"A Burglar Alarm watching over my Web Site” you say?
In terms of the challenges and opportunities that today's Internet
presents to business, we live in some very interesting times. There
is a Chinese curse: “May you live in interesting times.”
There are a lot of interesting things that an IT Manager must keep under
control. Things like:
- An e-commerce web-site that becomes mysteriously inoperative or unavailable.
- Email that doesn't get through, resulting in unhappy customers and frustrated staff.
- A company's web page that gets hacked and may now be serving up pornography.
- A denial of service attack that may now be underway.
- A company's domain name that may have been hi-jacked.
- A squirrel that may have chewed through the telco line a mile away.
- A CPU that may be quietly over-heating.
- A myriad of other problems that may slow down and even grind things to a halt.
These things happen. (If you need convincing, read the newspapers.) When
they happen, it is the IT Manager's responsibility to put Humpty Dumpty back together again.
Once upon a time
| |
Just before the elevator doors opened, the IT manager mentioned to his CEO: |
| |
"It's a challenge to repair the damage.
It's a frustration when you don't quickly know it occurred.” |
|
As he left the elevator, the CEO called over his shoulder: |
| |
"Perhaps we should get you a burglar alarm.” |
|
and the elevator doors closed. |
He had never thought about a burglar alarm before. He did not get off
at his floor. Instead, he took the elevator down to the Plant Security
Department. He asked some questions and got a few answers. Here's what he
learnt.
If something is important, we deploy a sentry to watch it. A common
example is unauthorized access detection. A more esoteric example
is temperature monitoring of the warehouse freezer. When a sensor is
triggered, the appropriate authority is notified so that corrective
action can be undertaken.
Historically, the manufacturers of sensor equipment also provided the notification
equipment. At first it was an alarm bell, but we quickly ran out of
bells as the number of critical systems increased. Further, an alarm
bell could only help if the appropriate authority was there to hear it.
Central station alarm monitoring service bureaus arose to meet the challenge.
They are the services we tend to rely upon to tell us when things go wrong.
The particular sensors employed by Plant Security were not relevant to IT.
However, learning how the central station alarm monitoring service
bureau fit into the equation raised his eye brows. His department
was awash in cell-phones and pagers, yet, particularly during
off-hours, it was hard to get hold of staff.
Back in the elevator, the word “ironic” kept coming to mind.
- Security was using tools to monitor events and trigger alerts for things that
were fewer in number and lesser in sophistication than those of his own department.
- Security was concerned about thieves trying to break in.
He was concerned about hackers trying to break in.
- Security was concerned that the freezer might thaw.
He was concerned that a CPU might over-heat.
- Proper dispatch procedures and technologies do exist. Central station
alarm monitoring service bureaus apparently do it all the time.
Yet, when it comes to the monitoring of Internet facilities, where
are they?
The IT Manager has now exited the elevator. I echo his question and offer something
of an answer:
The Internet presents an entire array of systems whose uninterrupted
operations are critical. Yet, the very “openness” and
pervasiveness of the Internet increases the likelihood of trouble.
Making matters even more challenging is the fact that some of these
critical systems are located many miles away and are not even under
our direct control. Yes, Internet facilities monitoring is critical.
It would seem that central station alarm monitoring service bureaus
don't yet know much about monitoring critical Internet systems. It
is time they woke up. The good news is that there are signs of
awakening.
Its name is Perimeter911.
Perimeter911 is an external service that monitors your Internet facilities through
the same channels that the public uses. Thus, Perimeter911 monitors
the real thing. It knows (because you tell it) how those facilities
ought to be responding.
- When something goes wrong, Perimeter911 alerts key personnel by
pager or cell-phone.
- And there's more. Perimeter911 is also closely
allied with Securitech, a central station alarm monitoring and
dispatch service bureau.
|
Hidden away in an underground security vault, Securitech alarm panels are
attached to some interesting devices that link up to Perimeter911.
When something goes wrong, Securitech dispatch operators can notify
your designated authorities. They do so with the same level of
professionalism with which they monitor all sorts of other critical
systems.
|
The future will probably see more of this sort of thing. In the
meantime, Perimeter911 stands out as quite unique.
Even companies that do their own in-house monitoring should be thinking
about Perimeter911.
PageMart Canada
It was quite unusual, but
then again, it is the “unusual” that this is all about. It
happened in August,2001.
The sophisticated self-monitoring equipment in the data centre suggested
that everything was ok. None of the 24/7 IT staff suspected
anything. They didn't know that their entire national paging service
network was effectively down.
However, the network manager knew something was wrong. He knew because
Perimeter911 told him.
Meanwhile, the switchboard was lighting up like a Christmas tree. The puzzled
operator didn't know what to tell customers. She didn't know until
the network manager walked up to her desk. Here's what he told her.
|   |
"There has been a fire at a critical Bell Canada central office.
It has nothing to do with us,
but that will be small consolation to our customers. Our credibility
will still be on the line. Ask our customers if they wish to be
notified the moment it has been cleared up.
“There may be reports of Bell Canada spokesmen saying
that the problem should be cleared up in an hour or so. Don't put
our neck on the line by repeating their assurances.
"Keep this pager next to your your phone. When things
are rectified, you'll get a message saying “ALL CLEAR.”
|
Behind the scenes, here's what Perimeter911 was doing:
- On behalf of the customer, Perimeter911 was monitoring its SNPP
services. (SNPP is an Internet protocol specialized for the paging
industry.)
- When it detected the failure, it issued 2 alarms:
- An Internet based alert (which could not be delivered).
- An operator dispatched alert (which was how they found out).
- Perimeter911 continued to monitor the situation.
- When things finally changed, it issued the ALL CLEAR.
There are other stories as well.
- One customer used Perimeter911 log records as evidence during a
litigation process. His ISP claimed that connectivity was stable,
but the log records proved otherwise. (There is something to be said
for 3rd party confirmation.)
- One customer's activities depended on the timeliness of data made
available by a weather service bureau. That customer was concerned
that the weather data might not be available or might simply be
out-of date. Not only did he want to be notified, but he
arranged for notification to be sent directly to the service bureau
in the event of failure.
- One customer's web-site contained critical links to a trading partner.
They needed to know if things were kosher at their trading partner's
site.
- Perimeter911 has had mom-and-pop small businesses whose owners simply had
to know. The owners didn't want to leave their cell-phones on all night.
Perimeter911 is something you should know about.
|