Access Query History

Screen overview

The query listing screen is divided into:

  1. Filter Zone.
  2. SQL query information.

SQL query list with main overview information

Field NameDescription
(Execution) DateTimestamp of the execution of the Query, in your local timezone
Progress

This column displays the status and the total execution time of the query.

Execution status:

  • GREEN: query completed successfully.
  • BLUE: query is still in progress.
  • RED: there was an error during the execution.

Please note that on mouse over the progress, a tooltip provides some more insight:

  • initialization time (time to parse the query, if more than 1ms)
  • pending time (time query was waiting for a slot, in case the maximum number of running queries is reached)
  • freezing time (in case the query was frozen due to Big Request feature activation)
UserUsername & IP address of the user that executes the SQL command
Source
  • Index(es) used to answer the query => Only when the query is a SELECT statement
    • When GREEN, the index contains all the columns of the query; Performance is optimum.
    • When ORANGE, the query fetches fine data from the disk.
SQL

SQL statement and number of results that this query generates

SQL query detailed information

When you click on SELECT statements, the user can access more information about the SQL query.

  1. SQL Query main overview information
  2. SQL Statement
  3. Progress
  4. Query Profile
  5. External Source Details
  6. (Used) Indexes

2. SQL Statement sub-panel

3. Progress sub-panel

The progress sub-panel displays the information regarding the progress of the query:

  • the number of results for a select, or number of inserted lines for a load
  • the number of tasks executed / to execute, for queries triggering sub-tasks
  • the progress indicator for queries when it is still in progress

4. Query Profile sub-panel

The query profile displays the Query identifier. This identifier is also available in queries.csv log file, to identify a query.

Commun Metrics

There are three common metrics:

  • memory used by the master node
  • memory used by all nodes (incl. master node)
  • network byte sent
Metric "Lines read from K-store"

When the engine needs to get detailed rows to answer the query, the engine provides information about the number of lines that have been scanned.

Metric "Spill To Disk"

When a query spills on a disk, Query Profile shows you how much memory Indexima has allocated to local storage to perform the query. 

More information on Spill-to-Disk is available here.

5. External source Details sub-panel

When a query is delegated to an external source (using an external table), the engine displays information about the external source & query.

More information on External tables is available here.

NameDescription
Delegation period

Duration to execute the query on the external data source

Query delegated executionPercentage of delegation period to send the query and wait for execution 
Rows fetching from sourcePercentage of delegation period to retrieve results
External sourceName of the external data source
TableTarget table in the data source  
Lines fetched from external sourceNumber of results fetched from the data source

6. (Used) Indexes

The Indexes' sub-panel displays the detail of indexes used to execute the current query.
Indexes used by the current queries are displayed in green whereas the unused indexes are displayed in grey.

Filter queries

To facilitate research, queries can be filtered. 

  • You can combine multiple filters.
  • You can go through the "quick filters" or type in the search bar directly.


The Select date field allows to select a date in the past, or to limit the time interval to analyze.

Please note that the day and the time interval are considered with the local timezone of your browser (whereas the timestamp in queries.csv log files are in UTC).

Please note that the query listing won't retrieve more than 10000 queries. In case the selected date / timeInterval contains more than 10k queries, you will see a notification that the results are partial, and you will need to reduce your time interval.

For a complete analyse of a large volume of queries, we recommend to use directly the queries.csv log files.

By default, the search is only performed through the SQL statement.

When a user needs to filter queries attributes, he can use both quick (modifiable) filters and other custom filters.

Available filters

FilterOperatorAvailable filter valuesDescriptionQuick filterApplied filter when using a quick filter
user: (means =)Any stringName of the user executing the queryMy Queriesuser:admin
hybridQueries: (means =)true/falseQueries requesting unindexed data on diskHybrid QuerieshybridQueries:true
external: (means =)true/falseQueries delegated to an external databaseExternal Queriesexternal:true
responseTime:<, :>integer + ms / s / minResponse time of the queryResponse TimeresponseTime:>5000ms
resultsNb:<, :>integerNumber of resultsResultsresultsNb:>5000
inError: (means =)true/falseQuery in errorErrorsinError:true
diskUsage: (means =)true/falseQueries using disk access for any reason (Spill to disk, Disk backed results or BigIndex)Disk UsagediskUsage:true
spillToDisk: (means =)true/falseQueries using disk for temporary storage during execution

diskBackedResult: (means =)true/false

Queries using disk for output buffering 



bigIndex: (means =)true/false

Queries using disk to store index on disk



freezing: (means =)true/false

Queries currently frozen due to Big requests feature



freezeDuration:<, :>integer + ms / s / minQueries frozen more than this duration
Eg: freezeDuration:>10s 


memoryUsed

:<, :>integer + B / KB / MB / GBQueries using more than a certain amount of memory to execute
Eg: memoryUsed:>2GB or memoryUsed:>500MB 


indexUsed: (means =)Any stringQueries using specific indexes
Eg: indexUsed:tb1_v0,tb1_v1    (no space between index names)


cachedQueries: (means =)true/falseQueries using cache

ip:192.168.1.1: (means =)One IP address

Queries launched by a specific IP address




Display Events

You can also display the "events" by clicking on the button on the right. 
Note that you cannot filter events.

More information: see Events API.