This is my first time hearing bout DAG,

with some  google search i found this (could it help):

-   http://research.cs.wisc.edu/htcondor/dagman/dagman.html

-   http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4095736/

-   https://github.com/frs69wq/biCPA         or    https://github.com/frs69wq/biCPA/blob/master/src/bicpa.c

-   https://github.com/dapurv5/distributedscheduling

-   https://github.com/JamesClough/dagology

Dear Peter,

Thank you for your answer, I am currently working on a Task Scheduling for multicore processor(Parallel systems) the problem seems to be Np hard and must be tackled By Heuristics, I have read a couple of books and papers, Dags are a model to represent the problem,I understand your point, but my work is limited to static scheduling.

thank you for your help, means a lot to me.

Regards,

Soroush

Dear Bekk,

Thank you for your answer, I will check the links,Thanks again.

Regards,

Soroush

Assume you have a Graph, with Nodes which each of them got number represent the computation time, and the edge are numbered with communication time, the makespan in the Scheduling literature means the maximum completion time, there are two kind of scheduling algorithms with respect to the parallel computing: List scheduling and clustering, my algorithm categorizes in the latter case,

If two nodes(or More) assign to on core of processor the communication time(edge between nodes) become zero clearly because they are using one core and don't have to migrate from one to another, but if one task(node)computed on one core and the other computed on another core then we have the communication cost.

the question is: what is the best assignment of tasks to cores in order to minimize the total completion time(A.K.A makespan)

if you are interested to get more information on my question please read:https://www.researchgate.net/post/I_am_Working_on_my_Thesis_which_is_about_Scheduling_of_cores_on_multi_core_processingParallel_systems_I_was_wondering_if_anyone_worked_in_this_Area/1

sorry it took a little bit long, thank you so much Peter.

regards,

Soroush  

DAGitty is a tool to create, edit, and analyze causal models

Dear Chetan,

Thank you For Your answer.

I have tried to use it but it seems like DAGitty is not able to calculate Critical path in DAGs, it is most likely to be used for checking the logic of loops and cause and effect aspect, Am I Wrong?

You can use optaplanner for the same.

Thank you Manmeet, I will search for it

In the work described in http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6645595 we started with a Fortran program:

i.  Expanded the sub-programs in-line

ii.  Turned all the arrays into scalars

iii. Separated the symbol lives

iv. Found the dependencies and computed the DAG

v. Applied scheduling heuristics to schedule execution for a multi-code fine-grain system.

All of this was done with purpose-built software.  Is any of this useful to your application?

DAGitty is software tool.  For more information visit http://www.dagitty.net/ 

I think this may help......

Dear John,

Thank you for your answer, I don't know about HPCs does it stand for High performance cluster? or high performance computer?

your explanation is very useful to my application, parallel systems, but I might need to develop my own tool.

I would like to read your article, if it is possible for you please send me a soft copy of that.

Best regards,

Soroush

Dear Dr.Bairam,

DAGitty is a very good tool for modeling and analyzing  DAGs but Unfortunately it is not able to calculate the critical path or any other mathematical aspect of DAGs. it is be used to analyze the cause and effect in epidemiology(to my best knowledge).

thank you for your answer and your time.

Regards,

Soroush

HAMA tool (BSP programming) with HADOOP solves the graph related problems.

Dear Chandio,

Many thanks for your answer, I am reading the tutorial on it's wiki, thank you again.

Regards,

Soroush

dagittty is not of much significance as it works with no mathematical significant principles and lacking various visualization principles