MPICH中的错误处理

碰到了两次了，写下来备忘！

error message: 

$ mpdboot -f  file :

failed to ping mpd on ... ; recvd output={}

 

solution: 

  1. just make sure, you got a correct working /etc/hosts file ...

     ... "akin" problem ... same error_message: mpicheck  /etc/hosts
 

  2. verify, you are using the SAME "MPD_SECRETWORD" defined in /etc/mpd.conf on every machine must user secretword=12345 ,can't  write like mpd_secretword=12345

  3. verify, mpich2 is available on every machine in the same directory and
      the environment variables PATH and LD_LIBRARY_PATH are set properly