dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
2178
grand total
join:2005-10-26
Mississauga
·Fido
MikroTik RB750Gr3
MikroTik wAP AC
Panasonic KX-TGP500

grand total to rjwells

Member

to rjwells

Re: [Voip.ms] Toronto server down again ?

said by rjwells:

Gonna try sticking an ATA on what it thinks is the fixed line, configured to use a different service

You have a good chance of success I think. I did this with an SPA3102 with it's FXS port connected to an SPA2102.

Davesnothere
Change is NOT Necessarily Progress
Premium Member
join:2009-06-15
Canada

Davesnothere to rjwells

Premium Member

to rjwells
said by rjwells:

Dave, to take the redundancy question a step further, why stick with just voip.ms?

If your device can failover between SIP registrations, why not failover to a different provider entirely ? ....

 
VERRRRRY Interesting !

Sounds like that ought to work, and it just might be the most compelling reason to date to swap out my trusty ol' PAP2T in favour of one of these OBI devices, besides the fact that SOME OBI models allow configuring for more that 2 VoIP providers in general.
rjwells
join:2013-04-29

rjwells to grand total

Member

to grand total
said by grand total:

said by rjwells:

Gonna try sticking an ATA on what it thinks is the fixed line, configured to use a different service

You have a good chance of success I think. I did this with an SPA3102 with it's FXS port connected to an SPA2102.

If it works, then I'm going to try and get the 2nd device on a different internet connection. I can buy one of those little data sticks and have it go out over expensive mobile bandwidth. Not ideal for regular use, but perhaps a lifesaver when my wired connection is down.
JeanInNepean
join:2012-09-19
Grenoble, FR

JeanInNepean to H_T_R_N

Member

to H_T_R_N
On voip.ms, you can route your DID(s) to a ring group of SIP URIs that redirect your calls to other servers. If the POP for that DID is having networking problems, that may not be of much help though. But it will help in the case you're having registration issues.

As far as DNS SRV goes, this is something that can be setup on a local domain server easily.
taytong888
join:2005-06-20
Nepean, ON

taytong888 to grand total

Member

to grand total
said by grand total:

said by rjwells:

Gonna try sticking an ATA on what it thinks is the fixed line, configured to use a different service

You have a good chance of success I think. I did this with an SPA3102 with it's FXS port connected to an SPA2102.

Hello,

1. Once the secondary ATA (SPA-2102 or OBi100/110) is connected via RJ11 cable to the FXS (PHONE) port of SPA-3102, then exactly how do you configure the PSTN Tab of SPA-3102 to allow for other VOIP service provider(s)?

2. Then how do we configure the secondary ATA such that both devices work well together?

Thanks in advance for your help.
SCADAGeo
Premium Member
join:2012-11-08
N California

SCADAGeo to Gobe

Premium Member

to Gobe
said by buzz :

With its history of problems, why does anyone use the Toronto server at all? In fact, why does voip.ms keep the server running?

From what I recall, the Toronto servers were working well after the upgrade and tuning period, then folks migrated back to Toronto and Toronto2.

 
said by VexorgTR:

It's odd, I see reports of Voip.MS and how some users really like it.. but the "Server Down" posts show up here often. With all those servers, I guess there's always an alternative.

There are resource issues during peak usage periods at the POP I tested.

My guestimate from tests performed over the past 8 days (not Toronto), is that we are currently in a situation similar to 3rd quarter of 2012.

 
said by PX Eliezer1:

Some servers seem prone to trouble more often than others.

According to their own data, a very [stable] server seems to be Dallas.

Some other servers seem to have had repeated episodes of trouble. It would be prudent to eschew those particular servers, as one does with walking alone through a bad neighbourhood at night.

---

I hope that Voip.MS works through this, as we need good-quality independent VoSP.

+1

 
said by H_T_R_N:

said by PX Eliezer1:

However, it seems to me that there are various IP choices available before returning to POTS.

I am going to try a few but really with 10 sales persons on the phone all day, if they are not making sales calls they are not making sales.

