如何停止AEM中运行的/oak:index/cqTagLucene等OOTB索引并重新索引

If you've got an out-of-the-box index like /oak:index/cqTagLucene chugging along for 4 days (thanks to those millions of PDF assets), here are the reliable ways to halt it cleanly before updating your exclude paths:

Method 1: Use the JMX Console (Most Straightforward) #

This is the go-to method for stopping in-progress index builds:

• Navigate to your AEM instance's JMX console at http://<your-aem-host>:<port>/system/console/jmx
• Search for the org.apache.jackrabbit.oak:name=IndexStats,type=QueryEngine MBean
• Find the operation named stopIndexing(String indexPath)
• Enter the full path of your index (e.g., /oak:index/cqTagLucene) in the parameter field and click Invoke
• Verify the index status by checking the getIndexStatus(String indexPath) operation—it should return IDLE once stopped

Method 2: Directly Modify the Index Node via CRXDE Lite #

If JMX isn't accessible or you prefer working with repository nodes:

• Open CRXDE Lite at http://<your-aem-host>:<port>/crx/de
• Navigate to the index node: /oak:index/cqTagLucene
• Locate the reindex property (it's set to true since the index is running)
• Change reindex to false and save the node
• Optional but recommended: Check the /oak:index/cqTagLucene:async child node—if it exists, delete it to clear any pending indexing tasks

Method 3: Run a Groovy Script (For Automation/Bulk Operations) #

If you're comfortable with scripting, this works well for repeat tasks:

• Open the Groovy Console at http://<your-aem-host>:<port>/groovyconsole
• Paste the following script, replacing the index path with yours:

import org.apache.jackrabbit.oak.api.QueryEngine
import org.apache.jackrabbit.oak.spi.query.QueryIndex

def indexPath = "/oak:index/cqTagLucene"
def queryEngine = osgi.getService(QueryEngine.class)

queryEngine.stopIndexing(indexPath)
println "Index ${indexPath} has been stopped. Current status: ${queryEngine.getIndexStatus(indexPath)}"

• Click Run Script to execute it

Post-Stop Tips Before Re-Running the Index #

• Double-check that the index status is IDLE using JMX or CRXDE (look for the indexStatus property on the index node)
• Add your excludePaths property to the index node (e.g., exclude large PDF asset folders that don't need tagging indexing)
• When you're ready to re-trigger, set reindex back to true—this will start the index build with your new exclude paths, which should be much faster

内容的提问来源于stack exchange，提问作者Mukesh Jha