Networking

Reader Response: Open source IP-PBX's: Ready for the enterprise?

Scott Lowe describes a telephony dilemma he's working through at Westminster College and poses questions regarding open source IP-PBX's to TechRepublic's readers.

Bear with me while I get to the information related to the title of this post.  I want to provide some background about why I'm asking this question.A few weeks ago, I wrote in the Servers and Storage blog a post indicating that my organization, Westminster College, was considering replacing our Nortel Meridian PBX with something else.  At this point, we've basically settled on an Avaya IP Office switch.  However, although the Avaya quote came in at an excellent price-well within budget and much less than what we're paying on the Nortel unit-I'm beginning to have some concerns.  So, I thought I'd come here, outline our dilemma, goals and concerns and get advice from the TechRepublic community.

Dilemmas

  • Interoperability. The Nortel Meridian system is very closed. From requiring only Nortel IP phones to easily integrating with Exchange Server 2007, the Nortel system has not proven easily expandable. We've also had incredible difficulty integrating with Exchange Server 2007's Unified Messaging and continue to have problems. More than likely, a large part of our problem is our support vendor who had admitted that their knowledge of Meridian's newer features--including VoIP and SIP--is limited.
  • Cost. The Nortel system's maintenance costs are relatively high and will be making a jump this year as some work that was done last year comes off warranty. Worse, though, is Nortel's interest in seriously nickel and diming their customers to death. As I indicated, we did some work on the system last year. The end goals were to bring the switch back into compliance for support purposes and to make sure the switch was ready for version 5 of the Meridian software and to make sure the switch could use SIP, which would allow direct integration with Exchange Server 2007 Unified Messaging. The only promised outcome that was met was met was getting the switch upgraded to a point where support could still be maintained. Once the version 5 software came out, Nortel nailed us for a power supply upgrade that they failed to include in the original upgrade. And, once we rolled out Exchange 2007, in order to make it work, we had to buy SIP licenses. When we questioned it, Nortel told us, "The switch is ready for SIP just like we promised. But now you have to buy SIP trunk licenses." Not exactly a very customer-focused way to do business.
  • Complexity. Managing the Nortel system is not an easy affair. Granted, it's an enterprise level switch and we don't have one person dedicated to it, but we require the assistance of our support vendor for a lot of simple tasks. This is far from the first time I've had telephony in my portfolio, but it's been the most difficult.

Goals

In my previous posting, I outlined a number of goals that we have.  I'm repeating those goals here and expanding on them a bit.

  • Cost. As always, cost is a huge factor. Right or wrong, for a number of reasons that I won't go into, this is actually the #1 factor for this project. Most importantly, I don't want to be nickel and dimed anymore! Both up front acquisition and ongoing maintenance costs must fall within my budget parameters.
  • Easy to manage: I have a pretty small staff, so whatever solution we get has to be easy to manage. I'd like for adds, moves and changes to be easy to handle.
  • Open: If at all possible, whatever solution we put into place needs to be as open as possible. I don't want to be locked into a particular type of phone and I'd like to see a system that is more easily integrated into our other services than the Meridian.
  • Reasonable feature set: We don't need anything crazy. Just normal features that any 200 person office (plus a few hundred students) would need. Probably the most advanced feature that we need is a conference bridge.
  • IP-based with analog capability: For most of our faculty and staff, we'd like to provide IP-based devices to replace existing digital sets. However, we also support students in dorms and have a need for a relatively high number of analog ports as well.

Solution

Since I wrote my previous post, we've moved away from Pingtel as a final solution, but only because Avaya came back to us with a solution and pricing that made sense.  The Avaya solution we've settled on is actually a pair of IP Office units-an IP Office 500 to support the admin/faculty side of our needs and an IP Office 412 to support student needs.  At the beginning of the process, one of the selling points of the IP Office was free software upgrades and very, very reasonable maintenance costs, as well as a good initial acquisition cost.  Moreover, the IP Office feature set is exactly what we need.