With people constantly on the phone, you're in a (maybe not so) great position to note when peak resource issues first appear.

 
said by H_T_R_N:

I have been through several providers, 8x8, zingotel, callcentric, VOIP.ms and a couple home grown asterisk servers. They all left me wanting. I am more than likely going to go back to POTS with an asterisk server and digium cards. That was by far the most reliable package we had. It wasn't until the cost savings were sought that we dumped the POTS for a VOIPSP.

said by VexorgTR:

there are a few copper fail over's in place. In the even the Sip trunk pukes out, we go from 12 voip channels down to two copper... but that hasn't yet happened.

said by mozerd:

When it comes to reliability it all depends on how VoIP is set up -- VoIP redundancy is not only the responsibility of the provider its also the responsibility of the adopter.

I like to have multiple providers, with fallback to cellular, just in case I encounter a worst case scenario.

 
said by H_T_R_N:

For outgoing I agree. There is no way to set up redundant providers for incoming calls on the same number. If there were, then that would be a whole nother ball game.

said by rjwells:

You can do something similar for inbound. You can configure your voip.ms account to failover to a SIP URI if it cannot connect to your device. Set it up so that it forwards your call off to your CallCentric or Anveo account, in the hopes that they can deliver the call to your device instead.

This obviously requires the voip.ms service is working well enough that it can bounce the call off, and you MAY wind up paying extra for having two different services process your call -- but wouldn't you rather have your phone ringing, even if it cost a little more?

And you may NOT have to pay more -- CallCentric in particular makes a great fallback because they do not charge you for calls that from their point of view don't involve the PSTN.

said by JeanInNepean:

On voip.ms, you can route your DID(s) to a ring group of SIP URIs that redirect your calls to other servers. If the POP for that DID is having networking problems, that may not be of much help though. But it will help in the case you're having registration issues.

As far as DNS SRV goes, this is something that can be setup on a local domain server easily.

Fortunately, the POP that I was testing has _not_ reached the stage where server resources are so severely stressed that calls are actually lost.

The creation of a ring group may be one way to work around the issue, and may yield both the reliability and cost savings desired.

The ring group can ring your extension, as well as forward to a SIP URI at another provider.
said by VoIP.MS Wiki :
Dialing out a SIP URI or using a SIP URI as forwarding does not generate cost(Outgoing call)

 
If the other provider has free inbound sip (CallCentric account still active?), there is no additional monetary cost.

The ring group can also ring a PSTN number, but at the prevailing outbound rate (if answered via PSTN).

Comparison of the CDRs will also yield data points about the percentage of calls that didn't directly ring the extensions, and, eventually, a temporal pattern that may be used in a predictive model.

 
said by N9MD:

Question:  What was Samuel Clemens pen name?

Answer:    BIC   

You got me with that one!

 
said by taytong888:

What can VOIP.MS practically do to improve reliability, therefore retain/attract customers?

said by PX Eliezer1:

I'm sure that Voip.MS [is] looking at the mid-range and long-term issues, not just putting out fires.

Experience is a great teacher, and I hope the valuable lessons learned from the 4th quarter of 2012 and the 1st quarter of 2013 will be applied.
grand total
join:2005-10-26
Mississauga
·Fido
MikroTik RB750Gr3
MikroTik wAP AC
Panasonic KX-TGP500

grand total to taytong888

Member

to taytong888
said by taytong888:

1. Once the secondary ATA (SPA-2102 or OBi100/110) is connected via RJ11 cable to the FXS (PHONE) port of SPA-3102, then exactly how do you configure the PSTN Tab of SPA-3102 to allow for other VOIP service provider(s)?

2. Then how do we configure the secondary ATA such that both devices work well together?

I'm not sure I understand what you are asking.

What I had was a SPA3102 to which I connected my phone. The VoIP side of the SPA3102 was configured to allow me to call or be called via provider A. The FXS port of the SPA3102 was then connected to a phone port of the SPA2102. The SPA2102 was configured to allow me to call or be called via provider B. The dial plan of the SPA3102 took care of directing outbound calls to the appropriate provider, and inbound calls from either provider rang my phone.
taytong888
join:2005-06-20
Nepean, ON

taytong888

Member

Hello,

Originally I wasn't sure how your setup works and what exactly software configuration entails. Now I understand everything thanks to your clear explanation.

