[App_rpt-users] Allstar and Motherboards

[R. Wayne]

My mind is boggled by Beagle, Beagle-X, Raspberry, Raspberry-Pi... I am an internet professional that only knows mid-towers and rack mount servers with true power! But now you ask me to place a computer a little bigger than a couple packs of cigarettes on top of an inaccessible mountain? LOL.

We are intending on an aggressive link-in program once we work out all of the bugs and get our first repeater dialed in. We have two more already built and ready for deployment plus a couple of TKR-750’s. As it turns out the venerable Motorola MSF 5000 VHF with its 30KHz channel spacing and 2MHz RX window (for multi-channel remote base operation) is a poor option in Los Angeles. It hears too much off frequency too well with a 15KHz band plan. At any rate, we have engineered a main hub that will be housed in my companies level 3 data center. Not the company Level 3; a level 3 high availability data center. No cost to us. Perks of the job. We are further planning sub-hubs. We have a metro Los Angeles region; a desert region and a mountain region. Each region has resources dedicated to it, such as a 440 machine and a 900 machine in the desert. We have another 440 that attaches to the Mountain system and Metro its own resources.

The reason for this is segregation and diversity in an area of the country plagued by forest and wildland fires in the mountains, deserts and even within what is called the urban interface. An example is Griffith Park in Los Angeles. One would think that they are in the mountains when they are only in the foothills minutes from downtown. Our earthquakes also know no pattern. In the city; in the mountains or desert. Or both. Or all. In the event of a regional emergency we feel that we can better control resource allocations that will follow the sub-hub if/when removed from the main network. No need for Metro to carry Mountains fire EOC traffic. But Mountain may cross over into Deserts area. Enough said. By using sub-hubs and co-located resources the computer can better handle the loads on USB resources or RTCM’s (our preference for non-col-located resources).

So now you see that without planning routing of resources could be tedious. As well, those that might need to  use the web transceiver for access can just call up the sub-hubs node number and find that they have three repeaters in the group. There is a reliance on the PC and now the question for a mission critical application. Do we want to use a regular PC, like a Dell Optiplex Desktop PC w/8 USB  2.0 ports. Loaded with Linux they scream. But they run on 110vac. A Beagle-X? 12 volts but with limited USB? Help me out here. Our repeater sites in So. California are truly awesome at an average elevation of over 5,000 feet. But the trade-off is often inaccessibility during the winter months or even during a fire event when stuff breaks because it can. We have no illusions that we replace public safety communications, but we recognize what hams bring to the table. Reliability is often one that public safety has screwed up. Remember Katrina? Enough said there.

With all of this setup and explanation which board is the better option for reliability? We can rack mount a Beagle board easily enough Why? Earthquakes toss stuff around. A repeater offline because a cable came unplugged is unacceptable when it can be mitigated and planned for. No hangin’ doo-dads for us. But a Dell even with a 12v PS? $125 refurbished on eBay all the time with a 3.GHz CPU, 4GB of RAM... these things blow away a Beagle for a fraction of the cost. Switch the HDD to a SSD and are we talkin’ now?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.keekles.org/pipermail/app_rpt-users/attachments/20150422/6293ed61/attachment.html>