Stop staging the 9.2 GB ChEMBL DB into every task workdir
GET_FINAL_METABOLITES_STATIC took chembl_db as a `path` input, so Nextflow copied the ~9.2 GB SQLite DB into each task's work directory. With ~29 concurrent tasks that is hundreds of GB of I/O, against a process that also declared only 1 GB of memory. Every task failed with exit 1 and retried ten times, producing 746 errors and an empty 1b_final_metabolites/ output. Pass the DB as a `val` path instead so tasks read it in place from the dreamdock-data PVC, and give the process parameterised growing memory.
This commit is contained in:
@@ -274,18 +274,23 @@ process GET_FINAL_METABOLITES {
|
||||
}
|
||||
|
||||
process GET_FINAL_METABOLITES_STATIC {
|
||||
memory 1.GB
|
||||
// The ChEMBL DB is ~9.2 GB. It is passed as a `val` path (not a staged `path`
|
||||
// input) so Nextflow does not copy it into every task work directory — with
|
||||
// concurrent tasks that multiplied into hundreds of GB of I/O. The DB lives on
|
||||
// the dreamdock-data PVC and is opened read-only/immutable, so sharing one copy
|
||||
// across tasks is safe.
|
||||
memory { params.chembl_initial_memory.toInteger().GB + (task.attempt - 1) * params.chembl_growth_memory.toInteger().GB }
|
||||
|
||||
container "${params.container_chembl}"
|
||||
containerOptions "${params.containerOptions}"
|
||||
publishDir "${params.outdir}/${params.project_name}/1b_final_metabolites", mode: 'copy'
|
||||
|
||||
errorStrategy { task.attempt <= 10 ? 'retry' : 'ignore' }
|
||||
maxRetries 10
|
||||
errorStrategy { task.attempt <= params.chembl_max_retries.toInteger() ? 'retry' : params.chembl_fail_action }
|
||||
maxRetries { params.chembl_max_retries.toInteger() }
|
||||
|
||||
input:
|
||||
path smiles_csv
|
||||
path chembl_db // Add ChEMBL SQLite database as input
|
||||
val chembl_db // Path to the ChEMBL SQLite DB on the PVC (not staged into the workdir)
|
||||
|
||||
output:
|
||||
path "${smiles_csv.simpleName}.txt"
|
||||
|
||||
Reference in New Issue
Block a user