VoIP vs "traditional"?

[archalot]

Its time (way past time) to replace our outdated Executone 228 system. Office has about 55 extensions. We currently have 2 T1's with around 24 voice lines and 512K data. An important part of our business is a small call center with 4-5 agents.

I'm trying to decide whether or not to entertain VoIP solutions, or stick with the "tried-and-true" switches.

Suggestions and/or warnings to be aware of would be greatly appreciated.

Archalot
Columbia, SC

 

[johnpoole]

works great most of the time

that's my concern with voip, unlike cell phones my customers expect tdm/pstm to work perfect all the time. we have cisco voip wireless and yes it works great most of the time, of course we have the added plus of limited site coverage in a large hospital. interns (techie) love the wireless, but if they end up in an office away from the main building (17 medical parks) it doesn't work. walking from mp 5 to 3 ( 100 ft) and your past the range. those are inherent to wireless. the voip side works 90 percent of the time, 90 percent is not a very good batting average in the or.

have to agree with scott (srmega), it's the right app, at the right time, in the right place. as far as pure voip with zero tdm backup, i wouldn't make that leap yet

john poole
bellsouth business
columbia,sc

 

[srmega41]

Good point John. I think hawks hit the $64,000 issue on the head, possibly without really meaning to.

because we just use it to call back to the office.

What is missing from a topic like VoIP is this: CUSTOMERS!!!!! As I have said before, a multi-site environment is great for VoIP, as long as the customer is not involved very often. Can one employee handle talking to another and have sub-cell phone quality on the call; sure. If you add a customer to that formula, perceptions and expectations change. Your CEO will swallow a dropped call here and there if he/she is saving $500,000 a year in telco costs. He/she will NOT put up with that type of call quality if customers are a part of those calls, even for a much bigger savings.

Ask yourselves this question when looking at VoIP as a communications solution: Are customers going to be included in these VoIP calls?

If the answer is no, charge ahead and take as many risks as possible. Save money and play with the technology until you are blue in the face. If you are saving money, no one will bother you.

But........

If the answer to the above question is yes, proceed with EXTREME caution. It will only take one customer call that dropped or degraded at the wrong time for the entire comapny to came banging down your door and looking to burn you at the stake.

One of my favorite quotes from any movie that usually ALWAYS applies in IT belongs here.

You spent so much time trying to see if you could you never stopped to ask whether or not you should

Dr. Malcolm from Jurrasic Park.

Words to live by, words to live by.

Scott M.

 

[jpm121]

Wow, interesting thread. My observations:

- Numbers like 90% being bandied about -- if your VOIP system has 90% uptime, you need to fire your entire IT network department. If your wireless only covers 90% of your campus, you need to invest a bit more money on additional coverage, or fire your entire IT network department.

- LD carriers have been packetizing voice for years, and amazingly enough I can still call Mom every Mother's Day. The codecs may be different, and the hardware is different, but it's the same basic technology. VOIP isn't black magic, voodoo, or witchcraft.

- Routing critical voice over the public internet is best left to Vonage-type home service, IMO, and even then it's not for everyone. QOS is a must, whether it's on frame relay, point-to-point, or whatever.

- Customers -- a properly designed and maintained network should be able to serve customers just fine, whether it's using a traditional PBX or Cisco Callmanager or anything else. There are plenty of satisfied users out there who are happily providing high levels of customer service on a daily basis.

 

[johnpoole]

disagree slightly with the statement that tdm (pri/t1) is in any way simular to voip. but that is opinion. our network died thursday, email, all shared app's patient critial stuff at a major hospital. but the voip with redundant ccm's were not on the same (smoking) ups. the network was down, (the sky is falling) but what most people fail to see is the server farm was down, the backbone was up. we've used cisco wireless for about a year, small trial roll out with maybe 400 stations.. zero down time..

john poole
bellsouth business
columbia,sc

 

[srmega41]

jpm121,

You have all valid points, but one thing I would like to add to your observations on the LD side is this:

Who owns the network the LD carriers are using? They do. Who controlls the QoS on every one of thos calls? They do. And who dictates how many OC192 those "packetized" calls go throgh before the hit a codec and become TDM once again? They do.

