Intel® Moderncode for Parallel Architectures
Support for developing parallel programming applications on Intel® Architecture.

Capacity planning

aminer10
Novice
341 Views

Hello,

I have come to an interresting subject, if we have a distributed database and a webserver and HTML files and you want to do a capacity planning of your webserver this will complicate the things, cause the database server must be modelized as an hyperexponential distribution that is an M/G/1 queuing system , but as you have noticed since the database server system , in our network , comes before the internet connection that will be modeled as an M/M/1 queuing system, so you have to use a queuing network simulation to solve this problem , but if you have noticed, in capacity planning we have also to calculate the response time of the worst case performance, so this will easy the job for us cause in the worst case scenario since the M/G/1 queuing system of the database server have three exponential distributions for the reads and writes and deletes transactions, so we have to choose the worst service time that is exponentially distributed , so i think we have to choose only the  exponential distribution of the writes transactions, so this will make the queuing system of the database server a markovian M/M/1 queuing system, so this will transform our queuing network into a Jackson network so this will easy for us to calculate the response time of our web server, but since the bottelneck of the network is the internet connection so this will easy for us much more the capacity planning .. hence if you want to do a capacity planning of your webserver you have to concentrate on the very important part that is the M/M/1 internet connection and do your calculation.

I have done a M/M/1 and M/M/c queuing network simulation in Object pascal to easy for you to do capacity planning of
your webserver in Object pascal.


You will find the source code here:

http://pages.videotron.com/aminer/


Thank you,
Amine Moulay Ramdane.


Thank you,
Amine Moulay Ramdane.

0 Kudos
3 Replies
aminer10
Novice
341 Views

Hello,

Question:

But what is the purpose of the capacity planning using the queuing theory ?


Answer:

You have to be very carefull, if for example you are using a database
with your webserver and you are accessing this database from internet or intranet and you are doing reads and writes and deletes transactions on this database and after that the data of the databse is served through internet or intranet , as i told you you can use a worst case scenario cause the database queing node has an hyperexponential on its service side, so you can choose the worst case scenario, so if the reads takes much more time than the writes and deletes, you have to choose only the rate of the writes transactions , the rate means how many writes per second, as your lambda, but notice with me that the network node or the intranet node of your network is markovian, so if for example the bandwidth of your network internet connection is 30Mbps, since it's an M/M/1 markovian queuing node that means that the bandwidth will be reduced by 50% cause the Knee of the M/M/1 internet network queuing node is reached at 50%, and don't forget also
about the TCP/IP protocol overhead of ~20% , so  you have to be carefull when you do capacity planning..

And why you have to do capacity planning ? cause if you must
ensure a quality of service and set a constraint on the maximum waiting time, let for example say 5 seconds for each internet user to receive
it's webpage, you have to do capacity planning to set the maximum number of connections ofyour webserver correctly and do your analytic queuing network calaculations or you queuing network simulation.



Thank you
Amine Moulay Ramdane.

0 Kudos
aminer10
Novice
341 Views

I correct , please read again...

Hello,

Question:

But what is the purpose of the capacity planning using the queuing theory ?


Answer:

You have to be very carefull, if for example you are using a database
with your webserver and you are accessing this database from internet or intranet and you are doing reads and writes and deletes transactions on this database and after that the data of the databse is served through internet or intranet , as i told you you can use a worst case scenario cause the database queuing node has an hyperexponential on its service side, so you can choose the worst case scenario, so if the reads takes much more time than the writes and deletes, you have to choose only the rate of the reads transactions , the rate means how many reads transactions per second, as your lambda, but notice with me that the network node or the intranet node of your network is markovian, so if for example the bandwidth of your network internet connection is 30Mbps, since it's an M/M/1 markovian queuing node that means that the bandwidth will be reduced by 50% cause the Knee of the M/M/1 internet network queuing node is reached at 50%, and don't forget also
about the TCP/IP protocol overhead of ~20% , so  you have to be carefull when you do capacity planning..

And why you have to do capacity planning ? cause if you must
ensure a quality of service and set a constraint on the maximum waiting time, let for example say 5 seconds for each internet user to receive
it's webpage, you have to do capacity planning to set the maximum number of connections of your webserver correctly and do your analytic queuing network calculations or your queuing network simulation.



Thank you
Amine Moulay Ramdane.




0 Kudos
Bernard
Valued Contributor I
341 Views

>>>if for example the bandwidth of your network internet connection is 30Mbps,>>>

Do you mean Last Mile?

0 Kudos
Reply