In my case I have SPA-3102 as main ATA subscribed to voip.ms. I also got a spare OBi110 that could be used the same way as your SPA-2102. I have not done anything with the dial plan that originally comes with SPA-3102.

Again, thank you for your help.
rjwells
join:2013-04-29

2 edits

1 recommendation

rjwells to SCADAGeo

Member

to SCADAGeo
"Experience is a great teacher, and I hope the valuable lessons learned from the 4th quarter of 2012 and the 1st quarter of 2013 will be applied."

Yes that's true, and I think everyone understands that there are problems from time to time in any business. I have worked in businesses that deliver a (not really similar) service as well, and things have not always gone smoothly for us either. Having suffered through some tough problems and come out the other side, the advice I have to give voip.ms is this:

What would build up enormous trust, confidence, and respect from customers would be a clear and brutally honest explanation of what has gone wrong, what is being done about it, and when we can expect that it will be done.

At some level we've all hired voip.ms (and other firms) to do a job for us, running a voip service for us, and I think all anybody REALLY wants to know is that a competent team is on the job and dealing with the issues. And it's a lot easier to believe in that when there is a measure of transparency.

Maybe not all the time, but certainly in the places I've worked we found that it helped customers trust in us when things didn't go well. When everything is going well of course it's not necessary.

Anyway it's my 2c, and here in Canada now that we have eliminated the penny 2c rounds down to zero. So voip.ms can decide for themselves how they want to run their firm, just passing that along from my own experience in case it's useful.
pcunite
join:2010-04-10

pcunite to Gobe

Member

to Gobe
I've been using voip.ms for three years (maybe four?) and have found them to be excellent. However, I have really stable Internet service and a router to do Qos on my end. As a previous poster noted, we (society at large) need a good voice provider. I wish them well when I can and I find it disingenuous for someone who has had issue with VoIP in general to call out voip.ms on anything.

The network Voip.ms has in place is most likely not a $5million dollar setup. They have servers in racks at various places around the country and those servers die or the internet switches (that they share with everyone else in the coloco rack) goes down just like it does for any other business.

Until VoIP providers can have their own huge data centers in three places in the USA what can you seriously expect? Only Microsoft, Google, or Amazon has the infrastructure to do what some of you think you're getting with Appia or whoever writes your check.
PX Eliezer1
Premium Member
join:2013-03-10
Zubrowka USA

PX Eliezer1

Premium Member

Re: [Voip.ms]

said by pcunite:

They have servers in racks at various places around the country and those servers die or the internet switches (that they share with everyone else in the coloco rack) goes down just like it does for any other business.

Until VoIP providers can have their own huge data centers in three places in the USA what can you seriously expect? Only Microsoft, Google, or Amazon has the infrastructure to do what some of you think you're getting with Appia or whoever writes your check.

Voip.MS are a good company and I wish them the best---because we need a variety of strong independent VoIP providers.

-----------------

I disagree with what you said as your main point, which seems to be that going down is inevitable.

[Of course, SOME down time is inevitable. It happens to POTS too. The question is how often it happens, and the root causes].

Now, let's consider Anveo, which is a relatively new company at least in comparison to Voip.MS and CallCentric.

As I understand it, Anveo has 3 major POP's, and AFAIK if one goes down, automatic failover occurs. If that is true then that would be something for Voip.MS to work towards.

-----

An approach in the other direction is used by CallCentric, which manages all of its own equipment with its own personnel, not relying on anyone else's colocation rack.

This system served them very well for a long time, but as we all know the disadvantage of this occurred with the Sandy disaster.

So while redundancy [is] important---and CC must continue its efforts to redundantly enhance redundancy (power, data, geographic)---I would observe that there are also big advantages to running your own equipment, not depending on strangers.

-----

Finally, your post also suggests that if a VoIP provider's server is down, it is because of a broader problem in the colocation center.