My point here is that if we, as VoIP users and potential users, had control over the call 100% of the time it was IP, we would all do it (at least I would). The concern I always have is I am dependant on some other providers network to send a call from point A to point B. Until carriers ALL agree on a QoS or some other method of accepting, transmitting, and servicing, prioritized traffic, we are vulnerable to issues.

In TDM, the call is swiched, so the call path stays the same for every word, every syllable, every silence placed on the line. VoIP traffic packets can be sent on thousands of different paths, depending on network behavior at that moment, thus the bursty nature of data. If the call is LD, the bandwidth to make that call go through is so high, and provided by the same carrier all the way. The carrier is also dependant on that call being high quality, since the customers are riding that service expecting voice quality; VoIP using MPLS or Frame Relay do not get the same level of QoS on their meshed networks.

I also agree that there are many satisfied users out there, but most of it depends on their needs. If I have two sites on opposite sides of the US, and they are connected by an OC3, VoIP is a no brainer. If I have 100 sites all in one geograhic area, all using MPLS T-1 service, I need to tread a little more cautiously. This is not to say it can't and won't work; just that the needs of these two companies, and their network infrastructure, are vastly different.

VoIP will work for many people, it WILL NOT work for everyone. The challenge is to see how you fit in, while trying not to force a solution that will not fit.

Scott M.

 

[jpm121]

srmega, I think I'm seeing the source of some confusion here. VOIP doesn't (necessarily) mean Voice Over the public Internet. It means Voice Over Internet Protocol, which can be as simple as a switch on your LAN that has phones hooked to it, and a router that interfaces with your PBX. (I'm pretty sure you know this but you keep mentioning it...)

That's what I was getting at with my comment about Vonage type services. Any company with half a brain wouldn't rely on that type of architecture if they want to be in business long.

We're rolling out the 2nd iteration of a 6-location full VOIP solution, using dedicated point-to-point T1s between sites. For all intents and purposes, we control the variables -- at least as much as we would in a traditional PBX environment. A backhoe still can (and has) knocked out service to a location, at least until the backup kicks in. But by having control of the links, knowing the amount of bandwidth we have to work with, and applying the proper QOS applications, the variables can be mitigated.

My main objection to some of the comments in this thread are that they're FUD-based, implying that someone's voice traffic is going to be comingled with little Billy's illicit porn swapping ring.

 

[srmega41]

Certainly can understand that. jneiberger and I did a VoIP project in 2000 that would "hopefully" connect all 80 of our locations using our data network, replacing the TDM-voice, Frame-data, dual-network we had then. Using Nortel's ITG cards, we stared with two and expanded to six sites. That was the point we learned about the multiple encodings when calls were transferred between sites. We are still in a dual network environment, but are exploring the options once again.

I do NOT propose using the public internet for traffic, especially since I tried Vonage for home, using broadband cable access (bad idea). A dedicated broadband connection can work, but VoIP for public use, over the internet, is still in its infancy.

I am a big advoacte of VoIP in the corporate enviroment, but primarily as a replacement/enhancement to internal communications.

Scott M.

 

[cwc3]

I am 100 percent with you sr I am not ready to put all my
eggs in one basket !!also I think once voip and enhanced
wireless comes into play many of us will be getting out of the Biz.

