Hey wangJing, I know that we currently have a lot of memory problems with the graph processing. Depending on your graph it may take more than 20x the input size in memory. We are currently working on that, I already changed the graph part- but I'm missing the messaging improvements. As soon as we resolved that issues and tested it carefully we roll out a new release that will certainly do better :) Sorry for that inconvenience wangJing.
PS: I hope you don't mind that I forward this to the user list as it may help other people as well. 2013/3/15 qswm611 <[email protected]> > Hi,Thomas: > I'm sorry to bother you. When I run some small graph test on my cluster > 2Mb,graph data job can be succeed,but 10Mb or bigger can not. > The information is：INFO.ipc.NettyTransceiver: > id:.....=>id:.....INTEREST_CHANGED.and > so on. > > Do you have any idea what may cause this kind of fail?and how can I solve it > ? Thanks a lot! > > > wangJing > > > >
