KB482890: Fail to create cache, cube file or quick search index in clustered environment

• Strategy

This issue has been classified as a working by design behavior by MicroStrategy. The impact is limited to MicroStrategy Intelligence nodes and shared folder are not in same domain. A solution for this issue is available in current release.

SYMPTOM:

In Strategy 2020 and above, create cache, cube file or quick search index in the share folder fails if share folder and cluster nodes are not in the same domain.

Refer to Technical Document for configuring Cache, Cube, Inbox, and Session Recovery file sharing for a Strategy Intelligence Server cluster:
KB441125 How to configure Cache, Cube, Inbox, and Session Recovery file sharing for a MicroStrategy Intelligence Server cluster

Cache/Cube file cannot be created in share folder after configuring cache and cube share folders for Intelligence Server cluster.

The error log in DSSError.log is like:
Create \\ISERVERCLUSTER1\ClusterCube failed with error Failed to create directory '\\ISERVERCLUSTER1\ClusterCube'. (Server message 305419896) Access is denied. (Windows API error 5)

Also, File path in Developer is also a wrong path:

It should be a sub-directory with the share folder name, followed by another sub-folder whose name contains node name and project id.

When setting quick search index directory to a shared folder, the search index file created fails with below errors.


An IO exception from search engine: Command SECrawlToolCrawl; threw BaseException: A Java Exception: java.io.IOException: Cannot create directory: 
 <shared path>\index\38a062302d4411d28e71006008960167
at org.apache.lucene.store.NativeFSLock.obtain(NativeFSLockFactory.java:171)
at org.apache.lucene.store.Lock.obtain(Lock.java:72)
at org.apache.lucene.index.IndexWriter.<init>(IndexWriter.java:602)
at com.Strategy.search.CrawlTool.createEmptyIndexIfNotExists(CrawlTool.java:123)
at com.Strategy.search.CrawlTool.crawl(CrawlTool.java:77)

CAUSE:

The machine of the share folder and cluster nodes are not in the same domain.
Intelligence Server is started by system account if it's started as a service, so it uses system account to visit remote machine. On Windows, remote access for system account is granted only when the remote machine is in the same domain.

ACTION:

Join the machine of the share folder and all cluster nodes to the same domain to solve this issue.

Article Reference Number: KB482890

Comment

0 comments

Details

Knowledge Article

Published:

February 20, 2019

Last Updated:

June 6, 2022

An IO exception from search engine: Command SECrawlToolCrawl; threw BaseException: A Java Exception: java.io.IOException: Cannot create directory: <shared path>\index\38a062302d4411d28e71006008960167 at org.apache.lucene.store.NativeFSLock.obtain(NativeFSLockFactory.java:171) at org.apache.lucene.store.Lock.obtain(Lock.java:72) at org.apache.lucene.index.IndexWriter.<init>(IndexWriter.java:602) at com.Strategy.search.CrawlTool.createEmptyIndexIfNotExists(CrawlTool.java:123) at com.Strategy.search.CrawlTool.crawl(CrawlTool.java:77)