Error: FARMIT failed to accept job

[Abby Feng]

Hello,

I'm trying to run Trinity Phase 2 with HPC GridRunner. However, I ran into an error of  "FARMIT failed to accept job". The scripts I used are shown below:

1. SLURM.conf 

```

[GRID]
# grid type:
gridtype=SLURM
grid=SLURM

# template for a grid submission
cmd=sbatch --partition=cluster_short --mem=5G --ntasks=1 --cpus-per-task=1 --time=4:00:00

#############################################################################
# settings below configure the Trinity job submission system, not tied to the grid itself.
#############################################################################

# number of grid submissions to be maintained at steady state by the Trinity submission system
max_nodes=1000

# number of commands that are batched into a single grid submission job.
cmds_per_node=100

```

2. trinity-phase2_test_V2.run

```

#!/bin/bash -e
#SBATCH -p cluster_short         # partition name
#SBATCH --job-name=trinity-phase2grid-test-V2
#SBATCH --time=4:00:00      # maximum run time
#SBATCH --ntasks=1            # always 1
#SBATCH --cpus-per-task=1    # number of threads to use for Trinity
#SBATCH --mem=350G            # maximum memory available to Trinity

# load Trinity and HPC GridRunner
module load singularity

singularity exec -e /work/trinityrnaseq_latest.sif Trinity \
--CPU ${SLURM_CPUS_PER_TASK} --max_memory 20G \
--grid_exec "/work/download/HPC_GridRunner/HpcGridRunner-1.0.2/hpc_cmds_GridRunner.pl --grid_conf ${SLURM_SUBMIT_DIR}/SLURM_V2.conf -c" \
--genome_guided_bam /work/Aligned.sortedByCoord.out.bam \
--genome_guided_max_intron 50000 \
--SS_lib_type RF \
--output /work/trinity_test \
--monitoring

```

3. run command

```

sbatch trinity-phase2_test_V2.run

```

4. error messages

```

--------------------------------------------------------------------------------
------------ Trinity Phase 2: Assembling Clusters of Reads ---------------------
------- (involving the Inchworm, Chrysalis, Butterfly trifecta ) ---------------
--------------------------------------------------------------------------------



        *** Dispatching parallel commands to the compute farm:
Thursday, September 29, 2022: 11:00:00  CMD: /work/download/HPC_GridRunner/HpcGridRunner-1.0.2/hpc_cmds_GridRun
ner.pl --grid_conf /work/trinity_test/SLURM.conf -c trinity_GG.cmds
$VAR1 = {
          '[GRID]' => undef,
          'cmds_per_node' => '100',
          'grid' => 'SLURM',
          'cmd' => 'sbatch --partition=cluster_short --mem=5G --ntasks=1 --cpus-per-task=1 --time=4:00:00 ',
          'max_nodes' => '1000'
        };

FARMIT failed to accept job: sbatch --partition=cluster_short --mem=5G --ntasks=1 --cpus-per-task=1 --time=4:00:00  -e
/work/trinity_test/farmit.J2321222.cc21cluster21.2321222.1664416800/cmds/J2321222.S0.sh.st
derr -o /work/trinity_test/farmit.J2321222.cc21cluster21.2321222.1664416800/cmds/J2321222.
S0.sh.stdout /work/trinity_test/farmit.J2321222.cc21cluster21.2321222.1664416800/cmds/J232
1222.S0.sh 2>&1
 (ret -1)
FARMIT failed to accept job.  Will try again shortly.

```

Thanks for any help!

Best regards,

Abby

[Brian Haas]

Hi,

When running Trinity using singularity and also using a compute farm for dispatching the 2nd phase jobs, I've done it like so:

Trinity --seqType fq --max_memory 2G \

     --left reads.left.fq.gz --right reads.right.fq.gz \

     --SS_lib_type RF --CPU 4 \

     --grid_exec '/path/to/HpcGridRunner/hpc_cmds_GridRunner.pl \

     --grid_conf /path/to/HpcGridRunner/hpc_conf/BroadInst_UGER.conf -c ' \

     --output trinity_outdir_uger_simg_intra \

     --singularity_img /path/to/trinityrnaseq.simg

with my grid config as

===

[GRID]

# grid type: 

gridtype=SGE

# template for a grid submission

cmd=qsub -cwd -q broad -b y -l h_vmem=10g -pe smp 1 -l h_rt=240:00:00

# number of grid submissions to be maintained at steady state by the Trinity submission system 

max_nodes=100

# number of commands that are batched into a single grid submission job.

cmds_per_node=1

===

Note, for it to work this way, trinity needs to be installed (or at least just pulled down from github and not built) so it can kick off the work. All computes end up using the singularity image as specified in the Trinity command.

Hope this helps,

~b

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trinityrnaseq-users/1b590c24-d674-4e38-a610-430d1998a12bn%40googlegroups.com.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas