What is Automatic Camera Failover?
Automatic Camera Failover is a feature in Nx Witness which allows System Servers to automatically recover when a power or networking failure occurs on a Server. Automatic Camera Failover can be enabled on individual Servers. When enabled Failover tells the server to grab cameras from any failed Server in the System.
A single version 6.0 or higher server can handle a maximum of 256 cameras. So, for example, if a Server is actively recording 200 cameras it can act as a Failover Server for up to 56 additional cameras from other Servers in the System.
Failover takes approximately 1 minute in the instance of a network or power failure.
Automatic Camera Failover Across Multiple IP Subnets
Automatic Camera Failover works best when the failover server, another server, and cameras are on the same IP subnet. Although, being on the same subnet is not required as long as the failover server can reach the other server and its cameras (and vice-versa).
To allow the servers and cameras to be visible to one another, you must port forward the servers and cameras and take the following steps:
- If the cameras were added via autodiscovery:
Configure the failover server to accept multicast requests from the cameras by multicast forwarding to the subnet on the failover server using each camera's IP address as their source address. - If the cameras were added manually/using a known IP address:
Ensure each camera's IP address has not changed since their last server. The failover server will connect to these IP addresses using unicast.
Please verify that your network's routing is configured correctly to ensure a successful Automatic Camera Failover across multiple subnets. A good test is to ping the failover server from the other server and manually add the cameras from one server to another.
How to Enable Automatic Camera Failover on a Server:
The failover priority setting is a system-wide option and is synced across all servers in the System. When failover is enabled, if the last remaining main storage location is invalid or failed, failover is triggered automatically.
- Right-click on a server in the Resource Panel and choose Server Settings.
- In the Server Settings dialog, check Enable Failover.
- Enter the maximum number of selected server can handle (256 Maximum on Core i3 or above CPUs, 12 maximum on ARM CPUs)
-
Set the Server Location ID. By default, this value is 0 for all servers with failover enabled. Servers that share the same Location ID can failover to one another but not to servers with different Location IDs. This ensures that failover occurs between appropriate servers (for example, you may want to set servers near one another to the same Location ID and severs that are further away to a different Location ID).
- Click OK.
- A requirement since Nx Witness 4.0 is to make sure every Server with this feature enabled has the storage drives to allow failover recording. If there are no drives available to record footage from these cameras, the cameras will not be moved.
- If you want to add additional Failover Servers in the same System repeat the steps above for each Server.
It is also possible to establish a failover priority for individual cameras, so that the most important streams will be transferred first, lower priority cameras after that, and inessential devices can be set to not transfer at all.
To Setup Failover Priority:
- While still in the Server Settings dialog Click on the Failover Priority button.
- Assign a value of High, Medium, or Low priority to each camera.
Failover for System cameras will follow assigned priorities - with High priority cameras failing over first, then Medium priority, and finally Low. Failover assigns cameras to the first available Failover server recognized by the System. - Repeat this for all cameras that should be prioritized.
Questions
If you have any questions related to this topic or you want to share your experience with other community members or our team, please visit and engage in our support community or reach out to your local reseller.
Comments
2 comments
Lets say I have 2 remotely merged servers into one system, will failover still work? or should the other server have access to the cameras in the failed server's workspace?
what is the best scenario for such situation? vlan maybe ? port forwarding? But if port forwarding is the chosen option, should the cameras be connected to the first server by their external address?
Hi Kamal,
Failover will still work, but the cameras should indeed have access to the cameras in the first place. VLANs are the best practice, and in case of port forwarding, you need to add the through the external IP address.
Article is closed for comments.