I want to design a system which mediates user requests against the machine resources. Ideas i need...
Hi there;
I want to design a system which mediates user requests against the machine resources. Ideas i need.
I have multiple clients that need to execute a program remotely. My implementation regarding executable invocation is done but i need ideas to evaluate the request for starting and stopping that executable "against computer resources". Can you help me?
Kind regards.
I want to design a system which mediates user requests against the machine resources. Ideas i need.
I have multiple clients that need to execute a program remotely. My implementation regarding executable invocation is done but i need ideas to evaluate the request for starting and stopping that executable "against computer resources". Can you help me?
Kind regards.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
My first reaction would be to let the operating system do what operating systems do -- schedule. However, it appears you want something more.
It appears the problem is that the calculations are CPU intensive and you don't want too many running at once.
How many do you want to allow to run at once?
Is it possible to predict the execution time from the request?
Will there be jobs with higher priorities than others?
What are you planning to say to the users who are waiting? ("You are 7th in line. Estimated time is XXX." Giving an estimate only works if can do some prediction.)
If this is really demanding but also important, I suggest a server with many processors and lots of RAM. Depending on how much money your company is willing to spend, you might consider multiple servers. This gets advanced.
It appears the problem is that the calculations are CPU intensive and you don't want too many running at once.
How many do you want to allow to run at once?
Is it possible to predict the execution time from the request?
Will there be jobs with higher priorities than others?
What are you planning to say to the users who are waiting? ("You are 7th in line. Estimated time is XXX." Giving an estimate only works if can do some prediction.)
If this is really demanding but also important, I suggest a server with many processors and lots of RAM. Depending on how much money your company is willing to spend, you might consider multiple servers. This gets advanced.
ASKER
The users are expected to reside in a same LAN but may vary. The server machine ought to be very str
ong.
So, given those above, what kind of design against server resources for the very queing system?
I need ideas.
Kind regards.