Advise on VOIP provider - dslreports.com

Author	All Replies
carmelo join:2009-11-05 Canada	Advise on VOIP provider I am looking for some suggestions for a VOIP provider that is realiable and feature rich. I need this for a small business. I am considering a hosted PBX but I don't quite grasp the difference between that and a VOIP plan. What I am looking for is something that gives me plenty of features and customization: -Voicemail, call display, call conferencing -DID (Calgary) -Follow-me -A cell phone solution (to lower long distance costs) -Call recording -Good codecs support (at a min 711, 729, but would like GSM, 729) -Soft phone -Voicemail transcription (not sure who does that, but would love it!) -Advanced call forwarding features (e.g. fw to this number after 8PM) Do you think for these kind of things I need a hosted PBX solution? Or is there a provider that has some plan that would include these features? I have done some reading in the forum and came out with this list of providers: unlimitel primus callcentric iristel voicenetwork link2voip I really don't want to manage my own Asterisk box (something I did 5 yr ago, and don;t want to go back...). Any advise?
	to forum · permalink · · 2009-11-05 18:36:25 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON 1 edit	Voice Network in Canada and VoipStreet are both good, we've used them for years with few problems. Note that you need at least two providers, and they should both be long-established companies. We use a VoIP VPS (running Trixbox CE) hosted at Lylix in New York. Lylix is first rate and not terribly expensive. I really like having my own instance of Trixbox to manage myself although it was a huge pain to set up when we knew nothing about VoIP despite running on a 3Com VoIP PBX for years. Getting used to Asterisk was a challenge, but I sure am glad we made the move. Moving a little farther forward, I expect we will move to an instance of PBX in a Flash onto our own fibre in Bracebridge shortly. The reliability of this fibre is high enough that it worth the risk to move back on-net... edit: Sorry Carmelo, didn't notice until too late that you don't want your own server. Any particular reason why? The new stuff is way slick compared to what you went through five years ago... George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 19:34:09 ·
carmelo join:2009-11-05 Canada	Note that you need at least two providers
	to forum · permalink · · 2009-11-05 19:41:53 ·
carmelo join:2009-11-05 Canada	reply to GNca George Thanks George. Could you elaborate more on why I would need two providers? I don't mind managing things myself, but my experience with Asterisk back 4-5 yr ago was pretty painful (hopefully things got better in the meantime) and the thought of having to manage the whole thing on my own for a small business scares me -which is why I am looking for a canned solution-
	to forum · permalink · · 2009-11-05 19:47:42 ·
twizlar I dont think so. Premium join:2003-12-24 Brantford, ON	reply to carmelo Not sure why you would need two providers, you can do all of that with one. We use voip.ms and they have been great. -- Broadline Networks Inc.
	to forum · permalink · · 2009-11-05 20:00:43 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON	reply to carmelo Neither provider will be quite as good as a Bell landline. We have inbound 800 numbers for each team member through Voice Networks. We have outbound trunks through Voice Networks and VoipStreet. We hit a minor glitch this morning with VoipStreet outbound, just took a few seconds to switch to Voice Networks as the priority outbound route. Getting around an outage, at least outbound, is just that simple. We also have DIDs from VoipStreet, and publish all of them for people that really, really have to reach us. In addition, because we use Lylix in NY, way off-net, if we have a network glitch internally our calls just divert to our cell phones. This is possible with managed solutions, but not quite as simple as configuring your own. Please do a search for the Orgasmatron Google Voice edition for PBX-in-a-Flash and see what you think. Its pretty cool. I particularly want the Asteridex to Outlook Contacts integration, but there are many other parts of the build that are also pretty darn cool. Add in the VMware VM machine kernel (takes a couple of minutes to add, well documented) and your PBX will run just fine in a VM environment. That's important to us... George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 20:00:44 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON	reply to twizlar said by twizlar : Not sure why you would need two providers, you can do all of that with one. We use voip.ms and they have been great. Because I don't want our phone system to be down, ever, and VoIP providers typically don't provide the same 99.997% reliabilty that you would get directly from Bell. This is not me being paranoid btw, its generally accepted practice in the VoIP world because its so cheap to do it right. George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 20:03:42 ·
twizlar I dont think so. Premium join:2003-12-24 Brantford, ON	I've really never had issues, I just have multiple servers set as backups. -- Broadline Networks Inc.
