OpenSearch Industry Algorithm Edition creates an image for uploaded data and data that is synchronized from data sources. If you modify the schema of an application or upload full data to the application, you must perform reindexing. The following four scenarios trigger reindexing:
Manual reindexing: If you modify the schema of an application or upload full data to the application, you can manually trigger reindexing.
Scheduled reindexing: Full data is synchronized from a MaxCompute data source to an application every day.
Modify Offline Application: If you modify the schema of an application or modify index or attribute fields, OpenSearch triggers reindexing.
Clear data or Clear documents: If you clear data or documents, OpenSearch triggers reindexing.
Scheduled reindexing
After you create a scheduled reindexing task, OpenSearch performs reindexing once at a specified time every day.
1. Log on to the OpenSearch Industry Algorithm Edition console. On the Instance Management page, find the instance for which you want to create a scheduled reindexing task, and click Details in the Actions column.
2. Click Reindex.
3. Specify a time when the scheduled reindexing task is triggered. The minimum interval at which the scheduled reindexing task is triggered is 24 hours. This means that the scheduled reindexing task can be triggered once a day.
Notes:
An application can have up to two versions, which are the online version and the offline version. If an application has two versions, the offline version will be automatically deleted in a scheduled reindexing task. The offline version must be deleted in a manual reindexing task.
When the scheduled reindexing task is complete, the created offline version is published to replace the online version. OpenSearch automatically deletes the replaced version within 8 to 16 hours after the version is replaced.
Manual reindexing
To manually trigger reindexing, perform the following steps:
1. Log on to the OpenSearch Industry Algorithm Edition console. On the Instance Management page, find the instance for which you want to manually trigger reindexing, and click Details in the Actions column.
2. Click Reindex.
3. Select Reindex Only or Re-import Data and Reindex. Then, click OK.
Reindexing procedure and progress
Reindexing is performed in the following procedure: 1. Import data. 2. Process the data. 3. Build indexes.
After data is synchronized from a data source, OpenSearch processes the data and delivers the data to the build service for index building. If full data is synchronized, OpenSearch performs reindexing on a regular basis. The time that is required for each reindexing is 30 minutes. Therefore, the reindexing of full data takes some time even though only a small number of documents are involved. The OpenSearch console displays the progress of reindexing based on the status of each step.
Notes:
If both the online version and the offline version of an application exist, only the online version provides search services. When you use an application to search for data by specifying the application name in an API operation, the online version is used.
If both the online version and the offline version of an application exist, the functions of offline change, reindexing, clear data, and clear documents will be unavailable. In this case, you can only publish either of the online version or the offline version and delete the offline version that you no longer need.
After you delete an application, you cannot recover it.
If you select Reindex Only for a manual reindexing task, only incremental data is synchronized. If you select Re-import Data and Reindex for a manual reindexing task, both incremental data and full data are synchronized.