HPC应用分析
收藏回复举报
HPC应用分析
t('forum.solved') 已解决
发表于2024-03-06 17:33:57
0 查看

1. 使用命令行方式分析namd应用,使用-d指定采样时间为20时,会出现以下报错,采样失败;同样的采样参数在web界面也会报错退出;

 70] Stack Traceback: 

  [70:0] linux-vdso.so.1 0xffff9e04693c __kernel_rt_sigreturn 

  [70:1] libucp.so.0 0xffff986bc970 ucp_worker_progress 

  [70:2] libucg_planc_ucx.so 0xffff9030b920 ucg_planc_ucx_p2p_testall 

  [70:3] libucg_planc_ucx.so 0xffff90306708 

  [70:4] libucg.so.1 0xffff981c2e74 

  [70:5] libucg.so.1 0xffff981c5fbc ucg_request_test 

  [70:6] mca_coll_ucg.so 0xffff981f2088 mca_coll_ucg_request_execute 

  [70:7] mca_coll_ucg.so 0xffff981f21d0 mca_coll_ucg_request_execute_cache 

  [70:8] mca_coll_ucg.so 0xffff981f5114 mca_coll_ucg_barrier_cache 

  [70:9] libmpi.so.40 0xffff9ada050c MPI_Barrier 

  [70:10] libmpicollector_ompi.so 0xffff9bf84850 MPI_Barrier 

  [70:11] namd3 0xb8e5e4 

  [70:12] namd3 0xb932b4 

  [70:13] namd3 0xb8cf0c 

  [70:14] namd3 0x473194 

  [70:15] namd3 0x46f790 

  [70:16] libc.so.6 0xffff9a975fbc 

  [70:17] libc.so.6 0xffff9a976094 __libc_start_main 

  [70:18] namd3 0x40f8f0 

-------------------------------------------------------------------------- 

MPI_ABORT was invoked on rank 70 in communicator MPI COMMUNICATOR 3 DUP FROM 0 

with errorcode 1. 

 

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes. 

You may or may not see output from other processes, depending on 

exactly when Open MPI kills them. 

-------------------------------------------------------------------------- 

[Hyper Tuner MPI 23.0.T30][INFO]enter _cluster_clean 

[Hyper Tuner MPI 23.0.T30][ERROR]Task 1(statistics) fail. 

[Hyper Tuner MPI 23.0.T30][INFO]******************************************************************************************************************************************************************** 

[Hyper Tuner MPI 23.0.T30][INFO]* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *    PACK DATA    * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 

[Hyper Tuner MPI 23.0.T30][INFO]******************************************************************************************************************************************************************** 

[Hyper Tuner MPI 23.0.T30][INFO]Number of task: 1 

[Hyper Tuner MPI 23.0.T30][ERROR]package failed! 

2. 指定采样时间为-u时,出现以下warning信息,可以采样成功,但采样时间较长。

mpool.c:44   UCX  WARN  object 0x29fc0bc0 was not returned to mpool ucp_requests 

mpool.c:44   UCX  WARN  object 0x29fc0d80 was not returned to mpool ucp_requests 

mpool.c:44   UCX  WARN  object 0x29fc0f40 was not returned to mpool ucp_requests 

mpool.c:44   UCX  WARN  object 0x29fc1100 was not returned to mpool ucp_requests 

mpool.c:44   UCX  WARN  object 0x29fc12c0 was not returned to mpool ucp_requests

我要发帖子