Background information

In standalone mode, the data migration job can be executed with one-click import:


  1. Run one-click import. Run import.bat in cmd.exe for Windows, and run bash for Linux.
  2. If you previously run this job, you are asked if you want to continue the job from the last breakpoint or if you want to run a new synchronization job. If you initiate a new data migration job, or have modified the synchronized source end/destination end, run the synchronization job again.
    Note After a job starts in Windows, a new cmd window appears showing the synchronization job in progress and the log. The job status in the old window is refreshed every 10 seconds. Do not close these two windows during the data migration process. In Linux, the preceding process is run in the background.
  3. When the job is complete, if a task failed, you are asked if you want to retry. Enter yto retry or n to skip this step and exit. To see why the upload failed, open the file master/jobs/local_test/failed_tasks/<tasktaskid>/audit.log and check the cause.
    After a job is submitted, the master splits the job into tasks, the workers run the tasks and the tracker collects the task statuses. After a job is completed, the ossimport directory contains the following:
    ├── bin
    │   └── ossimport2.jar    # The standalone version JAR
    ├── conf
    │   ├── local_job.cfg     # The job configuration file
    │   └──    # Configuration file of the system running parameters
    ├──            # The command line tool
    ├──             # One-click import script
    ├── logs
    │   ├── import.log        # Archive logs
    │   ├── job_stat.log      # Job status record
    │   ├── ossimport2.log    # Running log of the standalone version
    │   └── submit.log        # Job submission record
    ├── master
    │   ├── jobqueue                # Store jobs that have not been fully split
    │   └── jobs                    # Store the job running status
    │       └── local_test          # Job name
    │           ├── checkpoints     # The checkpoint record of the master splitting the job to tasks
    │           │   └── 0
    │           │       └── 034DC9DD2860B0CFE884242BC6FF92E7.cpt
    │           ├── dispatched      # Tasks that have been assigned to the workers but haven't been fully run
    │           │   └── localhost
    │           ├── failed_tasks    # Tasks that failed to run
    │           ├── pending_tasks   # Tasks that have not been assigned
    │           └── succeed_tasks   # Tasks that run successfully
    │               └── A41506C07BF1DF2A3EDB4CE31756B93F_1499744514501@localhost
    │                   ├── audit.log   # The task running log. You can view the error causes in the log
    │                   ├── DONE        # Mark of successful tasks
    │                   ├── error.list  # The task error list. You can view the error file list
    │                   ├── STATUS      # The task status marker file. The content is Failed or Completed
    │                   └── TASK        # The task description information
    └── worker      # Status of the task being run by the worker. After running, tasks are managed by the master
        └── jobs
            └── local_test
                └── tasks
    • For job running information, view log/ossimport2.log.
    • For the task failure cause, view master/jobs/${JobName}/failed_tasks/${TaskName}/audit.log.
    • For failed task files, view master/jobs/${JobName}/failed_tasks/${TaskName}/error.list.