[
https://issues.apache.org/jira/browse/NUTCH-2060?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14625936#comment-14625936
]
Sebastian Nagel commented on NUTCH-2060:
----------------------------------------
The dedup has been run with db.update.purge.404 == true. This could be the
reason since DeduplicationJob shares the mapper CrawlDbFilter with the update
job (CrawlDb).
> dedup is removing entries with status db_gone
> ---------------------------------------------
>
> Key: NUTCH-2060
> URL: https://issues.apache.org/jira/browse/NUTCH-2060
> Project: Nutch
> Issue Type: Bug
> Components: crawldb
> Affects Versions: 1.9
> Reporter: Steven Hayles
> Priority: Minor
>
> Using the standard bin/crawl script, Solr is never informed when a previously
> indexed document has been deleted.
> "bin/nutch update" sets db_gone status in the crawl db for requests returning
> HTTP 404 status.
> "bin/nutch dedup" remove entries with status db_gone from the crawl db .
> As a result "bin/nutch clean" never sees the db_gone status, so does not
> inform Solr.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
