adf:rdma
故障：CMA: unable to get RDMA device list
特征
出现类似如下提示：
CMA: unable to get RDMA device list CMA: unable to get RDMA device list CMA: unable to get RDMA device list CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 librdmacm: couldn't read ABI version. librdmacm: assuming: 4 librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 librdmacm: couldn't read ABI version. librdmacm: assuming: 4 CMA: unable to get RDMA device list CMA: unable to get RDMA device list librdmacm: couldn't read ABI version. librdmacm: assuming: 4 DAT: library load failure: libdaploucm.so.2: cannot open shared object file: No such file or directory librdmacm: couldn't read ABI version. librdmacm: assuming: 4
解决方式
在环境变量中，添加如下内容： export I_MPI_FABRICS=shm:tcp
adf/rdma.txt · 最后更改: 2017/04/12 20:45 由 liu.jun