summary.seqs in MPI

Hello,

I have some problem in MPI on summary.seqs command. Normally, I used batchfile for running mothur on MPI. For line of align.seqs can generate the result, however, when the step of line on summary.seqs is run, the problem occurs in below:

mothur > summary.seqs(fasta=project1seafinal.good.unique.align, count=project1seafinal.good.count_table)
[compute-1-2:3454] *** An error occurred in MPI_Recv
[compute-1-2:3454] *** on communicator MPI_COMM_WORLD
[compute-1-2:3454] *** MPI_ERR_TRUNCATE: message truncated
[compute-1-2:3454] *** MPI_ERRORS_ARE_FATAL: your MPI job will now abort

Using 16 processors.
--------------------------------------------------------------------------
mpirun has exited due to process rank 2 with PID 3448 on
node compute-1-2 exiting improperly. There are two reasons this could occur:

1. this process did not call "init" before exiting, but others in
the job did. This can cause a job to hang indefinitely while it waits
for all processes to call "init". By rule, if one process calls "init",
then ALL processes must call "init" prior to termination.

2. this process called "init", but exited without calling "finalize".
By rule, all processes that call "init" MUST call "finalize" prior to
exiting or it will be considered an "abnormal termination"

This may have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).
--------------------------------------------------------------------------
[compute-1-2:03445] 14 more processes have sent help message help-mpi-errors.txt / mpi_errors_are_fatal
[compute-1-2:03445] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

I’m not sure what is happen, Could you please suggest me? I used Mothur.1.34.1 version.
Thank you very much.

We actually just posted a new version of 1.34 to address MPI problems people were having. Can you pull down the latest version and try again?

Pat

Thank you very much for your suggestion. I just try a version of 1.33. Now, it does not have the problem. :smiley:

I think I have resolved the issue for you. Can you try version 1.34.2, http://www.mothur.org/wiki/Download_mothur?

Hi,

I am a IT person in our research institute. I am trying to install mothur 1.36.1 on our cluster for one of our researchers. But, while doing the test run by using summary.seqs function, I always got the same problems posted here. I thought this problem should have been resolved after version 1.34.2 and it should not be an issue for 1.36.1. Can anyone help me to fix this error? I built mothur 1.36.1 with gcc 4.9.2 and openmpi 1.8.4 on rhel 6 with kernel 2.6.32-573.1.1.el6.x86_64. Thanks a lot

LSW


Here is the output from my log file ```text mothur > summary.seqs(fasta=test.fasta, count=test.count_table)

Using 1 processors.

            Start   End     NBases  Ambigs  Polymer NumSeqs

Minimum: 1 420 137 0 2 1
2.5%-tile: 1 428 167 0 3 327307
25%-tile: 1 428 167 0 4 3273067
Median: 1 428 167 0 4 6546133
75%-tile: 1 428 167 0 4 9819199
97.5%-tile: 1 428 168 0 6 12764958
Maximum: 11 471 207 0 8 13092264
Mean: 1.00014 428.014 167.055 0 4.11019

of unique seqs: 478399

total # of seqs: 13092264

Output File Names:
test.summary

It took 17 secs to summarize 13092264 sequences.

mothur > quit()
[c132:6625] *** An error occurred in MPI_Recv
[c132:6625] *** reported by process [1152909313,6]
[c132:6625] *** on communicator MPI_COMM_WORLD
[c132:6625] *** MPI_ERR_TRUNCATE: message truncated
[c132:6625] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
[c132:6625] *** and potentially your MPI job)

I wonder do anyone have some idea about my summay.seqs error. Thank you.

Our non-MPI version runs faster than the MPI version. In our next release we will be removing the MPI option. I would recommend using the non-MPI version.