Copied to clipboard

Flag this post as spam?

This post will be reported to the moderators as potential spam to be looked at


  • Glen McMillan 2 posts 22 karma points
    Apr 06, 2010 @ 06:06
    Glen McMillan
    0

    Lucene unpublishing content?

    I have an Umbraco 4.0 site on Windows Server 2003 R2 with SQL 2005.

    A few days ago a page disappeared from both the website and the CMS as though it's been deleted, but so far as I can tell it's not through any user/admin action. The facts surrounding the disappearing page are:

    • The page does not appear in the CMS tree
    • The corresponding node in the database exists, but has the published flag set to 0
    • The 'updateDate' is set correctly to several weeks prior to the day it disappeared
    • There's no 'releaseDate' or 'expireDate' set
    • I know when it vanished because it contained a form generating dozens of datestamped posts per day which suddenly stopped
    • There are no unpublish, delete or move events in the logs for that node (nor even any logins around the time it disappeared)
    • There was an error to do with lucene not being able to update it's index around the time it disappeared (appended below)
    • Have checked the permissions on the index folder - they're fine
    • I am currently reindexing - there are 17,351 files to index and the index itself is 15Mb and growing

    I was able to resurrect the page quickly from a staging site, so that's no problem. My issue is prevention - understanding what caused this and preventing it from recurring. Why do I think this is Lucene? Largely due to similar issues posted here! Any insights appreciated.

    Error removing node from umbraco index: 'System.IO.FileNotFoundException: Could not find file 'E:\webroot\mysite.com\data\_systemUmbracoIndexDontDelete\_13g.fnm'. 
    File name: 'E:\webroot\mysite.com\data\_systemUmbracoIndexDontDelete\_13g.fnm'    
    at System.IO.__Error.WinIOError(Int32 errorCode, String maybeFullPath)    
    at System.IO.FileStream.Init(String path, FileMode mode, FileAccess access, Int32 rights, Boolean useRights, FileShare share, Int32 bufferSize, FileOptions options, SECURITY_ATTRIBUTES secAttrs, String msgPath, Boolean bFromProxy)   
    at System.IO.FileStream..ctor(String path, FileMode mode, FileAccess access, FileShare share)    
    at Lucene.Net.Store.FSIndexInput.Descriptor..ctor(FSIndexInput enclosingInstance, FileInfo file, FileAccess mode)    
    at Lucene.Net.Store.FSIndexInput..ctor(FileInfo path)   
    at Lucene.Net.Store.FSDirectory.OpenInput(String name)   
    at Lucene.Net.Index.FieldInfos..ctor(Directory d, String name)    
    at Lucene.Net.Index.SegmentReader.Initialize(SegmentInfo si)    
    at Lucene.Net.Index.SegmentReader.Get(Directory dir, SegmentInfo si, SegmentInfos sis, Boolean closeDir, Boolean ownDir)    
    at Lucene.Net.Index.IndexReader.AnonymousClassWith.DoBody()   
    at Lucene.Net.Store.Lock.With.Run()   
    at Lucene.Net.Index.IndexReader.Open(Directory directory, Boolean closeDirectory)   
    at Lucene.Net.Index.IndexReader.Open(String path)  
    at umbraco.cms.businesslogic.index.Indexer.RemoveNode(Int32 Id)'

  • Sebastiaan Janssen 5060 posts 15522 karma points MVP admin hq
    Apr 06, 2010 @ 11:29
    Sebastiaan Janssen
    0

    I doubt that Lucene has caused this, because it does not touch the content at all. If you look in your umbracoLog table you also see the nodeId for this error and you'll notice that it was a node that was being saved or published at that time.

    That being said, the Lucene version bundled with Umbraco isn't very good and can cause loads of these indexing errors. For me what has helped with this was installing SP2 on the Windows 2008 server hosting the site. This seems to have helped for some others as well.

    Please note that there are various debug flags in various config files that can be enabled and disabled, so you might not see all of the save, publish, unpublish, delete, etc. actions in your log file currently.

  • Glen McMillan 2 posts 22 karma points
    Apr 06, 2010 @ 23:07
    Glen McMillan
    0

    Thanks for the help Sebastiaan.

    I have have checked that debug events (publish, unpublish, delete, etc) are being logged and have checked again that there aren't any for the missing node.

    So accepting that the Lucene indexing errors are unrelated, are there any other ideas out there as to what could have caused this?

Please Sign in or register to post replies

Write your reply to:

Draft