Installation problems moving from labs3b to labs3stable

[efestione]

Hi everyone,
I'm triyng to migrate from a server (A) to another (B), but I obtain this error:

20:03:59,969 INFO  [org.alfresco.config.xml.XMLConfigService$PropertyConfigurer] Loading properties file from class path resource [alfresco/file-servers.properties]
20:04:08,016 INFO  [org.alfresco.repo.domain.schema.SchemaBootstrap] Schema managed by database dialect org.hibernate.dialect.MySQLInnoDBDialect.
20:04:08,906 INFO  [org.alfresco.repo.domain.schema.SchemaBootstrap] No changes were made to the schema.
20:04:09,906 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent] Index recovery started: 949 transactions.
20:04:10,172 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        10 % complete.
20:04:10,172 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        20 % complete.
20:04:10,172 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        30 % complete.
20:04:10,188 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        40 % complete.
20:04:10,188 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        50 % complete.
20:04:10,188 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        60 % complete.
20:04:10,203 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        70 % complete.
20:04:10,203 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        80 % complete.
20:04:10,203 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        90 % complete.
20:04:10,219 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent]        100 % complete.
20:04:18,469 INFO  [org.alfresco.repo.node.index.FullIndexRecoveryComponent] Index recovery completed.
20:04:18,500 INFO  [org.alfresco.repo.node.index.AVMFullIndexRecoveryComponent] Rebuilding indexes for 1 AVM Stores
20:04:18,500 INFO  [org.alfresco.repo.node.index.AVMFullIndexRecoveryComponent]     Rebuilding index for sitestore
20:04:18,563 INFO  [org.alfresco.repo.node.index.AVMFullIndexRecoveryComponent]   Reindex   100% complete
20:04:18,563 INFO  [org.alfresco.repo.node.index.AVMFullIndexRecoveryComponent] Finished rebuilding indexes for AVM Stores
20:04:18,750 INFO  [org.alfresco.repo.admin.ConfigurationChecker] The Alfresco root data directory ('dir.root') is: E:\Alfresco\alf_data
20:04:18,766 ERROR [org.alfresco.repo.admin.ConfigurationChecker] CONTENT INTEGRITY ERROR: System content not found in content store.
20:04:18,766 ERROR [org.alfresco.repo.admin.ConfigurationChecker] Ensure that the 'dir.root' property is pointing to the correct data location.
20:04:18,766 ERROR [org.springframework.web.context.ContextLoader] Context initialization failed
org.alfresco.error.AlfrescoRuntimeException: Ensure that the 'dir.root' property is pointing to the correct data location.
                at org.alfresco.repo.admin.ConfigurationChecker.check(ConfigurationChecker.java:312)
…

Servers are:

(A)
Windows XP
dir.root: C:/Alfresco/alf_data

(B)
Windows Server 2003
dir.root: E:/Alfresco/alf_data

I followed the steps described in the wiki section "General_Process_Upgrade".
Can the different drive ('E' instead of 'C') be the source of this problem? Or what else?

Thanks for your help.
Alessandro

[derek]

Hi,
My guess is that it occured during the upgrade process, assuming it didn't occur on 3b.  As you know we attempt to provide an incremental upgrade process along all versions for the Community releases, but don't spend much QA time on it - the focus, for Community, rather being on upgrades between major releases.
Regards

[efestione]

Hi,
I verified these dirty records were also on 3b.
Thanks again
Regards

[derek]

Was 3b the first version you installed or did you upgrade to it from an earlier release?
Anyway, as long as the issue doesn't arise again, at least we know that the issue isn't in the code as it stands.

[efestione]

Yes, it was the first version I installed.

[luca]

Hi,
I have quite the same problem, except I'm using Alfresco 3.0 Labs from the beginning.

I started to have the same errors in ftsIndexerJobDetail job:

09:53:24,686  ERROR [quartz.core.ErrorLogger] Job (DEFAULT.ftsIndexerJobDetail threw an exception.
org.quartz.SchedulerException: Job threw an unhandled exception. [See nested exception: java.lang.NullPointerException]
        at org.quartz.core.JobRunShell.run(JobRunShell.java:213)
        at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:529)
Caused by: java.lang.NullPointerException
        at java.io.StringReader.<init>(StringReader.java:33)
        at org.alfresco.repo.search.impl.lucene.ADMLuceneIndexerImpl.indexProperty(ADMLuceneIndexerImpl.java:1037)
        at org.alfresco.repo.search.impl.lucene.ADMLuceneIndexerImpl.createDocuments(ADMLuceneIndexerImpl.java:602)
        at org.alfresco.repo.search.impl.lucene.ADMLuceneIndexerImpl.updateFullTextSearch(ADMLuceneIndexerImpl.java:1502)
        at org.alfresco.repo.search.impl.lucene.fts.FullTextSearchIndexerImpl.index(FullTextSearchIndexerImpl.java:188)
…
‍‍‍‍‍‍‍‍‍‍‍‍‍

I tryed to do a full index rebuild, but the problem isn't solved.

If I search in the database with this query:

select * from alf_node_properties np join alf_locale l on (l.id = np.locale_id) where l.locale_str != '.default' and np.string_value is null;‍

I get more than 40000 records like these:

node_id     actual_type_n     persisted_type_n     boolean_value     long_value     float_value     double_value     string_value     serializable_value     qname_id     list_index     locale_id     id     version     locale_str
857191    0    0    f    0    0    0          106    -1    69730    69730    0    it_IT_
857227    0    0    f    0    0    0          106    -1    69730    69730    0    it_IT_
857071    0    0    f    0    0    0          106    -1    69730    69730    0    it_IT_
‍‍‍‍‍‍

Do I have to delete all these records? Won't it cause problems on the documents related to them?

If I search to get cm:title and cm:description properties without any value I also get more than 40000 records, am I doing anything wrong?

Can I delete them without any problem?

[derek]

Hi, Luca

In 3.0, the MLText (multilingual text) values were stored in the alf_attribute tables.  Upon upgrade, these values were pulled back into alf_node_properties.  It appears that null values in the originally-persisted MLText object - which was just a map - had been persisted and are causing your problems.

Your analysis of the alf_node_properties table is quite correct.   Notice the 'actual_type_n' values that are zero … that indicates a null value.  In other words, you can delete all alf_node_properties where the locale is not '.default' (ID 69730?) and the persisted_type_n is zero.  Perhaps just double-check that all string_value values are null where persisted_type_n is 0.

Regards