如何解决Sitecore发布后CD服务器上Lucene索引重建异常问题？

Hey there, let's tackle this Sitecore 8.2 Lucene index issue you're dealing with—duplicate content after publish, and your auto-rebuild strategy not kicking in like it should. I’ve worked through similar problems with Sitecore indexing before, so here’s a step-by-step breakdown to get this sorted:

The root cause of your duplicate content is almost certainly that the index isn’t being refreshed automatically after publish, leaving old and new entries coexisting. Let’s dig into why the auto-rebuild is failing:

1. Double-Check Your Publish Event Handler Configuration #

First, confirm your publish:end event is actually targeting your custom blog index. Open your merged config (via ShowConfig.aspx) and look for this handler:

<event name="publish:end">
  <handler type="Sitecore.ContentSearch.Tasks.IndexingEventHandler, Sitecore.ContentSearch" method="OnPublishEnd">
    <indexes hint="list">
      <index>your_custom_blog_index</index>
    </indexes>
  </handler>
</event>

• If your custom index isn’t listed here, the event won’t trigger a rebuild for it. Add it via a config patch if missing.
• Watch out for other config patches that might be overriding this event—use ShowConfig.aspx to verify the final merged setup.

2. Validate Your Index Update Strategy #

Your index’s update strategy needs to be set to trigger on publish end. Check your index config:

<index id="your_custom_blog_index" type="Sitecore.ContentSearch.LuceneProvider.LuceneIndex, Sitecore.ContentSearch.LuceneProvider">
  <!-- ... other settings ... -->
  <strategies hint="list:AddStrategy">
    <strategy ref="contentSearch/indexConfigurations/indexUpdateStrategies/onPublishEndAsync" />
  </strategies>
</index>

• Avoid manual-only strategies here. If you’re using a custom strategy, ensure it’s explicitly tied to the publish:end event.
• Try switching to onPublishEndSync temporarily to rule out async thread pool issues (common in older Sitecore versions) that might be silently failing.

3. Dig Into Sitecore Logs for Clues #

Logs are your best diagnostic tool here. Filter entries around publish times using keywords like your_custom_blog_index, publish:end, or IndexingEventHandler to look for:

• Exceptions like Lucene.Net.Index.IndexNotFoundException or permission errors writing to the index folder.
• Messages indicating the rebuild was skipped, timed out, or never initiated.
• Any warnings about missing index configurations or invalid item filters.

4. Verify Index Folder Permissions #

Lucene needs write access to its storage folder (usually App_Data/Indexes/your_custom_blog_index). If the application pool user doesn’t have Modify permissions here:

• Auto-rebuilds will fail silently, leaving old index entries intact (causing duplicates).
• Manual rebuilds might work because you’re running the control panel with higher permissions, but the app pool can’t execute the auto-rebuild.

5. Check for Indexing Rule or Filter Misconfigurations #

Duplicates can also happen if your index is pulling in unintended items or versions:

• Ensure indexLatestVersionOnly is set to true in your index config (so only the latest published version is indexed).
• Verify includeTemplates/excludeTemplates are correctly targeting only your blog items—no stray templates should be included.
• Check if field-level configurations are causing multiple entries for the same item (e.g., indexing all versions instead of just published ones).

6. Temporary Fallback: Custom Post-Publish Rebuild #

While you fix the auto-rebuild, you can add a custom handler to force a sync rebuild post-publish:

public class BlogIndexRebuilder
{
    public void OnPublishEnd(object sender, EventArgs args)
    {
        var blogIndex = ContentSearchManager.GetIndex("your_custom_blog_index");
        blogIndex?.Rebuild();
    }
}

Register this in your config under the publish:end event as a temporary fix, then remove it once the auto-rebuild is working.

内容的提问来源于stack exchange，提问作者Jordy