	to forum · permalink · · 2009-11-05 20:09:05 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON	Horses for courses, you pays yer money and takes yer chances. I don't worry that I'm too paranoid, I worry that I'm not paranoid enough. George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 20:27:07 ·
Vomio join:2008-04-01 ·odynet	reply to carmelo I use voip.ms for outbound long distance with POTS inbound and for local outbound. The POTS is also there for long distance fall over although so far it has not been necessary. My needs are pretty simple so I just use a /Sipura/Linksys/Cisco SP-3102 SIP box, it will also handle a local incoming POTS calls and then redirect outgoing over voip for long distance, handy in some circumstances such as cell phone use. Pretty much any non-skype softphone will work, I use Twinkle on my netbook. voip.ms has a fair number of features and offers Calgary DID. Give them a look, so far I've been happy.
	to forum · permalink · · 2009-11-05 20:29:48 ·
carmelo join:2009-11-05 Canada	reply to GNca George I guess it all depends on the nature of your business and how costly it would be do be down. How hard is it to setup and manage a hosted pbx using, say, Trixbox? And who offers voicemail transcription?
	to forum · permalink · · 2009-11-05 20:31:54 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON 1 edit	said by carmelo : I guess it all depends on the nature of your business and how costly it would be do be down. How hard is it to setup and manage a hosted pbx using, say, Trixbox? And who offers voicemail transcription? edit: Its hard to be a credible service provider when you can't even keep your phones running properly, so we take the phone sysem pretty seriously. It is very inexpensive to do it right compared to half-assed, so why not do it right? /edit I'm creeping away from Trixbox as its becoming more and more proprietary which I really don't like. My current suggestion is PBX in a Flash, which gives the same or better functionality and much better community support. Every iteration you will find is pretty much Asterisk with FreePBX underlying, so the quality of community support is important. Nerdvittles is a great source of add-ons for free and is incredibly well supported. PIAF is their target distro ever since Trixbox went quasi-proprietary. You can get pretty good transcription from PIAF. We are expecting to use it in our next iteration. Note that we are 100% VoIP, no landlines at all, so that perhaps explains why we want multiple providers for everything. When it only costs another five bucks a month, why wouldn't you? George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 20:38:24 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON	BTW, please forgive me for the pedantry. I've always been a big believer in business continuity, having worked in several industries where its a really big deal. I would always rather have redundant systems than deal with an unexpected outage. Saying that, I was at a demo this afternoon put on by a partner that blew me away. They recreated a real life disaster that happened a few weeks ago at a big Muni client. They showed a live iteration of bringing back a running implementation of some custom Muni software running on top of Oracle Financials for a Municipal customer, with the backup being restored and then operating from servers at a remote datacentre at 151 Front. They got the backup up and running, with the Muni clients all logged in and operating in less than 20 minutes. You had to see it to believe it. Holy Hotsite Batman, this is the real deal. George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-05 20:57:15 ·
MaynardKrebs Premium join:2009-06-17	said by GNca George : BTW, please forgive me for the pedantry. I've always been a big believer in business continuity, having worked in several industries where its a really big deal. I would always rather have redundant systems than deal with an unexpected outage. Saying that, I was at a demo this afternoon put on by a partner that blew me away. They recreated a real life disaster that happened a few weeks ago at a big Muni client. They showed a live iteration of bringing back a running implementation of some custom Muni software running on top of Oracle Financials for a Municipal customer, with the backup being restored and then operating from servers at a remote datacentre at 151 Front. They got the backup up and running, with the Muni clients all logged in and operating in less than 20 minutes. You had to see it to believe it. Holy Hotsite Batman, this is the real deal. George Not really..... this is the real deal..... »www.youtube.com/watch?v=qMCHpUtJnEI Pretty much out-of-the-box OpenVMS recovered in 13 seconds. With appropriate tuning of various system parameters, it could recover in about 1-2 seconds and not lose any pending transactions while doing so. When the World Trade Center went down, the world's leading bond broker at the time was running its OTLP system on a VMS platform. Until the power and comms lines were severed as the building fell, the systems in WTC continued to process. Once the lines were cut, identical clustered systems in Connecticut and London England completed the transactions that were left incomplete by the then crushed NYC systems. No data was lost. This company did transactions valued in trillions of dollars per day, and still exists thanks to the fact that it was able to continue processing its business on 9/11 that way using OpenVMS. OpenVMS clusters can span thousands of miles between nodes and support up to 96 physical boxes with up to 128 cpu's per box, and up to 4 cores per cpu. That's nearly 50,000 cpu's of clustered power operating as one system and managed by one console. I may be a little low in these numbers - I have not looked at the most recent system configuration constraints in about a year or so. Having built and managed applications processing billions of dollars of value each day, I can tell you that OpenVMS is THE king of the hill - even better than IBM zOS and Tandem for absolute reliability in a disaster tolerant way. BTW, there are free OpenVMS and compiler licenses available to hobbyists. You just need an appropriate box to run it on - an old DEC Alpha box can be had for not too much money on eBay, ditto for an early generation HP Itanium box.
