Service failure but Adapter's lights normal. Grrr.

[stevech]

9:15PM PT, 1/22/11

Inbound calls forward to my backup number.
Outbound call attempts: dial tone but fast busy after any number is dialed.Tried several different area codes.

Power-cycled the GS adapter. Seemingly OK now.

Inability to receive calls though the lights on the adapter are normal.
I did not know, could not know, that service was out.
Most displeased.

[stevech]

What precludes a VoIP provider from being proactive when a subscriber's ATA has not re-registered for x amount of time?
My ATA (a Grandstream) has run for a long time with the same router here, same configuration, etc. Suddenly, I get fast busy on call attempts. This may mean that the ATA is not in session with the server.

My expectation:
VoIPo has ATAs arranged to re-register every x hours. The default arrangment should be for VoIPo to note overdue re-registrations and take action, be that an automated email/SMS or a phone call to the backup number, etc.

Instead, apparently no action is taken, per the rather lame response from support. According to my router, my internet connection has been "up" for 20 days. And my PCs have had no glitches. No DHCP faults.

[VOIPoTim]

I looked up your ticket and I believe our system handled everything as designed by sending your calls to your failover number when we couldn't get them through to you.

While we do monitor our own network, we don't provide proactive monitoring of end user connections/management.

I understand that you are frustrated and wish that we would monitor things outside of our own network. Unfortunately that's not practical from a technological or economical standpoint and I don't know of any residential providers doing it.

Even ISPs don't do this level of monitoring for their customers to detect is a modem loses sync, etc. It's just not practical and we can only monitor our own network.

It would be extremely difficult and costly to design a system that could accurately monitor end user networks primarily because of all the different network setups, routers, and NAT situations out there. There's also the situation where sometimes users unplug their devices, sometimes they travel with them, etc. Sometimes ISPs renew IP addresses and they could be down momentarily, etc. There are hundreds of potential situations that could generate false positives. The support/customer service burden alone from those alerts would require more staff that our entire service does because of people complaining about false positives or special situations it didn't detect.

Our solution (and the one that is used by virtually every provider) is to provide failover forwarding where if we can't get a call to you for any reason, the call can be forwarded to an alternate number or send to voicemail. In your case, the failover worked fine so our system was working as designed.

Again, I understand you may want more, but there are limitations on what providers can do.

[stevech]

The ATA >>IS<< part of your network. The T&Cs say you own it!

Simply: why can't you monitor YOUR ATA's for re-registration faults and email/SMS so we know something's wrong?

Not asking you to monitor my router, just monitor overdue re-registrations - that's in your servers. Why not? If a customer habitually turns off their ATA, you can auto opt-out them.

[VOIPoTim]

The ATA >>IS<< part of your network. The T&Cs say you own it!

Simply: why can't you monitor YOUR ATA's for re-registration faults and email/SMS so we know something's wrong?

Not asking you to monitor my router, just monitor overdue re-registrations - that's in your servers. Why not? If a customer habitually turns off their ATA, you can auto opt-out them.

We do own the adapters, but when they are in possession of a customer, behind a customer-controlled router, and running on a third-party ISP, they are out of our control and I don't consider them to be on our network from a monitoring standpoint.

It's just not nearly as easy as it sounds for us to do this it sounds....both from a technological standpoint or an economic standpoint for the reasons I mentioned in my last post.

We only monitor our network and do not monitor anything as far as customer networks or devices in customer possession.

Again, I understand where you are coming from but I do not agree that we should monitor and proactively manage things not part of our network. I'm not debating that it would be useful for customers in an ideal world, but it's not practical for us to do it.

It's not something we offer or intend to offer at this time. Other companies agree with me that it's not practical since I don't know of any other residential providers doing it.

We offer failover for situations where there are issues on the customer side. Our failover solution worked perfectly in this case.

[stevech]

Please re-read. I'm asking why VoIPo does not/can not monitor VoIPo's registration servers. Not equipment at my location.

