Determining Secondary Index Usage

This section describes how to determine whether a query is using secondary indexes, depending on the query execution path used.

Determine whether your query uses the OJAI Distributed Query Service by following the steps outlined at Determining the Query Execution Path for OJAI Queries. The following sections describe next steps depending on the execution path.

Simple OJAI Queries

Using the tracing described at Determining the Query Execution Path for OJAI Queries, you will see the following in the log output:

2017-07-17 17:35:59 TRACE MapRDBTableImplHelper:703 - Scan on Index: 
'testIndex', Primary Table is: '/tmp/testTable', Index Scan QueryCondition: '((field < {"$numberLong":101}))', 
Index Scan startRow: '\x0FZ', Index Scan stopRow: '\x0F\x89\xCA\x80\x00'

The "Scan on Index", highlighted in bold, indicates OJAI used a secondary index to process the query.

Queries Requiring the OJAI Distributed Query Service

The tracing output described at Determining the Query Execution Path for OJAI Queries contains a queryId, highlighted in bold:

2017-07-17 18:51:14 DEBUG DrillDocumentStream:120 - DocumentResultsListener[1].queryIdArrived(queryId = 2692966d-0888-96e2-fa09-0d9befcd3173 , 
sql string = select t.`$$ENC00NQYF6YJUL5UW45AAL5UWI`,t.`$$document` from dfs.`/tmp/testTable` t where (t.`l0_a4_int` = -92) order by t.`l0_a4_int` ASC,t.`_id` DESC)

Use this queryId to retrieve more information through the Drill Web Console, including the query plan selected by the OJAI Distributed Query Service. See Query Profile for details.