Create a Checkpoint Using the CGE UI

Procedure for creating checkpoints on a database via the CGE Checkpoint interface

When a database is started for the first time its initial state is considered to be a checkpoint. When a change is made to the database, its state can be preserved by creating a checkpoint. This preserves a copy of the previous in-memory database. Creating a checkpoint creates a persistent record of the database state, which is written to the database directory in a file named export_dataset.nq.
Note: Checkpoints can only be created on running databases. If there are any queries or updates executing, it important to ensure that they finish executing before a checkpoint is created, otherwise the state of the database in the checkpoint may not contain the desired updates to it.
  1. Optional: Log on to the CGE UI by pointing a browser at http://machine-login1:3756/login, entering credentials and then selecting the Login button.
    This step is optional, depending on how the interface has been configured
  2. Access the Checkpoint Interface using one of the following mechanisms:
    • Point the browser at http://machine:3756/dataset/checkpoint, where machine is the machine running CGE’s web server.
    • Select Checkpoint from the Data Access drop down.
    This brings up the Checkpoint Interface, as shown below:
    Figure: Creating a Checkpoint
  3. Specify a location for the checkpoint in the Checkpoint Location field. This is the directory where the checkpoint will be saved. The server will generate an error if this directory does not exist or is read-only. 
  4. Select the Create Checkpoint button to create the checkpoint. This will return a success/failure message as appropriate, as shown in the following example output:
    Checkpoint created at /lus/scratch/cge/datasets/lubm/0/temp