Indexing exceptions after adding new node to the system
Search related exceptions observed in logs after adding a second node to an existing system.
The new node has been added in the System > Settings > Search landing page.
Search works on both nodes but an exception occurs in the logs of second node: "Failed to update the PegaRULES index - Cannot create directory:"
The reason is that this directory exists only in the first node and not in the second node.
User environment details
- Pega 7.1.8
- JBoss EAP 6.3.3 GA
- Linux version: 2.6.32-504.12.2.el6.x86_64
- Microsoft SQL Server 11.00.5058
- Microsoft JDBC Driver 4.0 for SQL Server 4.0.2206.100
[ PegaRULES:07.10] ( search.internal.RuleIndexer) ERROR - Failed to update ClipboardPages in the PegaRULES index - Cannot create directory: /var/lib/pega-app1/temp/PegaRULESIndex/DataIndex
[ PegaRULES:07.10] (ch.internal.SystemIndexerAgent) ERROR - [ Indexer ] The System Indexer encountered an error
com.pega.pegarules.pub.PRRuntimeException: Failed to update the PegaRULES index - Cannot create directory: /var/lib/pega-app1/temp/PegaRULESIndex/DataIndex
Caused by: java.io.IOException: Cannot create directory: /var/lib/pega-app1/temp/PegaRULESIndex/DataIndex
Steps to Reproduce
- Add a new node to the existing single-node system
- Add the node details to the Search Index Host Node Setting list in the System > Settings > Search landing page
- Set different paths for Search Index File Directories
- Check second node logs
A defect or configuration issue in the operating environment. Lucene indexing agents are running on the system.
The exceptions are caused by the SystemIndexer and SystemWorkIndexer agents that refer to the directory listed in DSS "indexing/explicitindexdir".
These agents can be stopped when search is in distributed mode (which is the default mode in Pega 7.1.8). The indexing is performed by FTSIncrementalIndexer agent, which is the only indexing agent to be kept running.
Stop the agents to resolve the issue.