Slashdot Mirror
Will VoIP Kill the PBX?
gManZboy writes "Following up on their last VoIP article, Queue just posted "Not Your Father's PBX?" from Jim Coffman at Avaya Labs. Looks like the PBX may survive, but it's going to have to evolve considerably. I guess eventually corporate telecom goes away as a kind of island in the MIS dept? Maybe that's already happened?"
I don't understand your post. If you are saying that Avaya is only traditional telco, they have been selling VoIP equipment for over three years now. The last World Cup matches had the entire setup using VoIP and WVoIP services provided by Avaya...
Personally, I'm intrigued by software like Asterix and its capabilities, but I have absolutely no telephony knowledge and I'm not really sure where to start, like what kind of hardware I'd need in order to set this up with POTS. Lots of modems? Special cards for the phones in the office?
;)
You need FXO hardware if you want to take a phone line from a telecom and make digitally share it, route, connect it to the phone system. FXO, or Foreign Exchange Office handles calls that can't be dealt with in your local exchange (your PBX and local telephone extensions..) If you want to hook up a telephone to your PBX (as in a normal analog POTS style telephone) you want FXS hardware. The cool thing with Asterisk's wide range of protocol support is you can easliy connect analog telephones or a wide range of IP phones up to it across the network. Asterisk native protocol. IAX is a great way to get past annoying firewall issues that usually plauge most SIP based VoIP implementations that leave the local area network. Yeah I guess there is a lot to talk about, more than I'm going to post here... but there aren't *that* many concepts that you have to know before getting rolling with Asterisk. I just setup my first working PBX last night
We installed a Telrad PBX system with VoIP, 30 VoIP phones for sales people around the country, and 85 hardwired in the building.
The PBX now sits in a 19" rack, along side the rest of the servers. Its console is web based for programming, its just another thing in the data center, if changes need to be made a request comes into the IT dept now rather then an outside consultiant.
"The word "genius" isn't applicable in football. A genius is a guy like Norman Einstein," - Joe Theisman
When the manual switchboards were replaced with analog/mechanical switching, it did cause some changes to the system. You couldn't just speak into the phone and be connected, you had to manually dial a number. That particular change cut both ways - it wasn't quite so convenient, but it was less prone to error and it did allow more people to have phone service.
Then, along came digital exchanges. Early digital exchanges had numerous programming bugs (to be expected) but these have now been largely ironed out. Digital exchanges are faster, more reliable and easier to maintain, but the changes haven't been really visible to end users.
Now, we're moving into the VoIP era. Instead of dedicated lines and switched circuits, we're looking at a packet-based system with routing. VoIP reduces the resources needed (it can - in theory - make use of any spare network capacity between the two points to be connected) and it simplifies some of the more complex types of call. (Multi-point phone calls over IP are as simple as a multicast, for example. Over a switched circuit, it takes a bit more effort.)
Will VoIP kill the PBX? It depends on how you define the PBX. If you think of the PBX as a person manually connecting you, then the mechanical relay exchanges killed the PBX. If you think of it as merely the mechanism (human or otherwise) by which two or more people can be connected, then routers become the "new" PBX.
Of course, true VoIP will only be possible with a migration to IPv6. There are simply too many phone numbers, which would need an IP address, to use IPv4. Also, IPv6 headers are simpler, which makes routing more efficient. This makes the complexity of routing over much more complex networks possible. Finally, IPv6 doesn't fragment, which means that packet garbling should be less common.
It'll also require much higher bandwidths. The Internet is just too crowded to support much in the way of high-quality audio traffic. Packet loss is a shade too high, and latencies need to be cut. Your computer can quite comfortably handle uneven packet transmission, but the human ear can't. To fool the ear, you need much smoother traffic flows.
Smoother flows mean you need lower hop counts. This means the backbone needs to be better connected. There's been a tendancy for backbones to move towards the simplest possible layout. That's great for economics, but it means that paths are maximised. Not good for VoIP. It also means that if there's any outage, there's unlikely to be an alternative route, which means that network segments will be disconnected. Also not good for VoIP.
