Say, I have a T1 1.554 Mbps, how do I know if I have enough for my
For example, I have 50 employees, some people using video conferencing,
some using WebEx, some checking email and surfing the net, some people
downloading stuff. They maybe doing at the same time or maybe not. Is there
a general formula or a way on how to check a line is "enough"? how? I
want some solid numbers.
There is no formulas that can cater for such unknowns. Remember that
the Internet is still growing close to exponentially. The US doubling
rate is up above a year, but elsewhere it may be down to the 7 months
I use more than 1 T1 for my own stuff, and I am just using personal
stuff. OK, some video and large web pages help to consume bandwidth.
E-mail is difficult to give a firm figure for. A simple, text only
e-mail is only a few K, but once you start hauling large office documents
or cad drawings around they can extend to hundreds of gigabytes, each.
It may be doable witl a T1 for a company of 50; but it will need work with
Some figures : VOIP consumes from 18 to 88 kilobits/sec per active voice call,
depending on the level of compression. With video the figure is 128k to around
4 mbit. Usually they are in the 348k-1m area. If it is 384 or 1m will
probably be a make-or-break figure for your case.
You will anyway need a firewall that can prioritize traffic in and outgoing,
and can do bandwidth enforcement. If you need a cheap one, look at
Need more data, and if you tell me more, I'm going to start charging you
by the hour. *grins*
The quick and easy method is to wait for people to start bitching about
slow performance. When the level exceeds your tolerance level, go get
A more detailed use required logging of your connection with reasonably
sophisticated traffic analyzer software, and a business needs analysis. Neither
of which is cheap or really off the shelf formula sorts of things.
You may be asking the wrong question because there is never 'enough'.
What you may want to do is to use switches in your LAN so that all
telephone calls get 'priority' over data and the interference between
the two are minimized.
You didn't mention users sending important LARGE files from engineering
to drafting ;-) at the same time as a vital call is closing a sale.
There is *NO* estimating formula.
What you do is get what you think is somewhat 'more than enough' capacity
initially, then you *monitor* actual usage. If your circuit is at 100% of
capacity for significant periods, you consider getting a 'bigger' pipe.