I was trying to run chimera.slayer as follows and an error was thrown:
mothur > chimera.slayer(fasta=input.align, template=silva.gold.align, window=100, processors=6)
Checking sequences from input.align ...
Reading sequences from silva.gold.align...Done.
Reading sequences from input.align...Done.
Only reporting sequence supported by 90% of bootstrapped results.
[ERROR]: unable to spawn the necessary processes.
I will try again on one processor to see if it still dies.
I forgot to mention before that I am using v 1.15 on a RedHat 4.1.2 system.
As expected, trying the command with only one processor is working.
We found a memory leak in the chimera.slayer command. It occurs when the template sequences that are chosen to be potential parents do not overlap well with the query sequences. It occurs whether you are using 1 or 6 processors, but using multiple processors uses more RAM than using 1 processor. In your case, it’s probably just enough more to cause the spawn to fail. It will be fixed in the 1.16.0 release later this month.