Batch Processing

UCloud batch processing applications are designed to run based on available resources without requiring end-user interaction. Users should select the relevant set of mandatory and/or optional parameters as if the executable program were initiated in a terminal window. For instance, consider the following command line:

$ kallisto quant -i transcripts.idx -b 30 -o results --genomebam -g transcripts.gtf -c chrom.txt reads_1.fastq reads_2.fastq

This same command can be executed as an instance of the Kallisto: quant application, which is available in the UCloud apps catalog. The program options for this specific case can be configured manually from the job submission page or imported via a JSON file, such as the one provided below:

    {
        "siteVersion": 2,
        "request": {
            "application": {
              "name": "kallisto-quant",
              "version": "0.45.0-5"
        },
        "product": {
              "id": "u1-standard-2",
              "category": "u1-standard",
              "provider": "ucloud"
        },
        "name": "Demo",
        "replicas": 1,
        "allowDuplicateJob": false,
        "parameters": {
              "input1": {
                "path": "/home/data/reads_1.fastq",
                "readOnly": false,
                "type": "file"
              },
              "input2": {
                "path": "/home/data/reads_2.fastq",
                "readOnly": false,
                "type": "file"
              },
              "i_var": {
                "path": "/home/data/transcripts.idx",
                "readOnly": false,
                "type": "file"
              },
              "o_var": {
                "value": "results",
                "type": "text"
            },
              "b_var": {
                "value": 30,
                "type": "integer"
              },
              "genomebam_flag": {
                "value": true,
                "type": "boolean"
              },
              "g_var": {
                "path": "/home/data/transcripts.gtf",
                "readOnly": false,
                "type": "file"
              },
              "c_var": {
                "path": "/home/data/chrom.txt",
                "readOnly": false,
                "type": "file"
              }
        },
        "resources": [],
        "timeAllocation": {
              "hours": 1,
              "minutes": 0,
              "seconds": 0
        },
        },
        "machineType": {
            "cpu": 2,
            "memoryInGigs": 12
        }
    }

In this example, the input files reads_1.fastq, reads_2.fastq, transcripts.idx, transcripts.gtf, and chrom.txt are located within the /home/data path on UCloud. The directory results is created after the run is completed in the job output folder.

Note

Batch jobs are completed as soon as the executable program terminates, provided the overall execution time is shorter than the job lifetime and sufficient resources are available throughout the runtime.