage growth projections say you will eventually need a 5 TB disk drive, even if you only need 1 TB right now, you will still have to lay out the CapEx for the 5 TB drive today. The cloud, in contrast, lets you pay for what you need when you need it.

Further advantages of the cloud include greater reliability and security. This is due to the fact that cloud providers can focus on maintaining the security of video data allowing business owners to focus on their business. They also have failover capabilities, so if a storage array at one cloud data center goes down, an array in a different place can rehydrate the missing video. Solutions that utilize on-prem storage exclusively would need to spend additional time in transferring the data from another on-prem device or risk losing the video data altogether. The cloud provider also has the personnel and expertise for managing their storage systems, which helps you to avoid hiring people to manage your own on-site NVR or other storage device.

Conclusion

NVRs capture video from IP cameras on a network. They are used mostly for video surveillance. Storage is an essential component of an NVR solution because security and compliance policies typically require video to be stored for a period of time. The amount of NVR storage required depends on many factors, such as the number of cameras, frame rate, video quality, and compression. Over-provisioning may be the best approach, though one that is not necessary in the cloud.

NVR owners have a range of choices for NVR storage, running from internal and external hard drives, to NAS and cloud storage solutions. Each has its advantages, with the cloud offering great scalability and flexibility, along with some potential financial benefits. The best practice is to acquire an NVR storage solution that aligns well with the NVR system’s operational purpose. The good news is that buyers have many options.