I've tried this before I send the posting. Nothing changes. I've set this als default in workspace.xml. No info in the logfile. I've tested to remove the Index (which offen helps) but in this case, the system crashes in the startup during the index build.
I'm checking why this all happen. Maybe there is something strange in the BundleManager and the myisam tables. I will activate innodb now for testing. I've seen in the log, that there is an Exception from the bundleManager. But the causing exception is not logged. So I'll make a test now with a patched BundleManager. (This solves not the reported problem, but maybe it will help in the future. Alexander Klimetschek wrote: > > On Tue, Jan 13, 2009 at 3:51 PM, Peter Heß <[email protected]> wrote: >> This will fail: >> >> Node node = getSession(false).getNodeByUUID(uuid); >> node.remove(); >> >> with an ItemNotExistException during the node.remove(), but the given >> UUID >> in the exception ist not the UUID of the node. >> >> Can anyone explain this? I think the repository is corrupted, but ist >> there >> an programitic way to solve this? > > There is a consistencyCheck option in the bundle persistence managers > that might solve your issue (and a consistencyFix option that will try > to fix those issues). The process requires to change the persistence > manager configuration (in the appropriate workspace.xml), ie. the > param consistencyCheck=true, restart the repository and inspect > Jackrabbit's log. The check (and the fix) will only run on startup. > > http://issues.apache.org/jira/browse/JCR-1428 > > http://jackrabbit.apache.org/api/1.5/org/apache/jackrabbit/core/persistence/bundle/BundleDbPersistenceManager.html > > Regards, > Alex > > -- > Alexander Klimetschek > [email protected] > > -- View this message in context: http://www.nabble.com/Problem-during-removing-a-node-with-childNodes-tp21413671p21442563.html Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