If a re-registration is overdue from a subscriber's ATA, way overdue, then send an email, SMS, or robo-call to the back-up number to warn that there may be (probably) no service.

OK. end of chatter.

[VOIPoTim]

Please re-read. I'm asking why VoIPo does not/can not monitor VoIPo's registration servers. Not equipment at my location.

If a re-registration is overdue from a subscriber's ATA, way overdue, then send an email, SMS, or robo-call to the back-up number to warn that there may be (probably) no service.

OK. end of chatter.

My responses encompass things from both your post and the ticket where you were asking about it.

No matter how we look at it, the bottom line is that we don't offer monitoring for anything but our own network. It's not feasible for us to monitor things on the customer side or actions devices outside of our network for the reasons outlined before.

While I understand you may want us to monitor more, I'm sure that if you knew all the behind the scenes things things that would go into doing that, you'd understand that it'd not practical. From a technological to financial to support perspective, it's not realistic to monitor anything but our own network.

Even ISPs are not providing that level of monitoring to detect if a modem is offline and their modems are not behind customer provided routers that we have no information on.

We provide failover and the failover for situations where there is some kind of failure on the customer side or a call can't get through their router. The failover worked perfectly for you so our system is working as designed.

[stevech]

This discussion isn't going well. Don't want an argumentative tone.
Three times I've asked, apparently without clarity:
Why can't VoIPo monitor VoIPo's own servers to detect overdue re-registration of customers' ATAs? And notify customer by email/SMS.

You keep responding that it's impractical for VoIPo to monitor customer premises equipment.

[VOIPoTim]

This discussion isn't going well. Don't want an argumentative tone.
Three times I've asked, apparently without clarity:
Why can't VoIPo monitor VoIPo's own servers to detect overdue re-registration of customers' ATAs? And notify customer by email/SMS.

You keep responding that it's impractical for VoIPo to monitor customer premises equipment.

I'm just referring back to my previous reply:

--

Even ISPs don't do this level of monitoring for their customers to detect is a modem loses sync, etc. It's just not practical and we can only monitor our own network.

It would be extremely difficult and costly to design a system that could accurately monitor end user networks primarily because of all the different network setups, routers, and NAT situations out there. There's also the situation where sometimes users unplug their devices, sometimes they travel with them, etc. Sometimes ISPs renew IP addresses and they could be down momentarily, etc. There are hundreds of potential situations that could generate false positives. The support/customer service burden alone from those alerts would require more staff that our entire service does because of people complaining about false positives or special situations it didn't detect.
---

Doing what you're asking is not by any means easy. If it were all providers would be doing it, ISPs would be doing it to monitor the loss of a modem connection, etc.

I don't know of any providers providing that level of monitoring because it's just not practical and would cost a small fortune to design something that could accurately do it while accounting for all the false positives, etc. Outside of that, it would require a massive increase in support staff due to all the new inquires that would come in and the irate people upset over alerts for false positives, missing special situations, etc.

Even the "Devices" link in the vPanel that shows the registration information generates hundreds of tickets a month with people upset over a variety of things like it not showing their virtual number so they don't know if it's up or not or they think the time listed for registration renewal means they will be billed when it re-registers or they are irate that it shows their IP address there because that's "not secure".

Some just don't understand it and that's fine. The problem is that a huge % of customers become extremely hostile over every little thing.

As an example, we had someone recently get a voicemail notification for a blank voicemail because the caller hung up and they filed complaints with the AG, BBB, FCC, etc right after contacting us about it because we wasted an SMS on their cell phone and clogged their e-mail box with a notification for a voicemail that didn't have anything it it and we should have suppressed the alert for it.


Not everyone does things like that, but enough customers do that it influences how we deal with every situation. People go ballistic over the smallest things every day and we unfortunately can't take common sense as a given. Again, it's a small % of our customer base but enough to have a huge impact.

