RESUME tells LOADPLUS whether to add data to an empty table space or to a table space that already contains data.
In addition to interactions with other options (which are documented with those options), the value that you specify for the RESUME option can affect the following aspects of your load job:
When invoking DSNUTILB, LOADPLUS passes this option to the IBM Db2 LOAD utility for processing.
NO
RESUME NO (the default) tells LOADPLUS that the table space is empty, or that certain partitions are empty. LOADPLUS terminates when either of the following conditions exists:
- The table space contains data and you do not specify the REPLACE option.
- You specify PART n REPLACE and a partition that you are not replacing contains data.
YES
RESUME YES indicates that the table space might or might not be empty. LOADPLUS adds the new data to any existing data.
Restriction
You cannot use this option with LOAD REPLACE. If you want to replace only specific partitions of a table space, specify RESUME YES INTO TABLE with the PART n REPLACE option. If you want to replace only specific tables of a segmented table space, specify RESUME YES INTO TABLE with the REPLACE option.
Additional considerations
Note the following information about RESUME YES:
- Unless you specify INDEX UPDATE or SKIPIX, LOADPLUS unloads the existing indexes and merges them with the new index keys from the input data. LOADPLUS then sorts the merged index data and uses that data to build organized indexes during the LOAD phase or the COMBINED phase.
LOADPLUS builds a new compression dictionary for the applicable table space or partition when all of the following conditions exist:
- You are loading an empty partitioned table space or partition.
- The table space is defined with the COMPRESS YES attribute.
- A compression dictionary does not currently exist for the table space or partition.
Important
For a partition-by-growth table space, if the first partition has a compression dictionary and any of the subsequent partitions are empty, LOADPLUS copies the dictionary from the first partition instead of building a new dictionary.
For all other LOAD RESUME YES jobs, LOADPLUS keeps any existing compression dictionary.
Important
If the existing dictionary is unusable, LOADPLUS continues but does not compress any rows.
| This option tells LOADPLUS whether to build or update an index when specifying RESUME YES. Restrictions The following restrictions apply to the INDEX option: - LOADPLUS does not build or update indexes when either of the following conditions exists:
- You are running an SQLAPPLY load.
- LOADPLUS is invoking DSNUTILB.
- If you specify SKIPIX SIX or SKIPIX NUSIX, LOADPLUS does not build or update the indexes that you are skipping.
| This option, which is the default, tells LOADPLUS to unload the existing indexes and merge them with the new index keys from the input data. LOADPLUS then sorts the merged index data and uses that data to build organized indexes during the LOAD phase or the COMBINED phase. | | This option tells LOADPLUS to add index entries to the existing indexes. With INDEX UPDATE, LOADPLUS does not reorganize indexes as it does when INDEX BUILD is in effect. Requirement When you specify INDEX UPDATE for a two-phase load, you must also specify UNIQUECHECK NO. If you specify UNIQUECHECK YES or CLUSTER, LOADPLUS fails and displays message BMC51419E. Restrictions The following restrictions apply to INDEX UPDATE: - LOADPLUS terminates if you are loading XML or LOB data.
- LOADPLUS changes INDEX UPDATE to INDEX BUILD when you specify TABLE tableName REPLACE.
- LOADPLUS invokes DSNUTILB when loading a temporal table and the following conditions also exist:
- A participating index is defined with BUSINESS_TIME WITHOUT OVERLAPS.
- Any SHRLEVEL value is in effect except SHRLEVEL CHANGE SQLAPPLY.
Additional considerations The following considerations apply to the INDEX UPDATE option: - BMC recommends that you limit the use of UPDATE to those cases where you are adding a small percentage of the total amount of existing data. If you are adding a large percentage, using UPDATE can impact optimal performance of the SQL that uses the index in processing.
- If you specify PART n REPLACE (unless you are replacing all partitions), LOADPLUS rebuilds the data-sorting indexes and updates the participating non-data-sorting indexes and nonpartitioned data-sorting indexes.
- When you specify INDEX UPDATE, LOADPLUS does not unload existing index records, so LOADPLUS cannot check for uniqueness in the PRELOAD phase. If duplicates exist, LOADPLUS detects them in the LOAD or COMBINED phase, loads the table space, deletes the duplicate records, and issues messages BMC50258E and BMC51477I. For a two-phase load, LOADPLUS does not write these duplicates to the discard data set and they do not count toward any limit that you specify with the DISCARDS option.
- When you specify INDEX UPDATE and a failure occurs during index update processing, you will need to recover your table space and indexes. If you attempt to restart, LOADPLUS terminates and issues message BMC51435S.
- When you specify INDEX UPDATE and index update processing completes, but the job fails in the LOAD phase, you must restart the job by specifying RESTART without PHASE. If you specify RESTART(PHASE), LOADPLUS terminates and issues message BMC51436E.
- (PTF BQU1929 applied) BMC AMI Load does not support INDEX UPDATE. If you invoke SmartSwitch and a job in LOADPLUS contains INDEX UPDATE, SmartSwitch ignores INDEX UPDATE while switching the job from LOADPLUS to BMC AMI Load.
|
|
