Cranking disk space optimization with Jackrabbit Oak in AEM 6.1

Oak is a scalable and hierarchical content repository and has been created by Apache Jackrabbit project.

This enterprise solution is the foundation of major enterprise platforms including Adobe Experience Manager (formerly knows as CQ by Day Software). The current version of AEM 6.1 consist of modules associated with release 1.2.2 of Apache Oak content repository. You can find more details on this initiative here.

Adobe has released compaction methodologies (offline and online) for enterprises experiencing unusual growth of their content repository. This can be caused by a variety of issues including the creation of too many nodes or size of the renditions etc. The online compaction mechanism allows an operations team to run scheduled maintenance tasks directly through an AEM maintenance dashboard. The details of the online compaction can be found here.

I have outlined the steps required to perform an offline compaction. It is a pre-requisite for performing online compaction if you haven’t removed the old checkpoints already.

Example Use Case:

The size of an Adobe installation increased many-folds during the use of a system by a client and the disk space on the operating system grew to approximately 160 GB. After cleaning up the JBoss temp directory and log files, the crx-repository still took approximately 85 GB of disk space before running the manual optimization steps. The following steps can be used if you are encountering similar issues

1. Shutdown your AEM instance.
2. Check the Oak Core version from the bundle list in the console: http://localhost:8080/lc/system/console/status-Bundlelist
   1. In my case it was found to be “org.apache.jackrabbit.oak-core (1.2.2)”.
3. Download the matching version of Oak Run tool from the maven repository. In my case, I used: http://mvnrepository.com/artifact/org.apache.jackrabbit/oak-run/1.2.2
4. Run the following commands to optimize the Oak repository:
   • (Find the old checkpoints using the Oak runnable jar file) java -jar oak-run-1.2.2.jar checkpoints C:\Adobe\Adobe_Experience_Manager_forms\crx-repository\repository\segmentstore

   

   • (Delete un-referenced checkpoints) java -jar oak-run-1.2.2.jar checkpoints C:\Adobe\Adobe_Experience_Manager_forms\crx-repository\repository\segmentstore rm-unreferenced

   

   • (Run the compaction command, sit back and relax) java -jar oak-run-1.2.2.jar compact C:\Adobe\Adobe_Experience_Manager_forms\crx-repository\repository\segmentstore

   

The Adobe folder size was reduced to 21 GB after the completion of the optimization tasks. Please start your AEM server after these tasks are complete.