Everything is complex with a large customer base and we have to account for that. To deal with something that's guaranteed to generate false positives and be so complex to monitor is going to cause 10x the effect.

Trust me, it's 1000x more complicated to do than it sounds especially when we can't count on all customers having common sense when utilizing tools like that. You may understand it and be able to use common sense with it and think we can opt someone out of it, but that doesn't mean everyone else will have the same common sense and we'd have to deal with the fallout from everyone.

It's nothing personal and I understand why you'd want something like this, but it's not practical to do. I'm not saying it's not practical based on speculation..I'm saying it based on experience and seeing the daily interactions with some of our customers and having an understanding of how many of our customers react to different things.

[TFlap]

...As an example, we had someone recently get a voicemail notification for a blank voicemail because the caller hung up and they filed complaints with the AG, BBB, FCC, etc right after contacting us about it because we wasted an SMS on their cell phone and clogged their e-mail box with a notification for a voicemail that didn't have anything it it and we should have suppressed the alert for it...
That's crazy!

[VOIPoTim]

That's crazy!

Just one of many examples. We deal with people doing similar things every day.

While in that example the complaints had no merit, we still have to respond and defend ourselves because unfortunately in today's "big bad business" society everyone is anti-business and pro-consumer even in situations where the consumer is wrong.

From complaints like that to showing up at our door slamming an ATA down to making threats like "I was on vacation and didn't use my phone while gone so give me a free month or I'll do a chargeback" we deal with it every day. I'm not sure if it's the stress of the economy in general or what but we've been seeing more and more people just going ballistic over minor things the last year or so.

The point is that doing something like outlined in this post that would generate false positives would create more situations like this where people would literally go ballistic and snap over simple things above and beyond the technical limitations of doing it. If they were opted out, we'd be the bad guy for taking them out and they'd want a refund for not fully delivering on something we advertised. If they were left in, they'd have false positives and complain about it.

It's more reasonable for us to monitor our own network because we understand the realities of monitoring and won't get upset over getting too many alerts, etc.

In anything we do, we have to find the right balance of providing functionality that people want while walking on eggshells around a % of our customer base with it. It's a balancing act and unfortunately there are some things like this that are harder to deal with than others.

That's why VOIPo moved to California....being 2 miles from the ocean has a calming effect that's needed. :)

[rainsux]

> I'm not sure if it's the stress of the economy in general or what but we've been seeing
> more and more people just going ballistic over minor things the last year or so.

Ditto.

[stevech]

I started the thread; I'll make the final post and hope the moderator locks this.

VoIPo avoided a direct answer to a simple question on why they can't monitor overdue re-registrations on their servers, not the customers' equipment. The answer must be "it costs".

And let me also add: I've had VoIPo for enough time to say, other than the AWOL ATA, the service is quite reliable.

END OF THREAD

[VOIPoTim]

I started the thread; I'll make the final post and hope the moderator locks this.

VoIPo avoided a direct answer to a simple question on why they can't monitor overdue re-registrations on their servers, not the customers' equipment. The answer must be "it costs".

And let me also add: I've had VoIPo for enough time to say, other than the AWOL ATA, the service is quite reliable.

END OF THREAD

Sorry you feel that we didn't answer your question. I feel that I've provided very detailed answers outlining our stance on this and explaining that it's not practical to do because of the registration action happens outside of our network and there would be too many false positives, etc as customers unplug equipment, move it around, etc. It would also require an astronomically complex system to handle it...there are lots of technicalities that would have to be accounted for to accurately monitor.

We're extremely open to feedback and have implemented a ton of stuff that people have suggested, but this is just one of those things that we don't feel we can in a practical way.

Everything is not as simple as it seems. I do understand that without knowing the specifics of everything required it can seem simple, but this is not something simple for us or any other provider to do by any means which is why no one does it.

I'll close the thread like you've requested at this time.