Some concerns have crept up, though:

  • Software costs. From another Avaya customer, I've heard that Avaya will begin charging for software updates. Whether software updates would continue to be provided to customer with maintenance agreements is unknown because the Avaya folks we're working with won't confirm or deny the rumor. My worry here is that my cost projections will end up thrown out of whack.
  • Nickel and diming. As I sat with our reseller working through the final configuration, everything quickly became a license fee. Although the initial acquisition price is good, my worry is that I'm going to be paying steep fees to make even minor adjustments to our configuration. Now, I'm not opposed to paying for what we use, but we're talking about things like SIP trunks and soft phones. Avaya's soft phones require a pair of relatively expensive licenses, for example.
  • Openness. Although the Avaya system will support SIP-complaint devices, there is a SIP licensing fee attached for each device.

What to do?

I've been considering a number of options and discussing the project with a lot of people.  Here's a look at my options:

  • Do nothing. Stick with the enterprise-grade Nortel system and suck up the maintenance costs, nickel and diming and complexity. I'd really prefer to move away from this system and the related costs.
  • Do nothing, but drop all maintenance. A fellow college CIO has the same system we do, except it's 10 years old. They haven't paid maintenance in years. When they have a problem, they replace parts from surplus or from eBay and a local freelance Nortel expert is called in to assist. This would definitely meet the cost goal of the project and the unit is very stable. Further, there are local experts that can help us and parts are plentiful. Even so, phones remain a critical element in the office environment. I don't think I could responsibly drop all maintenance and then hope for the best. As someone once said, "Hope is not a strategy." For some environments, this direction works well, though.
  • Go with the Avaya system we've specified. It would work and the initial cost is fine. The future is uncertain, though. As things stand now, the maintenance costs we're looking at are fine, too. But, who knows what will happen? Will Avaya continue to provide software upgrades? The nice part about the Avaya system is that it meets most of our goals. It's not as open as I'd like, but the other goals are met.
  • Look at open source. I was seriously considering Pingtel before. Pingtel is based on the open source sipXecs and provides a broad feature set. Pingtel provides the support I'd like to have for our phone system, too. Even Nortel has based one of their new products on sipXecs. But, there are a lot of open source options out there. Asterisk, the Asterisk-based Trixbox, which was recommended by one reader of my previous posting, and a lot more.

The open source route

So, my questions for you are:

  • Have you ever used an open source PBX in your organization? This would include commercial offerings based on open source offerings.
  • Do you think that a commercial-backed open source IP PBX is viable for an organization with up to 500 or 600 phones, many of which are analog sets (and would have to remain analog sets connected through gateways)?
  • Would you go a totally different route?

About

Since 1994, Scott Lowe has been providing technology solutions to a variety of organizations. After spending 10 years in multiple CIO roles, Scott is now an independent consultant, blogger, author, owner of The 1610 Group, and a Senior IT Executive w...

10 comments
larry
larry

I would go a different route ... the option of open source telephony is too uncertain for serious consideration in a mission critical environment. I'm not a particular fan of the IP Office product line either as it displays a lot of the worst characteristics of hybrid systems. A better solution might be Iwatsu's ECS with TOLee Unified communications. It will integrate cleanly with the Exchange 2007 platform to perfom all facets of Unified Messaging, Presence Management, and Mobility Management. The system offers an extensive array of features with support for SIP, H323, Digital and analog devices. The platform extends conventional functionality even to Skype or MSN messenger for remote users. And yes, there is going to be hardware,licensing and support costs .... but I think you will find this platform is incredibly well priced by comparison. (what is the value of a freebie anyway?) Iwatsu is the worlds oldest telecom manufacturer (75 years). They have a balance sheet that is more than enviable with virtually zero debt ... and they actually own their manufacturing facilities and processes. OBF rate .007% - MTBF rate 22 to 500+ years Check them out or contact me for more info .... I would be glad to help Larry Dougan President - Nexus Telecomm Services larry@nexusinc.ca 416.533.5333

