Star2Pass change in memory requirements triggers reexecution

[Alejandra Cervera]

If one job fails due to not enough memory, if memory is changed then all jobs are reexecuted since configuration has changed. Can you take a look and see if we can circumvent this somewhow? thanks

[Ville Rantanen]

Running first this script:

 val r=INPUT("../../anduril2/bundles/sequencing/functions/Star2Pass/testcases/case1/input/reads/").out
 val m=INPUT("../../anduril2/bundles/sequencing/functions/Star2Pass/testcases/case1/input/mates/").out
 val ref=INPUT("../../anduril2/bundles/sequencing/functions/Star2Pass/testcases/case1/input/reference").out
 val z=Star2Pass(reads=r, reference=ref, 
            genomeLoad="NoSharedMemory", 
            readFilesCommand="zcat",                                            
            threads=10, memory=20000,
            genomeParameters="--genomeSAindexNbases=5")

And then changing the memory to 40000:

 val z=Star2Pass(reads=r, reference=ref, 
            genomeLoad="NoSharedMemory", 
            readFilesCommand="zcat",                                            
            threads=10, memory=40000,
            genomeParameters="--genomeSAindexNbases=5")

does NOT trigger re-execution.

Cannot make test case where this behavior would show.

[Ville Rantanen]

Also, not re-executing with paired (mates) use case.

[Alejandra Cervera]

True enough. In your example it doesn't trigger reexecution. in one of my pipelines it did, but i cannot risk to run it again

[Ville Rantanen]

the log file should tell where (CONFIGURATION CHANGED), if you still have it, post it here. (attach files in the More -button)