>Hi,
>
>I started looking into asterisk cause we're looking for a real world
solution.
>(when I say we I talk about a 50+ HQ and a 10+ branch office).
>
>We currently use a Panasonic analog PBX, with home-made IVR and PSTN lines.
>
>We'd like to deploy most of Asterisk's capabilities through out our
>organization - to save on long distance and international calls.
>
>I've been playing with asterisk for a week now, and I am charmed (if
>not madly in love) with it.
>
>Today I went on and bragged to my boss about it and how we can
>implement it instead of buying something like Cisco's call manager.
>
>He got excited too and wants to have an estimate of hardware costs
>for a solution that'll work for us.
>
>Suddenly I was weak at the knees?
>
>I have a couple of questions for you guys.
>
>A. Has someone tried this in a real world, production environment?
Yes, I have. It works fine if configured and managed properly. Only
when you start to use the esoteric features do some ragged edges
start to show, or if you plan on scaling to very large installations
you will also need to dedicate more resources to the project (as
would be the case in all such installations.) Additionally, if you
pick hardware that is faulty or inappropriate for your environment,
your users will have a bad experience despite what great things
Asterisk can do.
>B. What is the Asterisk server hardware recommendation for
>managing approx. 75 extensions and 16 analog lines?
This is not a clear question, though you may think it is. Your
budget, your current configuration, your desire to move to VoIP
phones, and your ease-of-use requirements all fit into this equation.
Do you want VoIP phones on each desk? Do you want to keep all your
existing analog phones? What features can you live without? Are all
the analog lines in the same place? What is your bandwidth budget?
What are you most common calling profiles? What does your long
distance plan look like?
[hint: don't answer all of these questions in a reply - it's much too
long, and you should expect to find the answers in the history of the
list if you read enough posts. If you are unable to answer these,
then you should see my last comments in this reply.]
>C. What telephony hardware do I need in order to get all these
>extensions and lines connected to Asterisk
See answer to B.
>D. Do I need to replace our lines (analog) with a PRI line?
That would be optimal, yes, and would probably be cheaper for you in
the long run. It seems that the most common breakeven point for PRI
cutover is around 10-14 lines.
>E. Can I use the existing infrastructure (connecting the
>existing extensions to asterisk, for instance)
Yes. See B.
>F. What is there to be said for network bandwidth consumption
>for this size of a deployment?
Network between where and where? You haven't told us anything about
what you intend on doing, and what the network would need to look
like.
>G. Do I need dedicated bandwidth for it to work properly
See F.
>H. Anything else I might have left out?
Quite a bit has been left out of your description. Do you want
redundancy? What is your required uptime? Have you factored in a
development platform for testing? What are your plans for long
distance? What is driving this conversion, anyway? (price,
features, flexibility, long-term costs, ???)
>See, I'm not afraid of hard work, and I already started working on
>solutions not embedded into the Asterisk package (like broadening
>the web interface, having a directory application that connects to
>our internal phone directory, etc.).
>
>I am afraid that since this is not a commercially available product,
>since there's no guaranty it'll work I might find myself holding
>both ends of the short stick.
>
>For instance, can I get the hardware and return it after a while if
>it's not working? Where?
I don't know about this one. I would expect the answer is "No"
unless the hardware is defective.
>This way, if it's a no go, I'm not stuck with useless hardware that
>cost thousands of dollars.
Reward is not without risk.
>In short, can Asterisk be deployed to a real production environment,
>like ours?
Yes. You would probably be wise to consult with someone who has had
experience with building such systems before, so you avoid pitfalls
and ask the right questions before you start buying gear.
>Shoval Tomer, MCSE
>IT Manager
>Softov Advanced System Ltd.
>Email: <mailto:shoval@softov.co.il>shoval@softov.co.il
>Mobile: 972-55-229220
JT