Customer Responsibilities - Post Go Live
Oracle recommends that customers configure batch job streams to run the following batch jobs on a regular basis to support ILM:
• ILM Automation - New Subretention Value (F1-ILMSV): This process creates sub-partitions for Initial Measurements, Activities, and Device Events, based on specific sub-retention values defined in the ILM Configuration - MDM Sub-retention master configuration.
• ILM Automation - Measurement Compression (K1-ILMMC): This process changes the compression level from Advanced to Hybrid Columnar for Measurement data that is at least two months old which results in significant disk space saving. This compression job reorganizes the data and improves query performance. Parameters used by this process include: Table Owner, Degree of Parallelism, and Estimated Time Limit. This batch control should be part of monthly batch job streams so that its run once per month and during off peak / non-critical hours.
K1-ILMMC Sample Parameters:
Note: This batch job should be run with DIST-THD-POOL batch parameter with value of NOCACHE and during off peak hours.
• ILM Automation - Non-Measurement Compression (K1-ILMNC): This process changes the compression level from Advanced to Hybrid Columnar for non-Measurement partitions at least two months old and results in significant disk space savings. Parameters used by this process include: Table Owner, Maintenance Object Codes (used to define specific maintenance objects for compression), Degree of Parallelism, and Estimated Time Limit. This batch control should be part of monthly batch job streams so that its run once per month and during off peak / non-critical hours.
If the K1-ILMNC batch process is submitted without the maintenance object parameter defined, it compresses all maintenance objects except D1-IMD. In order to compress D1-IMD, submit a separate batch job and specify D1-IMD as the Maintenance Object parameter.
K1-ILMNC Sample Parameters:
Note: This batch job should be run with DIST-THD-POOL batch parameter with value of NOCACHE and during off peak hours.
• ILM Crawler Initiator (F1-ILMIN): The ILM Crawler Initiator process initiates the individual ILM Crawler batch controls as defined by the ILM Crawler Batch Control and ILM Retention Period in Days options on each maintenance object. his batch control should be part of monthly batch job streams so that its run once per month and during off peak / non-critical hours.
Individual ILM crawler batch controls invoke eligibility algorithms to determine if records are ready to be archived. Many maintenance objects use the ILM Eligibility Based on Status (F1-ILMELIG) algorithm to determine eligibility.
F1-ILMIN Sample Parameters:
Note: This batch job should be run with DIST-THD-POOL batch parameter with value of NOCACHE and during off peak hours.
• The ILM Automation - Drop Partition (K1-ILMDR) batch control provides the self-service capability to drop partitions for any ILM-applicable maintenance object without the need to raise a service request. Drop ILM partition support is also available for empty and old partitions for the all ILM-applicable maintenance object.
Customers using previous releases should run the following batch process in sequence:
a. Run the batch D1-IMDCL (ILM Crawler - IMD). This batch will update the ILM Archive Switch for drop partition eligibility. Guidelines for running this are as follows:
• If the requirement is to retain the IMD data from more than 3 months (which is the default), make sure to override the default retention value on ILM Master configuration.
• Don't run the batch with Override Cutoff Date batch parameter. Leave it blank to use current date so that batch should mark the IMD records based on the default retention value.
• Batch supports multi-threading, so run it with maximum available threads.
• Run the batch in the off peak hours.
• Expect to run this batch longer (may be hours), depending upon the number of IMDs.
b. Run the batch K1-ILMDR (ILM Automation - Drop Partition) that supports one month partition at a time. Guidelines for running this are as follows:
• Provide the following batch parameters:
• tableOwner: CISADM
• partitionMonth: First three letters of the oldest partition (for example Jan, Feb)
• partitionYear: Year associated with the oldest partition (for example 2022)
• tableNames: D1_INIT_MSRMT_DATA
• Run the batch in the off peak hours. Make sure that no other batches are running that are using D1_INIT_MSRMT_DATA and its child tables.
c. Raise a Service Request, in case of any issues and provide the following:
• ILM Master Configuration showing default retention period and D1-IMD retention period.
• Batch submission screen shots
• Batch run screen shots
• Both STDERR and STDOUT
d. Run these batches in any of your test environment first before running it in the Prod.
• Currently archiving is not supported. Only dropping existing partitions with or without data is fully supported.
Refer to Information Lifecycle Management in the Administrative User Guide for more details regarding ILM.