I came across to this paper ciel[1], which provides dynamic task graph composition. Just curious, is there any plan that hama may support in the long run as it seems to fit the goal of hama for scientific computation?
[1]. CIEL: a universal execution engine for distributed data-flow computing. http://www.cl.cam.ac.uk/~dgm36/publications/2011-murray2011ciel.pdf 2011/3/27 Edward J. Yoon <[email protected]>: > Hi, > > First of all, please compare the similarities and differences between > Hama Breadth First Search[1] and Pregel Breadth First Search[2]. (FYI, > Hama[1] was written earlier than Pregel paper.) > > The process() function which is described in Hama[1], is a current > bsp() function. And, you can understand that the expand() function > which is described in Hama[1], is the same with Google Pregel's > compute() function. > > I mean, the Pregel-like graph-specific programming model can be > implemented using BSP on top of Hama project. Basically, our project > goal is a generalized BSP computing engine[3]. But, I think it would > be nice if we can support the Pregel-like graph solution. (as a > sub-project of Hama) > > 1. http://people.apache.org/~edwardyoon/papers/Apache_HAMA_BSP.pdf > 2. http://blog.udanax.org/2010/06/summary-of-google-pregel.html > 3. https://blogs.apache.org/hama/entry/how_will_hama_bsp_different > > On Sun, Mar 27, 2011 at 4:43 PM, Tommaso Teofili > <[email protected]> wrote: >> Hi Edward, >> could you explain in more details what you have in mind? >> Cheers, >> Tommaso >> >> 2011/3/27 Edward J. Yoon <[email protected]> >>> >>> Hi all, >>> >>> I think it's too early stage to discuss but, it seems very important >>> in promotional view. :/ >>> >>> What do you think about Pregel-clone subproject on top of Hama BSP? >>> >>> -- >>> Best Regards, Edward J. Yoon >>> http://blog.udanax.org >>> http://twitter.com/eddieyoon >> >> > > > > -- > Best Regards, Edward J. Yoon > http://blog.udanax.org > http://twitter.com/eddieyoon > -- ChiaHung Lin @ nuk, tw
