search

marius-team / marius

Large scale graph learning on a single machine.
https://marius-project.org
Apache License 2.0
160 stars 45 forks source link

Always have sgmeantation fault (core dumped) when in DISPERSED mode #144

Open yichuan520030910320 opened 1 year ago

yichuan520030910320 commented 1 year ago

Describe the bug A clear and concise description of what the bug is. When I ran https://github.com/marius-team/marius/blob/eurosys_2023_artifact/experiment_manager/example/configs/ogbn_arxiv/marius_gs_disk_gpu.yaml using marius_train and change feature to this
features: type: PARTITION_BUFFER options: num_partitions: 32 buffer_capacity: 3 prefetching: true fine_to_coarse_ratio: 1 num_cache_partitions: 0 node_partition_ordering: DISPERSED it always has core dump To Reproduce Steps to reproduce the behavior:

  1. Go to eurosys_artifact
  2. follow the pip install in the readme
  3. 3.change the marius_artifact/experiment_manager/example/configs/ogbn_arxiv/marius_gs_disk_gpu.yaml in the entry of features to features: type: PARTITION_BUFFER options: num_partitions: 32 buffer_capacity: 3 prefetching: true fine_to_coarse_ratio: 1 num_cache_partitions: 0 node_partition_ordering: DISPERSED
  4. marius_train marius_artifact/experiment_manager/example/configs/ogbn_arxiv/marius_gs_disk_gpu.yaml

Expected behavior Segmentation fault (core dumped) Environment List your operating system, and dependency versions. You can obtain this by running marius_env_info from the command line. bash: marius_env_info: command not found? Additional context Add any other context about the problem here.

rogerwaleffe commented 11 months ago

Thanks for posting this issue! A similar issue seems to be occurring in the main branch which I investigated here: #143. Can you try running with the main branch and the config file I provided in that issue?