Jacker
Jacker
@Zarca @eedalong To load from HDFS or other distributed file systems, we need to implement the corresponding interfaces. Currently, only local file system is supported. Exactly to what you said,...
Show the code of how you launch the distributed processes.
Make sure the number of workers is not less than that of PS. More than one workers may connect to ONE PS. And we suggest the number of workers can...
I think [this](https://github.com/alibaba/graph-learn/blob/master/docs/graph_object_cn.md#32-%E5%88%86%E5%B8%83%E5%BC%8F--server-mode) is what you need.
Graph partition is a classical problem which was proposed many years ago. But till now, universal solutions are not found. In my point, we can develop kinds of task specific...
Good suggestion. May I know your OS and g++ version?
Good suggestion and the solution is coming.
We did not re-produce your problem. Would you mind to supply more details, about the size of data, the cluster info and so on. Besides, the performance will not always...
@jingmouren 不同环境编译方式可能有差异，手动改一下Makefile，在so编译时加上-lunwind
Yes, a distributed file system is needed. You can launch a distributed program on a local machine, or mount NFS to local directory to establish a real distributed model. Currently,...