fix "Failed" errors marked as "Completed"

[efigb]

When running qmap, all jobs were successfully completed but when running did not appear in "running" but in "pending" and then changed directly to "completed". @odoveestrella experienced the same problem. After checking the output this was the error:

[2116.colibactin] efiguerola@irbccn37:/data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin$ cat 20251128_colibactin_motif_20251201/000.err 
/var/spool/slurmd/job43392/slurm_script: line 10: ml: command not found
[2116.colibactin] efiguerola@irbccn37:/data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin$ cat 20251128_colibactin_motif_20251201/000.out 
[2116.colibactin] efiguerola@irbccn37:/data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin$ cat 20251128_colibactin_motif_20251201/000.info 
{
    "command": "Rscript   colibactin_motif.r   -i  /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/deepUMI/2025-08-05_rerun_Set1/mutations_vcf/TBEL_5549_COL_01.med.filtered.vcf  -o  /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin/results  -s  mutations_vcf",
    "status": "COMPLETED",
    "retries": 0,
    "executor_id": "43392",
    "job_cmd": "sbatch --parsable -c 2 --mem 30G -t 0-4 -D /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin -J colibactin.None -p bbg_cpu_zen4 --time 1440 -o /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin/20251128_colibactin_motif_20251201/000.out -e /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin/20251128_colibactin_motif_20251201/000.err /data/bbg/nobackup2/prominent/colon_duplex/rosana_collab/colibactin_motif/2025-11-28_colibactin/20251128_colibactin_motif_20251201/000.sh",
    "usage": {
        "disk": {
            "read": "0.01M",
            "write": "0.00M"
        },
        "memory": "0",
        "time": "00:00:01",
        "cluster": {
            "type": "SLURM",
            "nodes": "irbccn31"
        }
    },
    "exit_code": null,
    "error_reason": null

This is the header of my jobs file:

[pre]
ml load anaconda3
conda activate colibactin
[params]
cores = 2
memory = 30G
[jobs]

@migrau @FedericaBrando @CarlosLopezElorduy

[FedericaBrando]

The job is completed because it doesn't return any error from the command itself (usually is exit 1) because it happens in the pre-script.

If you login to that specific node, are you able to run ml load anaconda3 ? Maybe you don't have the alias and you need to specify the full command module load anaconda3.

[efigb]

I noticed when I'm in a screen with interactive module load anaconda3 does not work, while running this in the irblogin03 works. The way I could fix all of this and successfully run qmap:

Put this in my qmap job file:

[pre]
ml load anaconda3/2023.09-0-54w2amg
conda activate conda_env_needed

Then to run qmap in a screen, I did not use interactive (in interactive spack load py-qmap did not work either), so that I could load a conda environment where I have qmap installed. To do so: first I loaded the specific module to run conda that I obtained running module avail anaconda3. Then basically I activated the conda environment containing qmap and run the qmap jobs file:

ml load anaconda3/2023.09-0-54w2amg
conda activate qmap_env
qmap submit qmap_jobs_file.qmap 

Then all jobs were successfully done with qmap.

Hope this helps solving this!