Elastic Path Commerce for AWS Setup
Core Commerce Architecture
Cortex API Architecture
Tutorials & Code Examples
Elastic Path Production Tools
Core Commerce Release Notes
Cortex API Front-End Documentation
EP for Adobe Marketing Cloud Architecture
EP for AMC Developer Documentation
EP for AMC Release Notes
Cortex Studio
Scheduled Jobs (Batch Server)
Scheduled Jobs (Batch Server)
The batch server uses scheduled jobs to perform certain operations, such calculating top selling products and product recommendations. The quartz.xml file contains settings that configure how and when these jobs run. It is located in the batch/ep-batch/src/main/resources/spring/scheduling directory.
Some of these jobs can be configured to run at different intervals or to clear different amounts of data. For more information on how to do this, see Configuring System Settings.
Batch Server Quartz Jobs
The batch server has the following jobs:
| Job Name | Purpose | Default Reoccurance | Assosciated System Environment Variables |
|---|---|---|---|
| cleanupAbandonedCartsJob | Purges abandoned shopping carts. | 30 minutes | COMMERCE/STORE/ABANDONEDCARTCLEANUP/maxHistory, COMMERCE/SYSTEM/ABANDONEDCARTCLEANUP/batchSize |
| cleanupAnonymousCustomerJob | Purges anonymous customers. | 5 minutes | COMMERCE/STORE/ANONYMOUSCUSTOMERCLEANUP/maxHistory, COMMERCE/SYSTEM/ANONYMOUSCUSTOMERCLEANUP/batchSize |
| cleanupFailedOrdersJob | Purges failed orders. | 45 minutes | COMMERCE/SYSTEM/FAILEDORDERCLEANUP/maxHistory, COMMERCE/SYSTEM/FAILEDORDERCLEANUP/batchSize |
| cleanupExpiredOAuth2TokensJob | Purges expired OAuth tokens. | Daily at midnight. | None, configured in quartz.xml. |
| importJobCleanupProcessorJob | Removes completed import jobs from the database based on their age. | 172800 seconds (default max. age), 5 minutes (default timeout) | COMMERCE/SYSTEM/IMPORT/importJobMaxAge, COMMERCE/SYSTEM/IMPORT/staleImportJobTimeout |
| inventoryJournalRollupJob | Rolls up inventory journal. | Delayed for 30 seconds upon batch server launch, then every 24 hours. | None, configured in quartz.xml. |
| processImportJob | Processes import jobs. | Waits for 10 seconds, and repeats every 5 seconds. | None, configured in quartz.xml. |
| productRecommendationJob | Re-computes product recommendations. | 30 seconds | None, configured in quartz.xml. |
| releaseShipmentsJob | Calls OrderService.updateOrderShipmentStatus() to release order shipments. | Hourly, at the start of the hour (i.e. 1:00pm). | None, configured in quartz.xml. |
| searchTermsAggregatorJob | Aggregates search terms. | Delayed for 30 seconds upon batch server launch, then every 30 seconds. | None, configured in quartz.xml, CE/pom.xml, CE/pom.xml. searchTermsAggregatorJob properties are generally set during the build process, not runtime. |
| staleImportJobProcessorJob | Processing stale import jobs. | Delayed for 30 seconds upon batch server launch, then every minute. | None, configured in quartz.xml. |
| topSellerJob | Determines the latest top selling products for testing and demos. | Delayed for 30 seconds, then every 30 seconds. | None, configured in quartz.xml, CE/pom.xml, CE/pom.xml. topSellerJob properties are generally set during the build process, not runtime. |
Computing Product Recommendations
A trigger is declared for computing product recommendation data. A SimpleTriggerBean is typically used to compute recommendations every 30 seconds for demonstration purposes. For production, use a CronTriggerBean to compute recommendations with a longer interval (for example, an interval of 24 hours).
Running the product recommendations job frequently on a large catalog with many orders can have a negative impact on performance. In a busy production environment, it is recommended that you schedule it to run at intervals of one hour (3600000 milliseconds) or more.
Computing the Top Sellers
A trigger is declared for computing the top selling product statistics. A SimpleTriggerBean is typically used to compute recommendations every 30 seconds for demonstration purposes. For production, use a CronTriggerBean to compute recommendations with a longer interval (for example, an interval of 24 hours).
Running the top sellers job frequently on a large catalog with many orders can have a negative impact on performance. In a busy production environment, it is recommended that you schedule it to run at intervals of one hour (3600000 milliseconds) or more.