Reporting

What about archiving/offline storage?

Explorer

What is the recommended way to export/archive a large amount of historical data for retention or offline storage? I would like to clean old data from splunk but still maintain our regulatory requirements on retention.

Tags (2)
0 Karma
1 Solution

Splunk Employee
Splunk Employee

The coldToFrozenScript (also specified in indexes.conf) runs just before the frozen data is erased. The default script simply writes the name of the directory being erased to the log file $SPLUNK_HOME/var/log/splunk/splunkd_stdout.log. If you want to archive frozen data rather than delete it, you'll need to substitute your own archiving script.

To substitute your own script, add the following stanza to $SPLUNK_HOME/etc/system/local/indexes.conf:

[] coldToFrozenScript =

Note the following:

* <index> specifies which index to archive.
* <script> specifies the archiving script.
      o Define the <$script> path relative to $SPLUNK_HOME/bin. The script needs to be located in that directory or a subdirectory. 

Splunk ships with two archiving scripts that you can modify (or you can create your own):

* compressedExport.sh: Export with tsidx files compressed as gz.
* flatfileExport.sh: Export as a flat text file (not recommended for current performance and resource issues -- it can take a long time, and use a lot of ram, 2-3GB, while running). 

View solution in original post

Splunk Employee
Splunk Employee

The coldToFrozenScript (also specified in indexes.conf) runs just before the frozen data is erased. The default script simply writes the name of the directory being erased to the log file $SPLUNK_HOME/var/log/splunk/splunkd_stdout.log. If you want to archive frozen data rather than delete it, you'll need to substitute your own archiving script.

To substitute your own script, add the following stanza to $SPLUNK_HOME/etc/system/local/indexes.conf:

[] coldToFrozenScript =

Note the following:

* <index> specifies which index to archive.
* <script> specifies the archiving script.
      o Define the <$script> path relative to $SPLUNK_HOME/bin. The script needs to be located in that directory or a subdirectory. 

Splunk ships with two archiving scripts that you can modify (or you can create your own):

* compressedExport.sh: Export with tsidx files compressed as gz.
* flatfileExport.sh: Export as a flat text file (not recommended for current performance and resource issues -- it can take a long time, and use a lot of ram, 2-3GB, while running). 

View solution in original post