The UK Home Automation Archive

Archive Home
Group Home
Search Archive


Advanced Search

The UKHA-ARCHIVE IS CEASING OPERATIONS 31 DEC 2024


[Message Prev][Message Next][Thread Prev][Thread Next][Message Index][Thread Index]

RE: High Availability PC Based Home Automation


  • Subject: RE: High Availability PC Based Home Automation
  • From: "Nigel Giddings" <nigel@xxxxxxxxxxxxxx>
  • Date: Thu, 12 Feb 2009 08:53:20 -0000

Hi Rob,



I wonder why the decision to duplicate everything has been taken?



Any system is going to need occasional (?) repair and maintenance and
there is a balance to be struck between complexity and reparability.
Even if you go for a system that is resilient (in software) what happens
when a problem does occur, and it will, and how long before a repair can
be effected.



It is a reality of life that failure will occur, due to mechanical,
electrical or software failure. Sometimes due to a hidden design
problem...



I have been round this loop at home with SWMBO... The classic is
lighting control. I have opted for C-Bus in conjunction with Comfort. I
also have a PC (Server) running 24/7 providing more functionality over
the top of this. The key being that if the PC H/W or S/W fails the
lights still work. If Comfort were to fail, it hasn't yet, then I would
lose lighting logic. Of course should a dimmer fail you would loose the
associated circuits. You could argue this down to each bulb?



I am comfortable that I could obtain C-Bus and Comfort spares quickly,
less than 48 hours? I could opt to hold a full set of spares, at least
one of each separate unit but in the case of comfort this would double
the cost.



With regard to the PC hosting the application you are probably aware of
the Data Centre type servers which have built in Hardware redundancy.
The only issue I have experienced in the past was a BIOS battery which I
didn't replace in time, my problem as the H/W had been complaining for a
while... I have also hot swapped a PSU and HDD with no loss of service.
I run Homeseer and the biggest issue is Windows Patching.



As mentioned by others a UPS should be a given with shutdown S/W. Also a
lot of testing to ensure it shuts down gracefully and restarts when
power is reapplied...



Its not as exciting but I would keep things simple, and test it every
which way... by all means have a second machine there which can be put
into service quickly and has the benefit of allowing changes to be
tested off network before being applied to the live network...



The last thing I would say is to keep the PC H/W simple, avoid adding
H/W such as IO cards into the machine. I prefer to connect everything by
Ethernet and then the only H/W in the PC to be managed is the NIC. You
can then sub-divide your solution, for example 1-wire via a HA7NET using
Ethernet. If a repair has to be effected to the PC all the HA7NET
support is in S/W and none in H/W (except the NIC). If, on the other
hand, you have a 1-wire issue it is easily isolated and replaced (from
the local spares holding?).



My thoughts on resilient Home Automation....



HTH



Nigel



-----Original Message-----
From: ukha_d@xxxxxxx [mailto:ukha_d@xxxxxxx] On Behalf
Of craig
Sent: 12 February 2009 08:16
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] High Availability PC Based Home Automation



> So, people who are running PC-Based Home Automation Systems - - what
> lengths
> to you go to to ensure it's always available, and how do you implement
> it?

Ensure you use reliable 'branded' hardware with established drivers &
plugins.

Use redundancy in the hardware, (eg raid, psu, etc) & UPS.

Only install the bare minimum software and drivers required, after
testing
on a duplicate test system.

Monitor the software and environment.

> A potential client wants/needs a Very High Availability
home-automation
> solution (with Windows based software at the heart of it)- to the
> extent
> that he's considering purchasing 2 of every controller required (CM11,
> RAV232+, Dupline MasterGenerator, Cbus PAC etc etc) - - and connecting
> them
> to separate servers in an active-active or active-passive arrangement.
>
> Dedicated load balancing hardware will/may be placed "in
between".
>
> A Windows Cluster is one possibilty.
>
> 2 distinct servers set up with SQL replication is another.

I think you will struggle with this unless all the software, plugins,
network protocols, etc have been designed to be cluster aware.
Enterprise
software used in corporates has been designed from the ground up to be
used
in a cluster, to allow for scalability. Im not sure 'Home' windows
software
has been through the same design principles.

BTW - If the client wants a highly available system, I try to would
steer
them away from a CM11 :)





[Non-text portions of this message have been removed]



------------------------------------


UKHA_D Main Index | UKHA_D Thread Index | UKHA_D Home | Archives Home

Comments to the Webmaster are always welcomed, please use this contact form . Note that as this site is a mailing list archive, the Webmaster has no control over the contents of the messages. Comments about message content should be directed to the relevant mailing list.