 

[johnpoole]

or deeper into the business, my network techs here wanted the cisco wireless to be their baby, now they can chase people down for every problem, they have found it's a little hard to telnet into a coffee soaked voip station and do any good

john poole
bellsouth business
columbia,sc

 

[srmega41]

Too true. I started out a Meridan tech, with no asperations of EVER learning the data side (BIG MISTAKE). Now I manage both the voice and the data side, have received my CCNA, and have full visibility into the voice and data network. Had I kept my previous train of thought, I might not be working where I do today. I also think the opprotunities to grow expand exponetially(sp?) if you know both.

johnpoole is right on, learn to do both. Remember what it was like when you first learned the voice side of the house; exciting , interesting, challenging. The data side is the same way, only a whole new set of rules. The best part is you can use one to compliment the other, become a greater asset to any company who was two groups on opposite sides of the fence.

Scott M.
Network Services Manager
Data/Voice Networks and Systems

 

[cwc3]

IJMO but things were fine untill network people decided
to get involved in voice!!

 

[johnpoole]

i liked to do installs when phones had cranks on the side, but stuff happens

john poole
bellsouth business
columbia,sc

 

[cwc3]

yep I used to make 50+ amps a day by hand
back when there were only 3 tv stations!

 

[johnpoole]

that's a lot of amps for a butterfly, never got that fast, used 66e3 when i started, by the time amps came out, i was in the eq room leting the new guys work the floor

john poole
bellsouth business
columbia,sc

 

[cwc3]

yeah ia2 good times!

 

[jsaxe]

Geez-o-pete!

I've found the old-timers corner. Found my old butterfly under a huge pile of dust the other day, wondering whether to make a nutcracker or a lamp out of it...


jsaxe


"When the going gets weird, the weird turn pro."
Hunter S. Thompson (R.I.P. Doc)

 

[OldPhoneGuy]

...used 66e3 when i started...

How about 44A blocks and 50 type key equipment with lamps.

 

[cwc3]

yeah back when alexanders bells wife was the operator!!
hoy hoy

 

[johnpoole]

still have my butterfly in the shop, doubt ebay has one for sell.. keep it another 50 years and it may be worth a quarter or so. the 1st phone i installed had a crank on the side and my wife wanted an 8 track for the truck.

john poole
bellsouth business
columbia,sc

 

[aarenot]

archalot,
It sounds like you have decided to go with a VOIP solution, now the question is which one.
The data end solution companies have come up with a concept of stability and performance called the five nines. They put an aim of keeping the system up and running 99.999% of the time, as you probably know. What they should really call it is the three zeros and a one standard <0.001%, which is what the real goal is here. This is a hard goal to meet with any windows based system as your experience with windows based systems probably confirms that assumption. There is a concern with virus issues which is ever changing with windows based systems. I feel I should tell you in my experience, I have never seen a non windows based system go down to a virus in my 15 years experience. That is a risk you should consider seriously. Also, the windows based systems have all the windows issues and costs involved with them as an added factor in the ongoing cost of the solution. Cisco would be one of the major windows based providors, and probably the most experienced with windows based systems. Cisco in their expert position has opted to discontinue using the windows based solution for it's voice solutions from what I have read. So the windows based solutions for voice are in concencus not the ideal for reliability.


The voice history solution providors did not create the five nines concept as it would generaly in my experience be a step down in performance for even the lower quality voice history system providors. Generaly they have considered windows based solutions inferior without even being the most experienced with windows based solutions. Cisco seems to agree, if actions are considered to speak louder than words.
Also, have you ever heard of a windows based solution that lasted "way past time" for replacement like your executone system? Being on the cutting edge of technology is a risk, and can become the bleeding edge very quickly. If windows is the protection between the cutting and the bleeding, do you want your carreer throat leaning on that edge?
I suggest going with a hybrid system which provides the traditional telephone set independent of the LAN and it's wiring infrastructure. One that also supports functionality via the LAN for IP phones as well, whether soft or hard IP phones. They can all do IP site to site, via your network, VPN, etc., and if the LAN goes down as they do >0.001% , you only lose site to site and your desk top phone manager applications.

O.K., I have exposed my bias, and I suspect I will hear it from some. In fact I would appreciate any correction as it is after all what a wise man embraces.