Upgrading a dataset schema

When a dataset requires a schema upgrade, CM Enterprise Studio displays Schema upgrade required behind it in the list of datasets.

1. Expand Datasets.

   If the dataset is not listed, you need to register it.

   See Registering an existing Content Manager dataset.

2. Right-click the dataset and from the Schema menu, select Upgrade.

   The Upgrade Version dialog box appears.

   If your dataset schema is already up-to-date, the dialog box will display the message: Database is already at current version - no upgrade required.

   • Dataset Name - the current dataset you are viewing
   • Dataset Identifier - the unique ID for the current dataset
   • Upgrading Dataset - the From field contains the current schema number, the To field contains the new dataset schema number
   • Number of upgrade processing threads - To improve the speed of upgrade, you can specify a number of threads that will be used to run various update steps concurrently. The upgrade process is divided into approximately 8 phases, within each of these phases, various upgrade work items can be run in parallel.
   • Support GIS columns and spatial indexes - If you are upgrading from a schema prior to that when GIS features were introduced, you can choose whether you wish the upgrade to enable these features as part of the process. If you choose not to enable them, there is a separate option you can run later to turn them on.
   • Count document viewed events to initialize record view counter - select this option to get the schema upgrade to initialize this view counter based on the number of Document Viewed events stored in the online audit log. This will only be useful if your database is currently auditing document view activities for all record types. For large datasets, as this initialization step could take quite a long time, it is recommended that this process is run after upgrading using the Initialize Document Views Counter option.
   • Limit how many rows can be updated at any one time - Many of the schema upgrade functions have an impact on the transaction roll back logs of the database. If you don't have a large enough roll back log, it is possible that one of the SQL statements run as part of the upgrade will fail due to the filling up of one of the rollback log files. To prevent this, you can limit how many rows are affected by any single update statement. This means that updates will be limited to a smaller number of rows and therefore a larger number of update statements are needed to be executed. If you need to use this parameter it is likely you have a large database, so having a small number in this field will likely cause extended delays. Try with a minimum value of 1,000,000 or larger.
     • Use CTAS approach when rows to update exceeds - this is optional - CREATE TABLE AS SELECT (CTAS) allows for the restructuring of tables by copying them over to a new table. Specify a threshold value, past which the CTAS technique should be used. The specified number needs to be greater the number entered for the Limit how many rows can be updated at any one time option
   • Table Location - choose the table location, if required
   • Index Location - choose the index location, if required

   If a schema upgrade is being run on an Oracle database, the following options are available:

   • Block size when using DBMS_PARALLEL_EXECUTE.CREATE_CHUNKS_BY_ROWID - there is a special technique used in Oracle upgrades to split a large transaction up into pieces. This parameter determines the size of block to use for each thread that employs this technique.

   • Parallel level for DBMS_PARALLEL_EXECUTE command - this allows you to specify how many threads to use when running one of the special split Oracle transactions.

   • Use PARALLEL and NOLOGGING for CTAS/CREATE INDEX, with parallel_level - this is used for other Oracle upgrades and allows you to specify how many threads Oracle should use when running CREATE TABLE AS SELECT and CREATE INDEX commands.

3. Click OK to start the upgrade process.

   The Dataset Work In Progress window appears.

4. Select the Pause After option you want to apply:
   • Each Step - pause after each step is completed
   • Major Steps - pause after each major step is completed
   • Complete All Steps - complete all steps without pausing
5. Click Start to start the upgrade process.
6. Once all steps are completed, click Done to close the Dataset Work In Progress window.

   The new dataset appears in the list of datasets..
   After the repair, Content Manager Enterprise Studio displays a dialog with the path to the change log file. Click OK to view the change log.

7. After upgrading the dataset schema, deploy the new configuration to all running Workgroup Servers