dbrooks
dbrooks

The short answers, then soap box. 1. Yes I have used and installed open source. Hard to make a living at it and putting the parts together don't always work as expected. If your vendor can't earn a living, they won't be around to support you in five or ten years. 2. 500 to 600 is a large installation for any open source company. Very few, will have working multiple references in this size. Get at least ten references. This is a middle size or small size for the major players. 3. Maybe. Nobody has the perfect system. Look for experience and a people you get along with. If you don't like the sales person or the techs, you will find reasons to hate the system. You are going to live with this system for years, find some one to trust. Is open source ready? - You get a weak maybe. A lot depends on how you support the system. If your staff is willing and capable to support the an open source system then it may work. Start with a small project and buy quality servers and phones. I would not rely on open source to support a large analog configuration. If you go open source, go SIP. A significant reason why proprietary phone manufactures can, and do, charge more for their phones is because they do more. A phone manufacture can create phones with custom features that simply are not available when you go pure SIP. SIP is still struggling with how multi line phones should work, and the SIP standard, is laced with RFC terms like like 'may' and 'shall' that allow a company to have a SIP compliant phone, that does not support all the features of your SIP PBX. If you are looking at price, you should also look at the second tier manufacturers like MITEL and NEC. Your size is in their sweet spot for PBXs. NEC recently purchased Sphericall, a reliable IP PBX that uses open standard phones and could easily support a lot of analog phones. More important than price is to find a company that you trust. One that you feel right about working with. There is more good equipment out there than there are good support groups and just because a company has world wide support, does not means all the offices have the same level of competence. (Have I mentioned trust?) I only have about thirty years experience in the telecom industry. I have a little less with VOIP - about twelve years.

robo_dev
robo_dev

The real question is "What problem are you trying to solve?" The PBX market is very competitive, and the second-tier players are, indeed, very hungry. Excellent observations about SIP phones. I was involved several years ago in evaluating various IP-based PBXs verus the conventional systems, and there was a huge 'feature gap' back then. Of course, the salesman will go on a long rant about 'how many features do you need'. One poster in this thread mentioned the use of DECT phones using SIP and Voip. That brought back flashbacks of a very expensive failed implemenation of this sort at a company I used to work for. The object lesson there was that ANY system you implement needs to be tested and tested and tested. In the case of the failed DECT phone implementation, the issues were that voice quality was terrible, and simple tasks like call forwarding were very difficult for normal humans. VOIP in general means you need to have a well-engineered and well-managed LAN and WAN. Putting VOIP over a shaky LAN/WAN infrastructure is like transporting a load of anvils in a Piper Cub.

luba
luba

Hello, Is Open source IP PBX ready for enterprise: definitely yes. I am a system integrator working mainly with Asterisk. We have deployed, an asterisk based system for a multinational with premises in USA (Michigan), Europe (Luxembourg and Germany), Asia (China). The system is 100% based on Asterisk, with SIP protocol and a mix of wired IP phones and DECT phones. It serves around 1,200 devices... the deployment started march 2005 and some more premises will be added in near future. Regards, -- Luba Vincent luba@novacom.be http://www.novacom.be

Bas.wijne
Bas.wijne

In my company I am responsible for telephony and with various contact centres this is not a small task and key to our deliverables. We have implemented Avaya 1,5 years ago and I can recommend it. It has to be said though that the best thing you can do is to specificy 'exactly' what you agree upon for deliverables from Avaya. Why not try and get a clause included for future upgrades and pricing of such? We pushed Avaya for full data-access across their platform for integration with our BI-solutions and it works well. Another thing to consider is stability. Avaya has been very good for us so far and you should ask yourself how much of a headache an open platform (potentially not as well tested in the real world) would cost you, not even mentioning support-time. Just my 2p

robo_dev
robo_dev

While I whole-heartedly support the open source movement, my advice would be to 'proceed with caution'. It's not about the technology or the fact that it's open source.... the risk is 'Who supports it?' There are high expectations for dial-tone in any enterprise. If something breaks, you may need software or hardware help very quickly. The capabilities and response time of the major PBX vendors support and service teams are a good chunk of what you pay extra for. For an enterprise, the service and support options for open-source IP PBXs are not robust enough to put their faith in, just yet. While, from a technical perspective, the open-source solutions are just about there. From a support/maintenance standpoint, there would be considerable risk in going that route for an enterprise of any size.

