When VPN Solutions Aren't All That
I can really relate to Mike Fratto's Jan. 21, 2002 article,
"Our Long and Winding Road to VPN", and I'd like to share some of the reasons why the company I work for switched from an Internet VPN to a frame network.
My organization acquired a company that used an Internet VPN for communication between branch offices. One company provided all the Internet connections for each office, and unfortunately that company was Epoch (it eventually went broke and left some facilities with no communications whatsoever). Meanwhile, we had a point-to-point, leased-line hub configuration in place within our parent company network.
The carrier kept the traffic on its network as much as it could (cold potato routing), but it was always possible for network traffic to be routed any which way across the Internet. Not all our problems originated with our service provider. There were many times I had to call an Internet service provider to [complain about] a bad router or a routing problem (which I had discovered through a trace-route command). Sometimes a provider would tell me there was nothing wrong because, in fact, there was nothing wrong when [routing was] tested from inside its network--but traffic from outside its network was indeed being routed improperly.
Another problem with the VPN solution was that some issues involved multiple vendors--all of which liked to point fingers at each other. The phone companies would always say there was no problem or that the problem had to do with customer premise equipment and then, magically, the circuit would start working.
We eventually switched to an AT&T managed frame relay solution so our internal support wouldn't be stuck trying to figure out hardware failures on private point-to-point networks or in Internet VPN networks.
It all comes down to what you're willing to pay for and how much in-house support you're willing to provide. If you've got the manpower to manage [the system] and troubleshoot problems (and you don't mind some downtime), you don't need a premium service. For us, having AT&T know there's a problem on the circuit before we do--and getting to work on it right away--is a good thing.
Dean Ott
Technical Project Leader
Gaylord Container
djott@a1mail.gylrd.com
 |
Leaving Layers 5 & 6 Behind
I agree wholeheartedly with Peter Morrissey's opinions regarding the current usefulness of Layers 5 & 6 ("Looking to the Future by Ignoring the Past," Jan. 21, 2002). I believe that those two Layers were always destined to become a function of the desktop operating system.
But we must take into account that at the time the model was created, Layers 5 and 6 were necessary to allow logical placement within the networking structure of IBM's SNA LU6.2 and X.25's PAD function.
X.25's PAD may have been excluded from the model safely, but had the ISO excluded IBM's SNA LU6.2, the model immediately would have been discredited. Remember, at that time IBM still controlled probably 75 percent of the corporate desktops in America (a conservative estimate).
Good riddance to half of the top four Layers. Now let's work on combining Layers 3 and 4.
On a side note: While it's not palindromic like the ISO OSI, the FDDI TP-PMD has always been a rhythmic favorite of mine.
Bill Hawk
Senior Engineer
Navigant Consulting
bhawk@navigantconsulting.com
 |
Third Time's a Charm
I hope Doug Barney's message on getting out of the recession ("A Recession? Buy Your Way Out," Jan. 21, 2002) gets out to people. I've been downsized [out of a job] again, this time by a telecom. But, really, it's no big deal: This is already my third downsizing. Personally, I like to think I'm on the bleeding edge of some new professional working style.
I've come to the conclusion that the only thing most middle managers think about is cost-cutting. I guess that's just their job: Figuring out ways to do absolutely everything for the minimum amount of money.
But being thought of as overhead is getting pretty old, and it seems that most IT jobs are considered just that. I'm now doing some work for a few small businesses, and I doubt that I'll ever go back to a large corporation.
New ideas and revenue generation are the keys to prosperity, just like Barney outlines.
Rob Reilly
Consultant
PKR Products and Services
Rreilly@cfl.rr.com
 |
HAL's in the House
Mike DeMaria's article "Home Smart Home" (Jan. 7, 2002) is very good. I'm an electrical contractor in Madison, Wis., for a company that's been in business since 1962. That's quite a long time to be in the same business, but the interesting part is that it has allowed me to see all the ups and downs of home automation.
Smart houses died because of exorbitant prices. Would you pay $20,000 to $75,000 to get your house wired so that you could obtain automated appliances for some unknown price at some unknown time down the road? Of course not. However, these days people are putting LAN networks, phone systems, Cat 5/satellite TV systems and audio systems into their homes. We've even installed a few fancy lighting systems in the past few years. The good part is that now there are many quality manufacturers coming out with one-box systems to make the wiring look as neat as a pin.
Steven Mylrea
Electrical Contractor
Academy Electric
steve@academyelectric.com
 |
Correction
In the print version of "NetIQ Shows Analysis Smarts" (Jan. 21, 2002), the NetIQ product name should be WebTrends Reporting Center.