Yes, learning on a dedicated Spark cluster and predicting inside a Storm bolt is quite OK :)
Thanks all for your answers. I'll post back if/when we experience this solution. E/ 2014-06-19 20:45 GMT+02:00 Shuo Xiang <shuoxiang...@gmail.com>: > If I'm understanding correctly, you want to use MLlib for offline training > and then deploy the learned model to Storm? In this case I don't think > there is any problem. However if you are looking for online model > update/training, this can be complicated and I guess quite a few algorithms > in mllib at this time are designed for offline/batch learning. > > > On Thu, Jun 19, 2014 at 12:26 AM, Eustache DIEMERT <eusta...@diemert.fr> > wrote: > >> Hi Sparkers, >> >> We have a Storm cluster and looking for a decent execution engine for >> machine learned models. What I've seen from MLLib is extremely positive, >> but we can't just throw away our Storm based stack. >> >> So my question is: is it feasible/recommended to train models in >> Spark/MLLib and execute them in another Java environment (Storm in this >> case) ? >> >> Thanks for any insights :) >> >> Eustache >> > >
