Read again....

[Ramine]

Hello.........


I was working all those previous days on my new projects,
i have worked on two projects, my first project cons=ist of
an optimal implementation that uses Dijkstra's algorithm with
a binary heap that takes a time complexity of E*log(V), V is
the number of vertices and E is the number of edges. This library can be
used in parallel clusters manner by dividing your graph in many parts to
speed much your parallel algorithm, also i have added an option that
permit you to pass the edges of the graph that you can substract from
your graph to be able to give you algorithm more control if you want for
example to ignore congestions in some roads...

This was my first project that i am finishing.

My second project is about capacity planning , i have come
with an algorithm that permit you to fix the number of internet
connections easily and have control on the waiting time
of the internet users that are using for example your Ecommerce
website, my new algorithm uses a new semaphore that is
FIFO fair because it must be FIFO fair to be be more optimal,
and i will also show you how to model an Ecommerce with queuing
theory that contain a queue with an hyperexponential service that
can be approximated with an M/M/1 queue when the writer transactions are
less or equal to 30% of the total transactions and the Ecommerce have a
read-mostly workload.


I am right now finishing my second project.




Thank you,
Amine Moulay Ramdane.

[Ramine]

Hello...........

I have come as i have talked before to my post about Ecommerce website
modeling with mathematical queuing theory...

As you have noticed Sir/Madam that i am a programmer and as you
have noticed i have invented many Synchronization algorithms,
and i have implemented many parallel algorithms.. here they are:

https://sites.google.com/site/aminer68/

But during all those years i have come to a conclusion that
the job of programming is not a sufficient condition to become more
professional, and i have come to a conclusion that Operations research
is a necessary requirement for me as a programmer, because mathematical
queuing theory and linear programming etc. in Operations research is a
requirement to be able to implement and model for example an Ecommerce
website and for a better QOS, this is why in this post i will show you
how to model a database server that has an hyper-exponential service, so
how can we use this M/G/1 queue that has an hyper-exponential service to
model mathematically an Ecommerce website, i think that to not get into
simulation, since many Ecommerce websites have read-mostly workloads,
the hyper-exponential queue that is an M/G/1 can be approximated with an
M/M/1 queue when the writer transaCtions are less or equal to 30% of the
total transaCtions and the Ecommerce website has read-mostly workloads,
and to not get into mathematical proofs i will give you the mathematical
equations of the M/G/1 queue that have an hyper-exponential service that
permit us to model a database server, here they are:

The mean that is the the mean time of the service is:

M1 = p1/a + p2/b...

and the second moment is:

M2 = 2*p1/a^2 + 2*p2/b^2...

and the variance is:

variance = M2 - M1^2

a and b are the service rates of the different transactions such us
read,write and delete...

and p1 and p2 are the percentage of the transactions.

and that's all:

So when you calculate and get M1 you will then calculate the mean
service time that is 1/M1 and you will plug this on your M/M/1 THAT
is an approximation of the above case using the arrival rate and that's
all, and now you can model mathematically and easily your Ecommerce
website by simplifying a little bit the model by using queues that are
interconnected in a serial manner.