Troubleshooting Indexing

This section explains how to troubleshoot exceptions that occur when indexing activities are carried out.

AnalyticsTableNotAvailableException

Sample Error Log	TID: [-1] [] [2018-02-13 04:49:40,345] ERROR {org.wso2.carbon.analytics.dataservice.core.indexing.IndexNodeCoordinator} - Error in processing staging index data: Error in processing index staging entry: [-1234:NEW_CUSTOMERS_BY_QUARTER_PRODUCT_AND_POD] does not exist {org.wso2.carbon.analytics.dataservice.core.indexing.IndexNodeCoordinator} org.wso2.carbon.analytics.datasource.commons.exception.AnalyticsException: Error in processing index staging entry: [-1234:NEW_CUSTOMERS_BY_QUARTER_PRODUCT_AND_POD] does not exist at org.wso2.carbon.analytics.dataservice.core.indexing.IndexNodeCoordinator.processStagingEntry(IndexNodeCoordinator.java:961) at org.wso2.carbon.analytics.dataservice.core.indexing.IndexNodeCoordinator.access$600(IndexNodeCoordinator.java:69) at org.wso2.carbon.analytics.dataservice.core.indexing.IndexNodeCoordinator$StagingDataIndexWorker.run(IndexNodeCoordinator.java:997) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)
Occurence	`When the INSERT OVERWRITE operation is carried out.`
Possible Reasons	The corruption of meta information and indexing that could have resulted from one of the following: Deployments or the removal of deployments in the nodes of the DAS setup. Configuration changes in the DAS nodes of the setup. Creating/deleting tables in the event store or the process store.
Recommended Action	To resolve this issue, reindex your data following the steps below: Shut down all the DAS nodes in your deployment. Truncate the following tables. Table names are encoded in characters that cannot be read by humans for all database types other than HBase and Cassandra. Therefore, if the database type is not HBase or Cassandra, you need to use the encoded versions of these table names depending on the database type when performing this step. `ANX_X1000__INDEX_STAGING_DATA_` `IDX_X1000__INDEX_STAGING_DATA_` Drop all the tables with names that start with `ANX` and `IDX` such as the examples given below. Table names are encoded in characters that cannot be read by humans for all database types other than HBase and Cassandra. Therefore, this step is relevant only if the database type used is HBase or Cassandra because the table names given below can only exist for these two database types. `ANX_X1000__INDEX_STAGING_DATA_` `IDX_X1000__INDEX_STAGING_DATA_` For each node, remove all the index data stored in the `<DAS_HOME>/repository/data` directory. Restart the DAS nodes.

Issues relating to the disk space

Occurence	Disk space related issues occur when a high volume of data is stored in the databases.
Possible Reasons	Tables with the ANX__8 prefix contain a huge amount of data and it takes the majority of the DB size. The known issue in Oracle databases where the previously used disk space cannot be reclaimed after deleting data from database tables
Recommended Action	Reindex your data by following the steps below: Shut down all the DAS nodes in your deployment. Delete all the contents of the `<DAS_HOME>repository/data` directory except the `local-shard-allocation-config.conf` file. In the `WSO2_ANALYTICS_EVENT_STORE_DB` database, delete all the tables with the `ANX__8` prefix. Open the <wso2das-3.1.0>/repository/conf/analytics/local-shard-allocation-config.conf file, and change the status of all the shards from `NORMAL` top `INIT`, so that the content looks as follows. 0,INIT 1,INIT 2,INIT 3,INIT 4,INIT 5,INIT If you find any other content in the `local-shard-allocation-config.conf` file, remove it. Restart the DAS nodes. Once the re-indexing is completed, the status of the shards in the `local-shard-allocation-config.conf` file changes from `INIT` to `NORMAL`. If you are using the Oracle database type, you need to shrink the database to reclaim the disk space previously used by the deleted data. To do this, follow the steps provided in this Oracle community discussion.

For more information about indexing, see Indexing Overview.