Telephone companies will be around for a long time, because they're about the only ones with the infrastructure and capital to build the highly connected networks required for VoIP. This is not a time for telephone companies to be concerned, this is their golden opportunity to demonstrate their continued relevence.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Organizationally, it started happening quite a while ago, at least in some industries. I worked as an IT director in a "Wall Street" firm for several years, and ended up with responsibility for telecoms, too. That wasn't because I sought it, or even wanted it -- I had to get up to speed on a whole bunch of new (to me) stuff -- but because it just made sense:
-
IT was itself the largest single purchaser of telecom services, since we had to provision links for market data, order transmission to the exchange, our private WAN, links to settlement / clearing agents, and so on.
-
The majority of telecom services had to interface, one way or another, with computer systems (e.g., to receive market data or to transmit trade data).
-
The PBXs and trading floor telephone systems were computer systems. (I can recall getting a new AT&T PBX installed. Their techs went to lunch while we were still testing. We found a little problem, which I looked up in the manual and fixed. The AT&T foreman was surprised at that: I told him "Hey, it's just a UNIX box.")
-
Following on to the last point, evaluating and choosing telecom systems steadily took more and more "systems-type" knowledge.
Buying a PBX was just buying a computer with some specialized I/O hardware; and it came with systems concerns -- security, for example, or the difference in performance between satellite and terrestrial links for TCP/IP.
Now, of course, we are seeing things like Asterix and VoIP, which will provide much tighter integration. Traditional voice comms are still important, but they're by no means something unto themselves.There are a couple of solutions that use some pretty robust hardware and software. We're in the process of evaluating some different IP telephony vendors, and have looked at quite a few.
Cisco's VoIP offerings run on Windows for the backend. Now, we're a Windows shop, but even our CFO who's a die-hard Windows guy expressed grave concern over the reliability of this approach.
Shoretel uses VxWorks as their software on a custom, 1U machine. VxWorks is pretty darn stable, and is what the Mars rovers run on.
Zultys looked very interesting, and runs on a PPC chip with Linux on top, but it didn't have the features that the other vendors (Avaya, Cisco, Shoretel, Seimens, etc) had, but were planned for their next software release. The other concern is the company isn't old, and they basically came out and stated that their entire goal was to make a good product so someone would purchase their company.
Avaya is still in the running for us, as is Siemens and Shoretel (albeit Shoretel is currently the most expensive -- they have a per user license fee that's really turning us off). They all seem to be pretty relaible from what we've seen.
Also keep in mind that Avaya's Audix voicemail system actually runs on Unix. SCO UnixWare, if I remember correctly. I might have to log into our voicemail system and check it out. I could be very wrong here, though.
Couple of answers for you. First, it is spelled Asterisk, like the web page you liked to. :-) Most of the hardware you need is available from Digium, the company that originally wrote, and still maintains and heavily contributes to Asterisk. http://www.digium.com and there is also a link from the Asterisk page you linked to above.
One to four POTS lines? Digium's WildCard TDM400 with FXO modules will fit the bill nicely. More than that, you will want to go with a T1 into one of their T1 interface cards. If all of the lines at your building are POTS, you will need a channel bank to convert them to the T1. Some people, including myself, have had limited success using a specific modem, but they are not nearly as reliable and trouble-free as Digium's hardware.
For your office extensions, you have several options. You can use several of Digium's solutions, including the IAXy which is ethernet-to-POTS, or the TDM400 card mentioned above with FXS modules for up to 4 extensions. If you have more than 4, you have to use those IAXys or a T1 interface card to a channel bank, then all of your phones attach to that.
Of course, there are several brands of IP phones you can use instead of the adapters above, such as Cisco and Grandstream. You would still need to attach to the PSTN phone system as mentioned above, but using IP phones would eliminate any worry for your office extensions.
I can't offer much more advice without knowing your needs, but if you want, go ahead and send me an e-mail with your situation and I'll help you figure out what you need.
Jeremy
Eek! I think you mean eke.
It's profitable to someone. The question is, who's got better lobbyists?
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Ethernet is designed to use four of the NON-VOICE wires in a standard 8 wire cable. All 8 wire, twisted pair (typically found connecting phones to a PBX or computers to your Ethernet HUB CAN run on the same wires. However, most people choose not to.
Basically, the savings is bull. Companies want ethernet separate from voice because they terminate at different devices.
