maker

[mictadlo]

Hi @abretaud, @nathanweeks, @johanneskoester, @kastman and @ArneKr,
I am running maker in the following way for 140 hours on 100 compute nodes. Is it normal that no output files have been generated until now?

#!/bin/bash -l
#PBS -N maker
#PBS -l walltime=150:00:00
#PBS -j oe
#PBS -l select=100:ncpus=1:mpiprocs=1:mem=4G

cd $PBS_O_WORKDIR

conda activate maker

mpirun maker -fix_nucleotides

Thank you in advance,

Michal

[abretaud]

Hi,
Is there any output produced in $PBS_O_WORKDIR? If it's working correctly, Maker should create a temp dir there with a lot of temporary files.
I tested it with slurm,no idea if there's anything to do when using pbs
Do you have a way to check what Maker is doing? (eating cpu or not ? strace?)

[mictadlo]

Hi,
Maybe the problem could be caused by:

> qpeek 4702790.pbs
Possible precedence issue with control flow operator at /lustre/work-lustre/waterhouse_team/miniconda2/envs/maker/lib/site_perl/5.26.2/Bio/DB/IndexedBase.pm line 805.
STATUS: Parsing control files...
ERROR: Could not determine if RepBase is installed
--> rank=NA, hostname=cl4n008

Where do I install RepBase with this package?

Thank you in advance,

Michal

[mictadlo]

Hi,
I fixed the problem by using mpirun maker -mpi -RM_off -fix_nucleotides. However, I got a new problem:

DBD::SQLite::db do failed: near ".": syntax error at /lustre/work-lustre/waterhouse_team/miniconda2/envs/maker/bin/../lib/GFFDB.pm line 498, <$IN> line 170017.

What did I miss?

Thank you in advance,

Michal

[abretaud]

For the repbase error, see my answer in #16501
For the DBD::SQLite::db I've never seen it... Do you give some GFF as evidences in your CTL file?

[bgruening]

@mictadlo please open a new issue if this is not resolved :)

Thanks!