Splunk: Exploring SPL
Last updated
Last updated
Splunk is a powerful SIEM solution that provides the ability to search and explore machine data. Search Processing Language (SPL) is used to make the search more effective. It comprises various functions and commands used together to form complex yet effective search queries to get optimized results.
This room will dive deep into some key fundamentals of searching capability, like chaining SPL queries to construct simple to complex queries.
Learning Objectives
This room will teach the following topics:
What are Search processing Language?
How to apply filters to narrow down results.
Using transformational commands.
Changing the order of the results.
Room Prerequisites
This room is based on the SIEM concepts covered in Intro to SIEM and Splunk: Basics rooms. Complete these rooms and continue to the next task.
Search & Reporting App is the default interface used to search and analyze the data on the Splunk Home page. It has various functionalities that assist analysts in improving the search experience.
Some important functionalities present in the search App are explained below:
Search Head: Search Head is where we use search processing language queries to look for the data.
Time Duration: This tab option provides multiple options to select the time duration for the search. All-time will display the events in real-time. Similarly, the last 60 minutes will display all the events captured in the last hour.
Search History: This tab saves the search queries that the user has run in the past along with the time when it was run. It lets the user click on the past searches and look at the result. The filter option is used to search for the particular query based on the term.
The Field Sidebar can be found on the left panel of Splunk search. This sidebar has two sections showing selected fields and interesting fields. It also provides quick results, such as top values and raw values against each field.
Some important points to understand about the sidebar are explained below:
2- Interesting Fields
Pulls all the interesting fields it finds and displays them in the left panel to further explore.
3- Alpha-numeric fields 'α'
This alpha symbol shows that the field contains text values.
4- Numeric fields '#'
This symbol shows that this field contains numerical values.
5- Count
The number against each field shows the number of events captured in that timeframe.
1) In the search History, what is the 7th search query in the list? (excluding your searches from today)
Answer: index=windowslogs | chart count(EventCode) by Image
2) In the left field panel, which Source IP has recorded max events?
Answer: 172.90.12.11
3) How many events are returned when we apply the time filter to display events on 04/15/2022 and Time from 08:05 AM to 08:06 AM?
Answer: 134
Splunk Search Processing Language comprises of multiple functions, operators and commands that are used together to form a simple to complex search and get the desired results from the ingested logs. Main components of SPL are explained below:
Splunk field operators are the building blocks used to construct any search query. These field operators are used to filter, remove, and narrow down the search result based on the given criteria. Common field operators are Comparison operators, wildcards, and boolean operators.
These operators are used to compare the values against the fields. Some common comparisons operators are mentioned below:
=
This operator is used to match values against the field. In this example, it will look for all the events, where the value of the field UserName is equal to Mark.
!=
This operator returns all the events where the UserName value does not match Mark.
<
Showing all the events with the value of Age less than 10.
<=
Showing all the events with the value of Age less than or equal to 10.
>
This will return all the events where the Outbound traffic value is over 50 MB.
>=
This will return all the events where the Outbound traffic value is greater or equal to 50 MB.
Lets use the comparison operator to display all the event logs from the index "windowslogs", where AccountName is not Equal to "System"
Search Query: index=windowslogs AccountName !=SYSTEM
Splunk supports the following Boolean operators, which can be very handy in searching/filtering and narrowing down results.
NOT
Ignore the events from the result where field_A contain the specified value.
OR
Return all the events in which field_A contains either value1 or value2.
AND
Return all the events in which field_A contains value1 and field_B contains value2.
To understand how boolean operator works in SPL, lets add the condition to show the events from the James account.
Search Query: index=windowslogs AccountName !=SYSTEM AND AccountName=James
Splunk supports wildcards to match the characters in the strings.
In the events, there are multiple DestinationIPs reported. Let's use the wildcard only to show the DestinationIP starting from 172.*
Search Query: index=windowslogs DestinationIp=172.*
1) How many Events are returned when searching for Event ID 1 AND User as James?
Answer: 4
2) How many events are observed with Destination IP 172.18.39.6 AND destination Port 135?
Answer: 4
3) What is the Source IP with highest count returned with this Search query? Search Query: index=windowslogs Hostname="Salena.Adam" DestinationIp="172.18.38.5"
Answer: 172.90.12.11
4) In the index windowslogs, search for all the events that contain the term cyber how many events returned?
Answer: 0
5) Now search for the term cyber*, how many events are returned?
Answer: 12256
Our network generates thousands of logs each minute, all ingesting into our SIEM solution. It becomes a daunting task to search for any anomaly without using filters. SPL allows us to use Filters to narrow down the result and only show the important events that we are interested in. We can add or remove certain data from the result using filters. The following commands are useful in applying filters to the search results.
Let's use the fields command to only display host, User, and SourceIP fields using the following syntax.
Search Query: index=windowslogs | fields + host + User + SourceIp
Use the search command to show all the events containing the term Powershell. This will return all the events that contain the term "Powershell".
Search Query: index=windowslogs | search Powershell
We can use the dedup command to show the list of unique EventIDs from a particular hostname.
Search Query: index=windowslogs | table EventID User Image Hostname | dedup EventID
Let's rename the User field to Employees using the following search query.
