Individual technologies will use different approaches. That’s why I said it’s a rule of thumb. Zwave is not the only technology we are dealing with here. It could be anything and unless told otherwise, it’s probably a miss of two polling periods before marking something offline.
If you want a fault tolerant and redundant system, you need at least two of everything.
- 2 sources of power (e.g. battery backup)
- 2 sources of comms (e.g. internet and cell)
- 2 OH instances running on two separate servers
- etc.
How much all of this makes sense to do and deploy depends on what threats/risks you are mitigating with the alarm system. Without that up front analysis you are just “securing all the things!” There is no such analysis here so I have no choice but to go to the worst case scenario.
Far more of a professional system will be implemented with microcontrollers which are far more reliable and far less prone to errors, down time, power outages, security vulnerabilities, etc. That’s not to say that they are immune to these but they are not as susceptible to them as a general purpose computer.
A well designed alarm system will develop a system in away that looks at the entire system as a whole and compare that against the identified risks. Many risks can be solved through choices like architecture, what to use to implement using a microcontrol versus a general purpose computer, and the like.
I recognize that I am not an expert in physical security but I am an expert in computer security. The parallels are strong enough to convince me that even I, with the knowledge and expertise and experience that I have, do not know enough to actually design and deploy something even as good as the cheapest options from most of the long lived respected security system companies. I’m not convinced I could even beat SimpliSafe, and they are IMHO, a pretty poor system.
The big problem isn’t in knowing how to do it. The big problem is in knowing what to do and why.
Maybe this is all an argument between Impostor syndrome - Wikipedia and Dunning–Kruger effect - Wikipedia.
I’m not going to argue on this thread any more. I’ve said my peace. IMHO, anyone who says you can design and build a security system unless you are physical security expert, with OH or any other system, either doesn’t know enough to realize how little they know about the subject or they have such a low set of requirements/threats and risks they are trying to mitigate as to result in a system that many wouldn’t call an alarm system.