It seems to me that this may be the case sometimes, but not all of the time.

So---if problems are due to a repeatedly bad [server] in terms of hardware or software, then that server has to be removed from active duty.

And---as I have said before, if problems are due to a repeatedly subpar [colocation center] then that vendor has to be removed from active duty.

It's almost always going to be one or the other.

-----

I just don't think that we should accept that repeated down time is inevitable, from ANY provider.
Mango
Use DMZ and you get a kick in the dick.
Premium Member
join:2008-12-25
www.toao.net

Mango to pcunite

Premium Member

to pcunite

Re: [Voip.ms] Toronto server down again ?

said by pcunite:

Until VoIP providers can have their own huge data centers in three places in the USA what can you seriously expect?

I seriously expect my VoIP setup to be just like a POTS line, with better features.

You absolutely don't need to operate a $5M datacentre for your service to be reliable. You can outsource to a quality datacentre (like we do), and have very high reliability. A quick Google search for "colo sla" gave me a big list of datacentres that will reimburse you if your equipment isn't online 100% of the time.

I'm sure you're aware that VoIP devices are capable of routing calls through any number of SIP switches, in case one is unreachable. They can also be configured to accept calls from multiple sources. Logic dictates it would be unlikely for multiple pieces of geographically dispersed equipment to be down at the same time. So, one can achieve near-100% uptime with reliable geographically dispersed equipment, if the transition from one switch to the other is unnoticed by the user.

All this to say if you can tolerate a little downtime, that's fine. But some of us have WAF to think about or use VoIP for business purposes and need it to be reliable.
rjwells
join:2013-04-29

3 edits

1 recommendation

rjwells to Gobe

Member

to Gobe
If people are seriously evaluating the failover capabilities of different VOIP providers, this is something that can only be done with information from the VOIP companies. There is quite a lot that can go wrong with failover that cannot be detected from the outside.

Most people think it's good enough just to have POPs in multiple locations, but that is not good enough. There's quite a bit more that is required.

First, what mechanism is used to fail traffic over from one POP to another, and how fast does that mechanism work? Will it take your device an hour to migrate to the alternate location? Or will it happen in seconds? Does it rely on assumptions about the capabilities of the customer's device (e.g., DNS changes) or does it happen automatically without any action by the customer (e.g., BGP routes changes). You MIGHT be able to tell this as an outsider, depending on the mechanism being used to fail traffic over.

Second, there are capacity issues. Does the second site have the capacity to handle all of its own traffic, plus all of the failed sites traffic, during the failover scenario? Nobody but the company itself can tell you this, and very often the company doesn't know, or thinks it knows but doesn't (i.e., they have never actually tried it and have some unrealistic expectations about the capabilities of their hardware).

Netflix is an interesting company because they constantly test out their own failover capability, intentionally failing their servers from time to time to prove that they have the ability to fail traffic over to working servers without interrupting service to customers. They know it works, because they routinely prove it works.

Most companies don't do that. Most companies create a design they believe will failover, but don't find out that it won't work until there is an actual problem. Very often the failover will happen slowly, with traffic not migrating the way that it should, and then when it gets to the alternate, the alternate will be swamped with too much load.

I would be curious to know if any of the VOIP companies can release a statement on their failover capabilities, and on whether they have tested the capability, and how often they test it, and whether they tested it at peak load or off peak.
pcunite
join:2010-04-10

pcunite to Mango

Member

to Mango

said by Mango See Profile
I seriously expect my VoIP setup to be just like a POTS line, with better features.

Mango and others ...
I appreciate the help and knowledge represented here. I want 99.9% uptime too and know that it is obtainable.

However, POTS has had millions of federal dollars to help build out the infrastructure. Testing and proven techniques has been discovered with society at large for over 50 years. Where does anyone expect the money, staff, and infrastructure for tiny VoIP companies to come from when they've only been around for 10 years? From the sales and use of their offerings! Strange as it may sound, the owners of VoIP companies probably want to provide for their family and live well too. Sacrificing for all of us can only go so far.