Search Query: index=windowslogs | fields + host + User + SourceIp | rename User as Employees
1) What is the third EventID returned against this search query?
Search Query: index=windowslogs | table _time EventID Hostname SourceName | reverse
Answer: 4103
2) Use the dedup command against the Hostname field before the reverse command in the query mentioned in Question 1. What is the first username returned in the Hostname field?
Answer: Salena.Adam
SPL provides various commands to bring structure or order to the search results. These sorting commands like head, tail, and sort can be very useful during logs investigation. These ordering commands are explained below:
Explanation
Each event has multiple fields, and not every field is important to display. The Table command allows us to create a table with selective fields as columns.
Syntax
| table <field_name1> <fieldname_2>
Example
| table
| head 20 # will return the top 20 events from the result list.
This search query will create a table with three columns selected and ignore all the remaining columns from the display.
Search Query: index=windowslogs | table EventID Hostname SourceName
Explanation
The head command returns the first 10 events if no number is specified.
Syntax
| head <number>
Example
| head # will return the top 10 events from the result list
| head 20 # will return the top 20 events from the result list
The following search query will show the table containing the mentioned fields and display only the top 5 entries.
Search Query: index=windowslogs | table _time EventID Hostname SourceName | head 5
Explanation
The Tail command returns the last 10 events if no number is specified.
Syntax
| tail <number>
Example
| tail # will return the last 10 events from the result list
| tail 20 # will return the last 20 events from the result list
The following search query will show the table containing the mentioned fields and display only 5 entries from the bottom of the list.
Search Query: index=windowslogs | table _time EventID Hostname SourceName | tail 5
Explanation
The Sort command allows us to order the fields in ascending or descending order.
Syntax
| sort <field_name>
Example
| sort Hostname # This will sort the result in Ascending order.
The following search query will sort the results based on the Hostname field.
Search Query: index=windowslogs | table _time EventID Hostname SourceName | sort Hostname
Explanation
The reverse command simply reverses the order of the events.
Syntax
| reverse
Example
<Search Query> | reverse
Search Query: index=windowslogs | table _time EventID Hostname SourceName | reverse
1) Using the Reverse command with the search query index=windowslogs | table _time EventID Hostname SourceName - what is the HostName that comes on top?
Answer: James.browne
2) What is the last EventID returned when the query in question 1 is updated with the tail command?
Answer: 4103
3) Sort the above query against the SourceName. What is the top SourceName returned?
Answer: Microsoft-Windows-Directory-Services-SAM
Transformational commands are those commands that change the result into a data structure from the field-value pairs. These commands simply transform specific values for each event into numerical values which can easily be utilized for statistical purposes or turn the results into visualizations. Searches that use these transforming commands are called transforming searches. Some of the most used transforming commands are explained below.
Command
top
Explanation
This command returns frequent values for the top 10 events.
Syntax
| top <field_name>
| top limit=6 <field_name>
Example
top limit=3 EventID
The following command will display the top 7 Image ( representing Processes) captured.
Search Query: index=windowslogs | top limit=7 Image
Command
rare
Explanation
This command does the opposite of top command as it returns the least frequent values or bottom 10 results.
Syntax
| rare <field_name>
| rare limit=6 <field_name>
Example
rare limit=3 EventID
The following command will display the rare 7 Image (Processes) captured.
Search Query: index=windowslogs | rare limit=7 Image
Command
highlight
Explanation
The highlight command shows the results in raw events mode with fields highlighted.
Syntax
highlight <field_name1> <field_name2>
Example
highlight User, host, EventID, Image
The following command will highlight the three mentioned fields in the raw logs
Search Query: index=windowslogs | highlight User, host, EventID, Image
STATS Commands
SPL supports various stats commands that help in calculating statistics on the values. Some common stat commands are:
Average
This command is used to calculate the average of the given field.
stats avg(field_name)
stats avg(product_price)
Max
It will return the maximum value from the specific field.
stats max(field_name)
stats max(user_age)
Min
It will return the minimum value from the specific field.
stats min(field_name)
stats min(product_price)
Sum
It will return the sum of the fields in a specific value.
stats sum(field_name)
stats sum(product_cost)
Count
The count command returns the number of data occurrences.
stats count(function) AS new_NAME
stats count(source_IP)
These are very important types of transforming commands that are used to present the data in table or visualization form. Most of the chart commands utilize various stat commands.
Command
chart
Explanation
The chart command is used to transform the data into tables or visualizations.
Syntax
| chart <function>
Example
| chart count by User
Search Query: index=windowslogs | chart count by User
Command
timechart
Explanation
The timechart command returns the time series chart covering the field following the function mentioned. Often combined with STATS commands.
Syntax
| timechart function <field_name>
Example
| timechart count by Image
The following query will display the Image chart based on the time.
Search Query: index=windowslogs | timechart count by Image
1) List the top 8 Image processes using the top command - what is the total count of the 6th Image?
Answer: 196
2) Using the rare command, identify the user with the least number of activities captured?
Answer: James
3) Create a pie-chart using the chart command - what is the count for the conhost.exe process?
Answer: 70
