Dvorak's article points out the problems of relying on a single server (or service) in a modern IT infrastructure.
The impetus for the article was Microsoft's WGA servers going down for 19 hours last weekend, but those of you tempted to rely on 'free as in beer' services might do a gut check (this applies to paid hosted services, too, of course.)
Skype's vaunted peer-to-peer technology didn't help them stay up - their outage lasted roughly the same period of time.
That's one reason why FreePBX® doesn't rely on an external service to run - whether to log you in or store your settings.
With Moore' Law at your back, why share? It is a valid question, which is sometimes met with one of the following answers:
1) offsite is supposed to be more reliable than on-site, because there are more qualified people making it their number one priority to keep it up
2) your team is distributed and they still rely on one service anyway
As the WGA and Skype outages show, these justifications sometimes fall down.
After all, at your office you can put in a backup system, if needed. It is hard to find a second service provider to be your backup.
When trixboxPro was released I set up a VM system and started poking around. While setting up a trunk I suddenly could not administer the system any longer. The "Administration" of a xPro system is hosted. Something happened, either a DDOS or server malfunction or whatever shut down the hosted portion. WGA, Skype and also trixboxPro have had recent outages. Certainly the trixboxPro outage is the least of them. It is illustrative of the point just the same.
I think there are useful applications to be hosted such as web sites and such. Real time communications such as VoIP systems… not so much.
Robert
"All men can see these tactics whereby I conquer, but what none can see is the strategy out of which victory is evolved."—Sun Tzu
I have been in the webbased biz for MANY MANY years, we have never had any lost data.
The uptime for a medical billing software has been up 99999 for YEARS.
Katrina came to town two years ago, many folks lost it all, we did not lose a thing; never did any of our systems go down, as we are real providers and the “whole racks data / VPS’s” are sent to a second data center in another state.
I got 7 days battery power and ten days of generator power, not many office have that
As the network admin for a few big companies; I can tell you my webbased network will be up for more then the local net’s.
I can see where a webbased app is a better fit for some folks; where as a local app is better for some others.
And please… who would compare MS to any other setups, we all know they engineer failure as a rule.
Hang on… seven days of batter power? I know this is going off topic but if you have a server room that can run for 168 hours on batteries alone, I would LOVE to see a pic of the kit being used!
At the end of the day, Asterisk was built around the notion of getting rid of the nickel & dime middle man and finally taking control of your telecom in a manner that was reasonable to manage unlike the legacy PBX systems, yet with most of the features & stability to match (i.e. v1.2 seems to hit that mark pretty well, other than echo in the Zaptel non-hardware-echo boards).
Not sure many have changed their minds about wanting to return to the annuity, charge for every handset (whether you use it much or not), offsite supported, dependency laden business model? If I remember correctly the “hosted” or “PBX as a service” model was tried for two decades without too much success by PacBell as the “Centrex” service.
Great idea admittedly, the easiest implementation I think there is (i.e. got a phone, make a call - instant PBX). Nevertheless, it never did that well for various reasons, the least of which being the above mentioned.
Just some thoughts. At the end of the day, our opinions mean little as people’s/businesses choices will ultimately make the final call as to who survives and who does not based on what they do/don’t want …
So it’s always good to keep the options supported, especially the ones that triggered the avalanche originally - the original Asterisk mission, purpose, focus.
ksDevGuy
P.S. Philipel & crew, you guys are doing an appreciated and good job of forwarding FreePBX from where it was left at the last transition. Thanks.
Yes but…
“two decades” ago you big office’s full of folks, now you have the “work from home / small sat offices”.
“two decades” ago you could not get fiber to your door, broadband was only for the bigger offices.
We are moving toward a more mobile office type of needs for folks; more companies are looking at reducing “floor space” by reducing the number of folks who work in a office building.
And as more and more folks are starting up new “at home jobs/ business’s” the “home office” may really be at home.
I just took apart a accounting firm, the seven employee’s all moved to a home office / or got office space else where.
The asterisk box is a remote hosted box; so they just took the IP phones with them.
Only one had a issue with the phone not working at her home, she wanted a headset anyway… so zoiper and 40.00 usb headset and now they are just like they where to the outside world.
Had this been a onsite box they would need to move the box and hope that the broadband at where ever they took it was good for VOIP.
Times… they are a changing…
Just look over the forums for the past year; more and more folks are looking at VPS (a host solution) and rack mounted servers for data center’s.
I guess we see it more than some; because of the hosting / rackspace we provide for asterisk.
a guick browse of the notes shows 7 or more request’s for VPS quotes last month and we have not offered that service to the public for over a year.
Am very aware of what the “forums” show, but there’s quite a bit more to that than face value to be sure. Being in software, I am aware of your article’s message and just a bit more.
Ultimately, it really doesn’t matter all that much. The actions of the FreePBX crew for example, are of great value.
As I said before, time will tell - and possibly not even all that much of it.