Video Distribution...

sonny

Always nice to get the "mad" call on Sunday afternoon from the client trying to watch the game. Although the concept of video distribution makes sense at a large residential complex, I sometimes wonder if this isn't too much rocket science going on to simply watch tv...

1) User Presses button on R4
2) Zigbee gateway gets message
3) Zigbee Passes to Ethernet Switch in Pool House
4) Pool House Switch passes to Rack Switch in House
5) Master Control processes...(TV On via Serial, Turn Cable Box On, Switch Video, Switch Audio)
6) Video over Cat5 to TV
7) Audio over Cat5 to Amp in Pool House

I would like to hear both philosophical arguments for/against centralized video, and techniques you've used to try to "harden" this solution to make it as reliable as possible. In this case the Zigbee gateway is offline, I can cycle the power on the gateway and reboot the processor remotely, but that hasn't fixed the problem.

vining

Well with local video there are less things to go wrong and quite possibly if something does you'll have nothing to do with solving the issue. Local cable box, TV and using the cable remote. On the other hand that's really no work or income either.

The problem now is with HDMI and the desire to have the ability to distribute 1080P. Componant can be relatively bullet proof as long as the control system functions. In your case it's not.

Is the R4 dead? Batteries or does it need a reboot? I give users an easy to access reboot button on the R4s just for these reasons. I'll still get the call but it's easier to have them go to a utilities page then it is to get them to hold the input and back button at the same time. I would have them reboot the R4 and see if re-acquires the gateway. Usually takes a few minutes. You've alread already rebooted the gateway and I assume you see it in the online tree of NS.

In my systems any remote or TP can be used in any room by simply selecting the room you want to control on the UI you want to re-purpose so if something goes wrong with a UI they can grab any other UI that isn't being used to replace the one not working.

sonny

Turns out the Zigbee gateway had been unplugged from the network, so not really a "system" problem. I do setup TP's where they can control any room / system, but haven't done that with remotes. This is a legacy system I've taken over, which makes it a little more difficult.

Anyone ever do automated power cycling of devices that fall off-line via a controllable power conditioner?

Prodigal

I've done power cycles before, specially in a pool controller brigde communication that was constantly freezing. Every day or so, it would go down.

Since the device boot up was failry quick, I created a power cycle with a simple relay at around 4 am everyday.
Never had a problem, ever since.

[Deleted User]

Try RMS monitoring

sonny wrote: »

Turns out the Zigbee gateway had been unplugged from the network, so not really a "system" problem...

If the system was programmed to be monitored by Resource Management Suite, you cound have been sent an email when the Gateway / R4 went offline.

TurnipTruck

B_Clements wrote: »

If the system was programmed to be monitored by Resource Management Suite, you cound have been sent an email when the Gateway / R4 went offline.

Nice... or the poor man's way, use iEquipmentMonitorOut to send emails based on prolonged OFFLINE events.

vining

Or even cheaper yet just set up a server on your office master and send notifications directly.

the8thst

B_Clements wrote: »

If the system was programmed to be monitored by Resource Management Suite, you cound have been sent an email when the Gateway / R4 went offline.

The new version of RMS looks pretty slick (other than being built around IIS).

I am just waiting for the new stuff to drop and some payments to come in before I start adding it to most of our jobs.

[Deleted User]

the8thst wrote: »

The new version of RMS looks pretty slick (other than being built around IIS).

Thank you. The interface offers a rich web application reducing page flips and redraws to navigate to useful information. The user configured Dashboard and Widget tools offer a lot of flexibility on monitoring locations.

(Actually, we are using Apache Tomcat this time around. )

sonny

I probably need to look at Home Manager again. Has AMX thought about virtualizing this and offering it as a hosted solution?

I've been working on a health engine that will track errors, events, handle device resets, email logs, etc. One thing I've never been sure about, is there a way to get system information programatically, i.e. memory usage, buffer status, etc?

[Deleted User]

sonny wrote: »

I probably need to look at Home Manager again. Has AMX thought about virtualizing this and offering it as a hosted solution?

I've been working on a health engine that will track errors, events, handle device resets, email logs, etc. One thing I've never been sure about, is there a way to get system information programatically, i.e. memory usage, buffer status, etc?

It would be best to contact your AMX sales rep to answer these questions.

PhreaK

sonny wrote: »

I've been working on a health engine that will track errors, events, handle device resets, email logs, etc. One thing I've never been sure about, is there a way to get system information programatically, i.e. memory usage, buffer status, etc?

Yep, we do this on all our systems. Unfortunately the only way to get most of that info is to have the master telnet to itself. Not the most efficient solution but it works. Gives us full memory utilization info (volatile, non-volatile, duet), disk utilization, processor load, buffer utilization, code modification time/date, as well as some debug messaging.

felixmoldovan

Can you elaborate? Not efficient? In what exact aspect? Or you meant "not ideal", "not elegant"?

vining

felixmoldovan wrote: »

Can you elaborate? Not efficient? In what exact aspect? Or you meant "not ideal", "not elegant"?

You need to make the master open a socket to itself, you then need to log in with your user/pass, then send the desired telnet commands. Not as efficient or as elegant as having commands that could do this directly in Netlinx.

felixmoldovan

I know how. Thanks. I am asking why this will be not efficient. (Strictly referring to resource used less than optimal)

DHawthorne

I do it as a matter of course on every job so I can save the logs as a file that won't poof with a power reset. The only downside I see is it uses up an IP connection, and you only have a limited amount. That said, I've yet to run out (I think you can have 20 at once).

felixmoldovan

Thanks for clarifying. I am also using this on every project and never felt like I am compromising any of the stability or efficiency of the system. Not more (rather less) than using a regular socket to an external device.
And yes, I am aware of the 220 IP connections limit, but I have yet to come anywhere close to that from a single processor, so I am not going to consider that an efficiency drawback, i'll maybe call it a housekeeping challenge.

PhreaK

felixmoldovan wrote: »

Can you elaborate? Not efficient? In what exact aspect? Or you meant "not ideal", "not elegant"?

Not elegant is more of an accurate description. It's entirely functional, just a bit of a roundabout way of doing things. It also means that you can't kill telnet to lock the masters down to SSH only. However, if you're running RMS v3 you're stuck with this anyway as the RMSEngine module uses the same loopback telnet connection trick to get system info.