[Info-vax] Marketing ideas for VSI ?

[Arne Vajhøj]

On 12/15/2018 5:12 PM, Dave Froble wrote:
> On 12/15/2018 3:48 PM, Kerry Main wrote:
>> An analogy ..
>>
>> 6 cars lined up on a 6 lane highway that is exactly 100 miles long -
>> straight as an arrow with each car max speed = 200mph.
>>
>> The fastest any one car will make it to the end is 30mins (latency)
>>
>> Now, in this scenario, a total of 6 cars can make the trip in 30min's
>> (bandwidth)
>>
>> Adding an additional 6 lanes to the highway to improve the bandwidth 
>> means a
>> total of 12 cars can make it to the end in the same amount of time.
>>
>> However, the fastest any one car can make it to the end is still 30min's.
> 
> I totally understand that part Kerry, but, for that 6 lane highway, what 
> happens when there are 12, 18, 24, or more cars?  Is that much of a 
> problem, or do they just line up and go, and the small difference 
> between cars isn't much of an issue?

The point is that you can solve the more cars problem by adding more
lanes but you can not solve the time it takes to get all the way
problem by adding more lanes.

If we get back to IT.

Let us say that you have a database type application where the
data storage are distributed over WAN.

If the users start to store much larger data pieces, then you
can solve that by adding more network bandwidth.

If the bottleneck is that some records are being locked and
other tasks are waiting for that record to become available then
the latency becomes the issue. If the latency is X ms, then
enqueue lock request, enqueue lock response, write data request,
write data response, dequeue lock request and dequeue lock reesponse
each take X ms a total of 6X ms. That means that max TPM is
60000/6X. And it does not help to upgrade from 100 Mbit to 1 Gbit.
It may be possible to upgrade to a network with fewer boxes wasting
time, but the speed of light put a lower bound on X.

Arne