Scott Lowe
Scott Lowe

I have asked the vendor we're working with to get a contract clause from Avaya that includes software upgrades for 5 years. Our vendor indicated that Avaya was not willing to do so.

Jaqui
Jaqui

yes. even for the 500+ station type of implementation. though to truly get the benefits of the open source solution, you would need to use it on open source operating system. They can and will work on MS' systems, and on Macs. but to save the costs as much as possible, grabbing a dvd iso of a linux distro is more cost effective than buying a mac, or another MS Server system. I'm sure you have heard the stability argument in favour of open source before, but here is a salient point on that: When looking for phone and internet access for my home based business, while cost was a major factor, the simple facts of being able to contact the service provider was more important. Rogers-AT&T-Yahoo [ here in Canada ] business contact us.aspx form on their website, server error. call toll free number, modem screaming since the MS based PBX crashed on answering. I tried daily for three weeks and was never able to get in touch with them. if they were using open source solutions [ aspx = MS Server, VBscript and .NET ] maybe the web form would work If they use MS server for web pages, then they most likely use MS server for the PBX system also. open source, they may have gotten the contract for my business, since I would have been able to contact them. Then there is the free, open source GIZMO 5 sip software available, that turns workstation computers into internet telephone stations on top of that. Vonage, Skype, etc with only speakers and microphone required for making phone calls. [ conference calls even ]

jalperin
jalperin

Is open source telephony ready? Probably for many users with basic and even above-basic needs. [Disclaimer: I am an Avaya employee, and have spent several years watching/assessing the state of open source in general, including open source IP Telephony solutions. I have a certain vested interest in seeing you select an Avaya solution] However... you need to really consider what you perceive to be the benefit of the "open source" piece of the solution. Whenever you deal with a commercial entity offering an "open source" solution, you are really talking about "commercial open source", which often looks, smells and feels like any other commercial closed-source product. Pingtel and Digium/Fonality are all really giving you a commercial license, based on baseline open source technologies, which restricts you from code modifications and other things that come as benefits of a "pure" open source. One you use the Asterisk Business Edition, you are bound by its licensing terms, and this can impact your ability to leverage directions and enhancements being put forth by the "open source Asterisk community" directly. Instead, you would need to wait for Digium to release a new version, and obtain those updates as per their standard maintenance and support contracts. All in all, you are still going to be paying for maintenance and support services over the long run, and often from smaller companies that may not have the technical depth and certification requirements, nor the levels of compliance testing and interoperability with both network providers and 3rd party ISV/IHVs that Avaya has cultivated over many years. Just because a solution is claimed to have an "open source" element doesn't necessarily mean it is an "open" solution. Avaya products have open APIs, with technical support available through the Avaya DevConnect program, that can be (and are) leveraged by a large ecosystem of ISVs and IHVs to create interoperable solutions. While Digium and Pingtel may offer a lower "up front" cost by claiming to be "open source", you should look carefully and compare their support and maintanence offerings; their surround Ecosystem of SPs, SIs, channel partners, and 3rd party application partners; and their certification and compliance testing programs to ensure interoperability (as well as the technical competency of their channel partners delivering such service) as part of your overall assessment. Sometimes, Free isn't free, open isn't open, and the phrase "you get what you pay for" is all too often true....

Jaqui
Jaqui

Not every open source option is viable, nor is every proprietary option viable, for any particular need. It is up to the consumer to look at available options and pick the best fit for their needs. Any solution that works for the end user is the right solution for them. myself, if it requires windows, it is not an option. if it will work on linux, or any of the bsds, then it is an option. [ I will NOT buy a copy of any MS software, until they quit playing games with their protocols and formats ]