A better system would be some type of peer-to-peer collaboration among VoIP companies. If they don't they can expect Google, Microsoft, and others to replace them in the next decade anyway.
PX Eliezer1
Premium Member
join:2013-03-10
Zubrowka USA

1 edit

1 recommendation

PX Eliezer1

Premium Member

said by pcunite:

However, POTS has had millions of federal dollars to help build out the infrastructure. Testing and proven techniques has been discovered with society at large for over 50 years. Where does anyone expect the money, staff, and infrastructure for tiny VoIP companies to come from when they've only been around for 10 years? From the sales and use of their offerings!

I appreciate your commentaries.

I think that you are confusing the POTS phone system with the Interstate Highway System.

POTS in the USA and Canada was very much a project of private business (ie capitalism), not the government.

The Federal Government provided money mainly as a customer---a big customer, to be sure, but a customer nevertheless.

Yes, it [is] true that Federal money was spent by the Rural Electrification Administration to expand rural telephone service, but that was not that much out of the total picture nationally, and that money mainly went to smaller telcos and co-ops, not to the original AT&T/BellSystem.

AT&T/BellSystem prior to its breakup provided home phone service to about 85 percent of the population, and was very much a private creature. I suppose that you could say the government benefited them by helping them be a monopoly, but in fact they would have been one in most specific geographic areas anyway.

Mr. Bell was a pretty tough businessman, vanquishing others such as Gray and Meucci. Bell's partners and successors would be even tougher.

This was demonstrated when AT&T in the 1890's started the construction of the national long distance network ("Long Lines") and they refused to connect any independent local telcos to it.

Up until about 1911 when the Feds put a stop to that, AT&T had bought out thousands of local telcos in this manner---giving them the choice between being bought out, or cut off. That's how the Bell System got to be so large.

Sort of like the Foundation Federation in Asimov's "Foundation's Edge".
said by pcunite:

However, POTS has had millions of federal dollars to help build out the infrastructure. Testing and proven techniques has been discovered with society at large for over 50 years. Where does anyone expect the money, staff, and infrastructure for tiny VoIP companies to come from when they've only been around for 10 years? From the sales and use of their offerings!

I think another point that you are missing is that VoIP providers DO NOT NEED huge amounts of their own infrastructure.

There is this thing called the Internet.

Customers are already paying other companies to bring them the internet, and those companies (mostly big ones like Comcast or Verizon) help to support the operation of the Internet itself (although to a big extent it's run like a co-op itself).

Now, getting back to your original point, while the POTS companies were NOT built with federal funds, the Internet largely WAS a Federal project.

I'm not talking about Al Gore, but rather about DARPA and the universities that worked alongside.

Anyway, the old idea that you need a big building to have a phone company is obsolete. The Internet has changed that and so much more.

A better system would be some type of peer-to-peer collaboration among VoIP companies. If they don't they can expect Google, Microsoft, and others to replace them in the next decade anyway.

Wrong.

The future belongs to the nimble and the competitive.

A diverse ecosystem needs diverse inhabitants.

Microsoft's Skype was innovative because it was started by some eastern Europeans as Skype. It's not innovative as part of the Microsoft Empire.

And as the Soviet Empire crumbled, so will Microsoft. I don't know if Microsoft will even be around in the next decade.

We studied that in Earth Science in 8th grade: Eutrophication is what happens to swamps.

Where went Burroughs Corporation?

Where went Sperry Gyro?

Smith-Corona?

Kodak?

If they don't they can expect Google, Microsoft, and others to replace them in the next decade anyway.

Neither company has shown itself to excel at customer service.

The VoIP providers do not need to be a huge size, just the right size.

Switzerland is a nice country, and it does NOT propose to take over the whole world. Maybe it is happy and prosperous precisely because that is NOT its goal.

The world has a place for the Russian Federation, it has one for Switzerland, and it even has one for the Principality of Andorra.

Similarly, it has a place for Skype, a place for Vonage, one for Voip.MS/CallCentric/Anveo, and one for the smallest operators.