Package com.ibm.fhir.jbatch.bulkdata.tools

Class Main

  • public class Main
extends Object
    Tool to break large COS file into multiple ones. This allows us to use multiple partitions (one for each generated COS file) for the Bulkdata import job to import the same type of FHIR resources in parallel to achieve better performance.

    • Constructor Detail

      • Main

        public Main()

    • Method Detail

      • parseArgs

        public void parseArgs​(String[] args)
        Parse the command line arguments --cosapikey: the IBM COS API key or S3 access key. --cossrvinstid: the IBM COS service instance id or S3 secret key. --cosendpointurl: the IBM COS or S3 End point URL. --coslocation: the IBM COS or S3 location. --cosbucketname: the IBM COS or S3 bucket name to import from. --coscredentialibm: if use IBM credential(Y/N), default(Y). --cosfiletobreak: the file(COS/S3 object) to break. --numberoffiles: how many pieces (COS/S3 objects) to break into. --numberoftotalresources: the total number of distinct FHIR resources we want to include in the target files. used only when segBySize is 'N'(by number of records). --segbysize: break the file by file size(default) or by number of records("N"). (1) if by file size, then we get the total file size from the object meta directly, and each piece with size close to (total-file-size/numberOfFiles). (2) if by number of records (a) with numberOfTotalResources, then each piece has (numberOfTotalResources/numberOfFiles) distinct FHIR resources. (b) without numberOfTotalResources, then we go thought the COS/S3 once to get the total record number, then each piece will have (total-record-number/numberOfFiles) lines in it.
        Parameters:
        args -

      • main

        public static void main​(String[] args)
        Main entry
        Parameters:
        args -