Monitor if NVR is still online?

[xxHi-Fixx]

Curious if anyone has some ideas for monitoring off site NVRs. They are Windows based and often the customer doesn't realize the NVR is offline until something happens and they can't connect.

The NVR can't generate an email alert if it was unplugged or locked up. I'd love it if we could install software that would ping a server everyday like an alarm timer test.

Any ideas or products?

 

[Teken]

Curious if anyone has some ideas for monitoring off site NVRs. They are Windows based and often the customer doesn't realize the NVR is offline until something happens and they can't connect.

The NVR can't generate an email alert if it was unplugged or locked up. I'd love it if we could install software that would ping a server everyday like an alarm timer test.

Any ideas or products?

The first step is to determine root cause for the computer system failing to connect. Anything else you do is simply a band aid as it doesn't address the root cause. Having said this the SNMP protocol is used to monitor network based hardware.

There are lots of open source / paid services that can ping and monitor the same.

Sometimes people just want to skip the root cause and go for broke and incorporate a Band Aid. Regardless, one can use any of a dozen Watch Dog Monitors that will hard boot a unresponsive computer system. This is just one example of a all in one appliance that sends emails, pings, watch dog, and auto boots:

Web Power Switch! - 15 Day Free Trial! - Ships Overnight - From $149!

Web Based Power Controller $179!

www.digital-loggers.com

We've used this appliance in tens of thousands of installations over the years and have moved away from them as their warranty isn't worth the paper its written on. We always like to support our own economy or brothers in the USA before purchasing any foreign made hardware.

When the product works it just works . . .

Over the years Microsoft WMI has been implemented or used in concert with SNMP monitoring. Really comes down to how serious the network admin is and the infrastructure to support the same. It obviously goes without saying to impress the point that if you hard boot a windows computer system repeatedly expect a complete meltdown to follow.

Which can be from a completely corrupt operating system to blowing out a hard drive / MOBO.

Hence the importance of identifying at what point in the OSI model the problem is happening?!?! Keep in mind Pinging a network appliance does not tell you a computer system / network appliance is operating correctly! It simply tells you the NIC is present and replied back with a ACK.

It in no ways tells you the OS or application is running fine as the problem can be at the application, network, transport, session, physical, data link or anywhere in between.

Regardless, here's a quick snap shot of my network monitors . . .

 

[xxHi-Fixx]

Some good info, but to elaborate I'm not referring to one particular NVR locking up but a solution to monitor if any of our servers are still running. We have servers installed in mechanical rooms for property managers who are not onsite. Sometimes it's as simple as it got unplugged, other times the service didn't restart after a scheduled reboot. Whatever the cause many of our customers only look at cameras when an even happened and realize it hasn't worked for months.

The server software we are using has the ability to control network addressable outputs. I'm thinking that tieing one of these outputs to an existing alarm panel to create a heartbeat /timer test to a central station may be the best solution. It can be easily setup to notify a call list or email if we don't receive the signal every 24 hours and if it's automated it doesn't get forgotten about.

By running in a rule that requires the camera server service to be operating I think it will establish that
A. The NVR is powered up
B. The server software is running.

This won't catch every issue but I think it will catch most. Actual server issues / camera failures have no issue going through an email server if the system is online.

 

[Teken]

Some good info, but to elaborate I'm not referring to one particular NVR locking up but a solution to monitor if any of our servers are still running. We have servers installed in mechanical rooms for property managers who are not onsite. Sometimes it's as simple as it got unplugged, other times the service didn't restart after a scheduled reboot. Whatever the cause many of our customers only look at cameras when an even happened and realize it hasn't worked for months.

The server software we are using has the ability to control network addressable outputs. I'm thinking that tieing one of these outputs to an existing alarm panel to create a heartbeat /timer test to a central station may be the best solution. It can be easily setup to notify a call list or email if we don't receive the signal every 24 hours and if it's automated it doesn't get forgotten about.

By running in a rule that requires the camera server service to be operating I think it will establish that
A. The NVR is powered up
B. The server software is running.

This won't catch every issue but I think it will catch most. Actual server issues / camera failures have no issue going through an email server if the system is online.

As stated up above in the so called networking industry the use of SNMP is the primary method to track, monitor, and recall historic events on IP based system. It's important to understand there are thousands if not millions of scenarios where a network monitor does not or will not capture a fault.

This is why its imperative the company / team fully understand what the needs and goals are. If a person doesn't understand or know what the problem(s) are you're doomed to failure. Anyone who has used a computer for any length of time knows just because a computer displays the old school BSOD (Blue Screen of Death) has absolutely no relationship to other low level operations or functions of a computer system.

Hence pinging a network appliance offers zero assurance a piece of hardware is 100% operational. Even when a person installs a 3rd party agent on a target machine this too doesn't give 100% reliability that a machine is up and fully operational.

At this juncture I am going to over generalize for a moment to simply illustrate a point. When a piece of hardware is designed from the ground up to incorporate what most people call a watch dog timer. This is the only true method to know if something is fully operational. Because very few network related appliances have such things in place or don't monitor low level vs high level functions people in the networking industry employ multiple layers of monitors to provide more insight and raise the confidence level of (IF) a system is fully operational.

These methods employ push & pull telemetric's on everything that has to do with the system. That span drive space, temperature, memory, cpu, bandwidth, processes, etc.

• Identify root cause
• Incorporate push & pull system monitoring & verbose logging
• Integrate remote access & control
• Employ back up, fail over, and redundancy
• Test, validate, and Track incidence