	to forum · permalink · · 2009-11-05 23:41:59 ·
jfmezei Premium join:2007-01-03 Beaconsfield, QC	Maynard, I dont think there are any VoIP servers that run on VMS. And since HP is determined to cater only to the remaining installed base, nobody who doesn't yet have VMS should deploy new VMS stuff because its future is not bright.
	to forum · permalink · · 2009-11-06 01:09:28 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON 1 edit	reply to MaynardKrebs Sure, you can spend a zillion dollars and get remarkable results for big enterprises. OTOH, the failover solution I described brings very good DR to the SMB space, and only costs about $1000 per month all up including required hardware. That might pay for one empty shipping crate in an OpenVMS solution. It works perfectly in a VMware ESX 4 environment which is a pretty popular platform these days while requiring no more than a DSL line to keep the synchronization up. Last solution I used from EMC required a full DS-3 for synchronization of a similar amount of data, and that cost $27k per month at the time. George -- Tough Broadband for a Tough Crowd! GorillaNET.ca - 10Mbits to your desk, coming soon.
	to forum · permalink · · 2009-11-06 07:56:52 ·
jfmezei Premium join:2007-01-03 Beaconsfield, QC ·ELECTRONICBOX 1 edit	The trick to quick disaster recovery scheme lies in a distributed lock manager. While, on the surface, it doesn't seem related, it is an enabling technology that makes things possible. With VMS for instance, multiple nodes can read/write to disks directly because locks are common. If node A locks J.Doe's record, then Node B can't write to it until the lock is released. This allows multiple nodes to be active at the same time. When when a node drops, other nodes can continue without needing to rebuild anything. And the distributed lock manager also allows neat stuff because when a node drops, locks onwed by that node are automatically zapped. Here is an example: node A has 10.0.0.10 node B has 10.0.0.11 Both nodes have a "cluster" IP address of 10.0.0.50 First node that boots tries to get a lock for 10.0.0.50 and succeeds, so it ifconfigs the alias for 10.0.0.50 an send a gratuitous ARP to be known. Second node tried to get 10.0.0.50 and is refused but is put in the queue to get it when it is released. When node A goes down, the lock for 10.0.0.50 evaporates, so node B is granted the lock and then creates the ifcinfig alias for 10.0.0.50 andf sends the gratuitous arp to announce itself as the new owner of 10.0.0.50 and from then on, traffic flows to it. This can happen in about a second. VMS has this. So did Tru64. IBM's AIX has it. And Digital, as part of its suicide, gave that technology away to both Oracle and Microsoft. Oracle has implemented parts of it in it main database product, but I am not sure if it is a separate option or if it comes with the base database engine. HP inherited both VMS and True64 and abandonned plans to equip HP-UX with a dstributed lock manager. Microsoft never did anything with it. The problem with a DB-only solution is that the applications and OS don't benefit from this. Distributed lock manager allows wonderful things between applicartions to coordinate control of certain responsabilities for instance.
	to forum · permalink · · 2009-11-06 17:58:28 ·
GNca George GorillaNET Premium join:2008-07-12 Minden, ON	Ah the memories. We implemented SAP on Tru64 on Alpha when I was at Citrix.
	to forum · permalink · · 2009-11-06 18:10:14 ·
Sapper @clearnet.com	reply to carmelo Take a look at »www.onekingtelecom.com/ and their line of products.
	to forum · permalink · 2009-11-08 12:37:59 ·
freejazz_RdJ join:2009-03-10	reply to carmelo I also wanted to chime in about hosted IP PBX services versus IP trunking. In the latter, you run your own IP PBX, including hardware, software and connectivity to reach your trunking provider. In the hosted solution, you only run your local routing/switching and phones. The service provider hosts the server and provides all the inter and intra LATA connectivity and switching. This is an attractive solution for a few reasons, mainly lower capital cost, higher availability and simplicity. Most of these solutions have very high redundancy within 1 PoP and have several PoPs. I have seen clients with several providers, but frankly Allstream has been quite common surprisingly. In the hosted solution, you do need to have a good router onsite to do QoS and preferably some sort of edge router and WAN link redundancy. I usually see two types of WAN links (optical ethernet + T1/DSL) and a pair of routers using VRRP or HSRP.
	to forum · permalink · · 2009-11-08 21:59:44 ·


Wednesday, 25-Nov 23:19:50	Terms of Use \| Privacy Policy \| Hosting by www.nac.net - DSL,Hosting & Co-lo \| feedback \| contact over 10 years online! © 1999-2009 dslreports.com. page compression OFF