Alfresco Index rebuild going 100 % cpu and never ending

[emmanuelkasper]

Hello

After a user deleted some important datas, I had to reinstall Alfresco from backups. I choosed to reinstall, the software itself ( Alfresco 2.1), the postgres database, and the contentstore.

I configured index.recovery.mode=FULL in the repository.properties, and I see from the tomcat logs that the indexes are rebuilt.

maintenance@dakar:/srv/alfresco_data$ tail -f /opt/alfresco-2.1.0/tomcat/logs/catalina.out 
10:16:17,234 INFO  [node.index.FullIndexRecoveryComponent]    20 % complete.
10:17:12,781 INFO  [node.index.FullIndexRecoveryComponent]    30 % complete.
10:18:15,478 INFO  [node.index.FullIndexRecoveryComponent]    40 % complete.
10:19:50,181 INFO  [node.index.FullIndexRecoveryComponent]    50 % complete.
10:21:19,934 INFO  [node.index.FullIndexRecoveryComponent]    60 % complete.
10:22:50,441 INFO  [node.index.FullIndexRecoveryComponent]    70 % complete.
10:23:57,186 INFO  [node.index.FullIndexRecoveryComponent]    80 % complete.
10:25:07,464 INFO  [node.index.FullIndexRecoveryComponent]    90 % complete.
10:26:30,864 INFO  [node.index.FullIndexRecoveryComponent]    100 % complete.
10:26:30,871 INFO  [node.index.FullIndexRecoveryComponent] Index recovery completed.
‍‍‍‍‍‍‍‍‍‍‍‍

however after that CPU stays at 100 % , and it 's not possible to shutdown tomcat with bin/shutdown.sh
( I get

maintenance@dakar:/srv/alfresco_data$ tail -f /opt/alfresco-2.1.0/tomcat/logs/catalina.out 
10:16:17,234 INFO  [node.index.FullIndexRecoveryComponent]    20 % complete.
10:17:12,781 INFO  [node.index.FullIndexRecoveryComponent]    30 % complete.
10:18:15,478 INFO  [node.index.FullIndexRecoveryComponent]    40 % complete.
10:19:50,181 INFO  [node.index.FullIndexRecoveryComponent]    50 % complete.
10:21:19,934 INFO  [node.index.FullIndexRecoveryComponent]    60 % complete.
10:22:50,441 INFO  [node.index.FullIndexRecoveryComponent]    70 % complete.
10:23:57,186 INFO  [node.index.FullIndexRecoveryComponent]    80 % complete.
10:25:07,464 INFO  [node.index.FullIndexRecoveryComponent]    90 % complete.
10:26:30,864 INFO  [node.index.FullIndexRecoveryComponent]    100 % complete.
10:26:30,871 INFO  [node.index.FullIndexRecoveryComponent] Index recovery completed.
maintenance@dakar:/srv/alfresco_data$ less /opt/alfresco-2.1.0/tomcat/logs/catalina.out
maintenance@dakar:/srv/alfresco_data$ /opt/alfresco-2.1.0/tomcat/bin/shutdown.sh
 
Using CATALINA_BASE:   /opt/alfresco-2.1.0/tomcat
Using CATALINA_HOME:   /opt/alfresco-2.1.0/tomcat
Using CATALINA_TMPDIR: /opt/alfresco-2.1.0/tomcat/temp
Using JRE_HOME:       /usr/lib/jvm/java-6-sun-1.6.0.24/
01.03.2011 11:52:07 org.apache.catalina.startup.Catalina stopServer
SCHWERWIEGEND: Catalina.stop: 
java.net.ConnectException: Connection refused
   at java.net.PlainSocketImpl.socketConnect(Native Method)
   at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:35‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

)
I have to kill the java process. manually.

Is something going wrong in the index recovery ?

[mrogers]

From what you have posted the index recovery is complete.

[_sax]

Your log says that the reindexing is fast. Is that consistent with the size of your documents store (amount of non-binary-files)?
Does your Open Office-connection work (is the reindexing getting you content searchability)?

[emmanuelkasper]

Your log says that the reindexing is fast. Is that consistent with the size of your documents store (amount of non-binary-files)?
Does your Open Office-connection work (is the reindexing getting you content searchability)?

Well the content searchability never worked in my setup, so I am not going for it. Well after restarting alfresco with index=validate, it found my precious documents properly. I am right now getting them back with the WebDav interface.

[andy]

Hi

The index rebuild does not inlcude the time to reindex and transform all the content held in alfresco - all non ctent properties will have been reindexed.
The content indexing will continue in the background.

Andy

[_sax]

Thanks, I didn't know this. So, a high CPU usage after "finishing" could be a result of this content reindexing?