In conclusion, all this will do is move everybody from two wires -- computer and phone -- to two wires computer and IP Telephony Device.
Again, you can argue that the computer and telephone can be the SAME BOX, and you are right the capabilities have been around for ten years (or even longer), but desktop computers -- to this day -- are not considered stable enough (even though, in truth most of them are) to run something as ubiquitous and important as a phone.
Kinetic stupidity has a new brand leader: Allen Zadr.
I run an Altigen PBX system at my offices, it does just about everything including VOIP if I wanted to use it. We have one T1 comming in that does both data and voice. 50% of the channels go to voice with 50% going to data (so 768kb). However if a voice channel isn't being used it gets switched over to data. We can then hook up a bunch of analog phones (single pair) or VOIP phones (10/100Base-T) and assign numbers to them. Oh yes, did I mention that the version of Altigen we're running is about five years old?
"Have you ever thought about just turning off the TV, sitting down with your kids, and hitting them?"
However, Voice over IP and even open controlled analog/digital converted PBX systems (like Asterisk), will be able to converge into a single, re-assignable open standard.
If you are comfortable with interfacing your servlet engines with your phone system, Voice over IP (and H.323 standards) will allow you to do so.
Offtopic, My Ass.
Kinetic stupidity has a new brand leader: Allen Zadr.
More information than you'll ever want on Asterisk can be found here.
Ubiquity and Dynamicsoft have SIP Servlet containers implementing the spec; there's also a reference implementation here to play with.
is not going away any time soon. A good example is the University of Michigan which has run a large on-campus phone system for many years. http://www.itcom.itcs.umich.edu/telephone/about.ht ml They do have some VoIP service.
It is interesting to note that most students on campus (Ann Arbor) are going to 7 digit dialing (565 exchange) and that service at U Hospital is going over to SBC.
How comprehensive is the IAX coverage network? As I understand it, local Asterisk servers connect to remote Asterisks (sounds like line noise ;) with IAX, with each edge server connected to local PSTN. So calls route across the Internet (or other WAN carrying the IAX) between local PSTN gateways, avoiding tolls. How big are the holes in that local coverage? How much of my PSTN line capacity in NYC would be hogged by strangers' incoming calls, once my gateway server is online here?
IAX is not a network, it's a protocol. There's no predefined grid of Asterisk servers, you only contact the servers you configure in your dialplan (like your ITSP's servers, ITSP=Internet Telephony Service Provider, like Nufone, VoicePulse etc. or your own servers in another location). Strangers won't hog your PSTN line just as they don't if you have a normal PBX (unless you let people dial in and dial back out).
Pavlov. Does this name ring a bell?
Sure, you can join Free World Dialup, or IAXtel, which are free worldwide networks. However, your system will not route calls through that you are not a party to, unless you set it up to do so, and have the owners of the network include your box in the routing.
Think of it in normal telephony situations. You have an office with 100 extensions, so you attach a PBX to your 10 phone lines. Just because you are attached to the public telephone network, does not mean your box will be used to route outside calls through to another outside party. Now if you so chose, you can give certain people access to do so, but only then would your PBX be used to route non-internal-party calls. Asterisk is the same way. On basic terms, Asterisk is simply a PBX for IP telephony rather than packet switched and analog telephony.
That being said, some users of FWD have set up their asterisk boxes to allow other FWD users to call local PSTN numbers. That was fully their choice, and they understand the consequences (bandwidth, phone usage, etc) that come along with it. It doesn't just install and share your line like P2P filesharing software does.
Hope that helps!
Jeremy
Because Vonage does not allow direct connection to their network, you would need an FXO port to attach their adapter to your Asterisk server, then whatever adapter you wanted to use to connect your Asterisk server to the side of your head (IP phone, adapter, etc.)
If you went with Broadvoice or another more flexible company, you can connect directly without needing any interface between your Asterisk server and their service.
As far as bandwidth goes, to be on the safe side, figure for each call, you need approximately 80-90K for full quality, non-compressed voice. So if you went in via Vonage, and back out to another point on the Internet, you are talking about doubling that for each call. Now if you are coming in via Vonage, then to an adapter on your LAN, you don't have to count that double bandwidth if you understand what I am saying.