nslurm v1.1.5

nSlurm

node wrapper to slurm scheduler node jobManager_test.js -f ../config/arwenConf.json -e emulator

JOB OBJECT

A job is an Object. It contains variables and functions.

JOB ID

A job is defined by its input(s) and the jobID.json file. The jobID.json file is like :

{
  'script' : '/path/to/a/777_coreScript.sh',
  'exportVar' : {
    'flags' : ' --option1 --option2 ',
    'moduleScript' : '/path/to/this/script'
  },
  'modules' : ['blast'],
  'tagTask' : 'blast' // 'blast' is just an example. It can be 'clustal', 'naccess', etc.
}

where :

• script is a path to the _coreScript.sh (see the JOB CACHE CONTENT part)
• exportVar is a JSON
• modules is an array
• tagTask

JOB VARIABLES (into job.js)

• engineHeader = written into the .sbatch file, containing all the variables to parameter the scheduler
• submitBin = path to the binary file of the scheduler, to run the .sbatch file
• cmd = command to test the .sbatch (like echo "toto" for example)
• script = path to the source of the _coreScript.sh (to be copied into the cache directory)
• exportVar = JSON of variable to export. For example : {'myVar': 'titi', 'aFile': '/path/'}
• inputs = JSON of the inputs and their name. For example : {'nameInput1': 'contentInput1', 'nameInput2': 'contentInput2'}
• tagTask = specific to taskObject
• emulated = true if the job is emulated (on your proper computer), false if the job is run with a scheduler
• port = port to communicate via Netcat
• adress = adress of the machine that will run the job (IP adress)
• workDir = directory of the job (see the JOB CACHE CONTENT part)
• namespace = directory of the pipeline of the job
• cwd <> =
• cwdClone <> =
• ttl <> = deprecated ???
• ERR_jokers = counter for jobs emitting an error -> can be re-submitted 2 times more
• MIA_jokers = "Missing In Action" = counter for missing job in queue
• modules [] = array of the module to load into the _coreScript.sh
• debugBool =
• inputSymbols <> =

JOB FUNCTIONS (into job.js)

• start() =
• getSerialIdentity() = to create the jobID.json
• submit() = create a process to execute the .sbtach

JOB CACHE CONTENT

When a job is created, the JM create a directory where all files related to this job will be written. The minimalist content (where 777 is a uuid) :

• 777_coreScript.sh
• 777.batch
• jobID.json
• 777.err
• 777.out
• input , containing :
  • myInput1.inp
  • myInput2.inp Here, the job needs two input files, named myInput1.inp and myInput2.inp.

Note These two names (without the extension) are described as variables into the 777.batch file, to indicate their paths. Thus, the path to the input files of the job are known in 777.batch.