Considering this error output: >lamnodes: execution: >------------------- >n0 silk-node7:2:origin,this_node >n1 silk-node6:2: >n2 silk-node5:2: >n3 silk-node4:2: >n4 silk-node3:2: >n5 silk-node2:2: >n6 silk-node1:2: >error message: >------------- >execution: ./mpiblast.sh on the masternode (silk-node 7 n0) >Wed May 18 14:29:55 IST 2005 >Sorry, mpiBLAST must be run on 3 or more nodes >[0] MPI Abort by user Aborting program ! >[0] Aborting program! >p0_15547: p4_error: : 0 >Sorry, mpiBLAST must be run on 3 or more nodes >[0] MPI Abort by user Aborting program ! >[0] Aborting program! >p0_15548: p4_error: : 0 >Sorry, mpiBLAST must be run on 3 or more nodes >[0] MPI Abort by user Aborting program ! >[0] Aborting program! >p0_5535: p4_error: : 0 >Sorry, mpiBLAST must be run on 3 or more nodes >[0] MPI Abort by user Aborting program ! >[0] Aborting program! >p0_5536: p4_error: : 0 >----------------------------------------------------------------------------- >It seems that [at least] one of the processes that was started with >mpirun did not invoke MPI_INIT before quitting (it is possible that >more than one process did not invoke MPI_INIT -- mpirun was only >notified of the first one, which was on node n0). > >mpirun can *only* be used with MPI programs (i.e., programs that >invoke MPI_INIT and MPI_FINALIZE). You can use the "lamexec" program >to run non-MPI programs over the lambooted nodes. >----------------------------------------------------------------------------- > > My understanding is that the Scalable Informatics mpiBLAST RPMs have been compiled against mpich, whereas you are using LAM/MPI. mpiblast is calling mpich's MPI_Init and LAM never gets an MPI_Init call, leading to the above error. Either switch to mpich or rebuild mpiBLAST on your system with LAM. -Aaron