Splunk® Enterprise

Search Manual

Splunk Enterprise version 9.0 will no longer be supported as of June 14, 2024. See the Splunk Software Support Policy for details. For information about upgrading to a supported version, see How to upgrade Splunk Enterprise.

Use the stats command and functions

This topic discusses how to use the statistical functions with the transforming commands chart, timechart, stats, eventstats, and streamstats.

About the stats commands and functions

The stats, streamstats, and eventstats commands each enable you to calculate summary statistics on the results of a search or the events retrieved from an index. The stats command works on the search results as a whole. The streamstats command calculates statistics for each event at the time the event is seen, in a streaming manner. The eventstats command calculates statistics on all search results and adds the aggregation inline to each event for which it is relevant. See more about the differences between these commands in the next section.

The chart command returns your results in a data structure that supports visualization as a chart (such as a column, line, area, and pie chart). You can decide what field is tracked on the x-axis of the chart. The timechart command returns your results formatted as a time-series chart, where your data is plotted against an x-axis that is always a time field. Read more about visualization features and options in the Visualization Reference of the Data Visualization Manual.

The stats, chart, and timechart commands (and their related commands eventstats and streamstats) are designed to work in conjunction with statistical functions. The list of statistical functions lets you count the occurrence of a field and calculate sums, averages, ranges, and so on, of the field values.

For the list of statistical functions and how they're used, see "Statistical and charting functions" in the Search Reference.

Stats, eventstats, and streamstats

The eventstats and streamstats commands are variations on the stats command.

The stats command works on the search results as a whole and returns only the fields that you specify. For example, the following search returns a table with two columns (and 10 rows).

sourcetype=access_* | head 10 | stats sum(bytes) as ASumOfBytes by clientip

The ASumOfBytes and clientip fields are the only fields that exist after the stats command. For example, the following search returns empty cells in the bytes column because it is not a result field.

sourcetype=access_* | head 10 | stats sum(bytes) as ASumOfBytes by clientip | table bytes, ASumOfBytes, clientip

To see more fields other than ASumOfBytes and clientip in the results, you need to include them in the stats command. Also, if you want to perform calculations on any of the original fields in your raw events, you need to do that before the stats command.

The eventstats command computes the same statistics as the stats command, but it also aggregates the results to the original raw data. When you run the following search, it returns an events list instead of a results table, because the eventstats command does not change the raw data.

sourcetype=access_* | head 10 | eventstats sum(bytes) as ASumOfBytes by clientip

You can use the table command to format the results as a table that displays the fields you want. Now, you can also view the values of bytes (or any of the original fields in your raw events) in your results.

sourcetype=access_* | head 10 | eventstats sum(bytes) as ASumOfBytes by clientip | table bytes, ASumOfBytes, clientip

The streamstats command also aggregates the calculated statistics to the original raw event, but it does this at the time the event is seen. To demonstrate this, include the _time field in the earlier search and use streamstats.

sourcetype=access_* | head 10 | sort _time | streamstats sum(bytes) as ASumOfBytes by clientip | table _time, clientip, bytes, ASumOfBytes

Instead of a total sum for each clientip (as returned by stats and eventstats), this search calculates a sum for each event based on the time that it is seen. The streamstats command is useful for reporting on events at a known time range.


Examples

Example 1

This example creates a chart of how many new users go online each hour of the day.

... | sort _time | streamstats dc(userid) as dcusers | delta dcusers as deltadcusers | timechart sum(deltadcusers)

The dc (or distinct_count) function returns a count of the unique values of userid and renames the resulting field dcusers.

If you don't rename the function, for example "dc(userid) as dcusers", the resulting calculation is automatically saved to the function call, such as "dc(userid)".

The delta command is used to find the difference between the current and previous dcusers value. Then, the sum of this delta is charted over time.

Example 2

This example calculates the median for a field, then charts the count of events where the field has a value less than the median.

... | eventstats median(bytes) as medbytes | eval snap=if(bytes>=medbytes, bytes, "smaller") | timechart count by snap

Eventstats is used to calculate the median for all the values of bytes from the previous search.

Example 3

This example calculates the standard deviation and variance of calculated fields.

sourcetype=log4j ERROR earliest=-7d@d latest=@d | eval warns=errorGroup+"-"+errorNum | stats count as Date_Warns_Count by date_mday,warns | stats stdev(Date_Warns_Count), var(Date_Warns_Count) by warns

This search returns errors from the last 7 days and creates the new field, warns, from extracted fields errorGroup and errorNum. The stats command is used twice. First, it calculates the daily count of warns for each day. Then, it calculates the standard deviation and variance of that count per warns.

Example 4

You can use the calculated fields as filter parameters for your search.

sourcetype=access_* | eval URILen = len(useragent) | eventstats avg(URILen) as AvgURILen, stdev(URILen) as StdDevURILen| where URILen > AvgURILen+(2*StdDevURILen) | chart count by URILen span=10 cont=true

In this example, eventstats is used to calculate the average and standard deviation of the URI lengths from useragent. Then, these numbers are used as filters for the retrieved events.

Last modified on 04 November, 2019
About calculating statistics   Use stats with eval expressions and functions

This documentation applies to the following versions of Splunk® Enterprise: 7.0.0, 7.0.1, 7.0.2, 7.0.3, 7.0.4, 7.0.5, 7.0.6, 7.0.7, 7.0.8, 7.0.9, 7.0.10, 7.0.11, 7.0.13, 7.1.0, 7.1.1, 7.1.2, 7.1.3, 7.1.4, 7.1.5, 7.1.6, 7.1.7, 7.1.8, 7.1.9, 7.1.10, 7.2.0, 7.2.1, 7.2.2, 7.2.3, 7.2.4, 7.2.5, 7.2.6, 7.2.7, 7.2.8, 7.2.9, 7.2.10, 7.3.0, 7.3.1, 7.3.2, 7.3.3, 7.3.4, 7.3.5, 7.3.6, 7.3.7, 7.3.8, 7.3.9, 8.0.0, 8.0.1, 8.0.2, 8.0.3, 8.0.4, 8.0.5, 8.0.6, 8.0.7, 8.0.8, 8.0.9, 8.0.10, 8.1.1, 8.1.2, 8.1.3, 8.1.4, 8.1.5, 8.1.6, 8.1.7, 8.1.8, 8.1.9, 8.1.13, 8.1.14, 8.2.0, 8.2.1, 8.2.2, 8.2.3, 8.2.4, 8.2.5, 8.2.6, 8.2.7, 8.2.8, 8.2.9, 8.2.10, 8.2.11, 8.2.12, 9.0.0, 9.0.1, 9.0.2, 9.0.3, 9.0.4, 9.0.5, 9.0.6, 9.0.7, 9.0.8, 9.0.9, 9.0.10, 9.1.0, 9.1.1, 9.1.2, 9.1.3, 9.1.4, 9.1.5, 9.1.6, 9.2.0, 9.2.1, 9.2.2, 9.2.3, 9.3.0, 9.3.1, 8.1.0, 8.1.10, 8.1.11, 8.1.12


Was this topic useful?







You must be logged into splunk.com in order to post comments. Log in now.

Please try to keep this discussion focused on the content covered in this documentation topic. If you have a more general question about Splunk functionality or are experiencing a difficulty with Splunk, consider posting a question to Splunkbase Answers.

0 out of 1000 Characters