Failed to use distributed graph partitioner

I followed the sample program in example/pytorch/graphsage/dist/partition_graph.py and used the instructions in the readme file to run the program, but the partition program quit halfway through. Since there are no errors, I don’t know exactly where the problem is.
My server is ubuntu20.04, the memory is 128GB, and the dgl download is a non-GPU version.

This issue could be Out of Memory. But 128GB should be enough.

I tried on my side and it works. my DGL is 0.9.1. which DGL version are you using? And could you double confirm the RAM of your container via free -h?

Thanks for your reply, it seems that it may be because of memory constraints, I have successfully run it on another server.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.