Sebastian Nagel created NUTCH-2041:
--------------------------------------
Summary: indexer fails if linkdb is missing
Key: NUTCH-2041
URL: https://issues.apache.org/jira/browse/NUTCH-2041
Project: Nutch
Issue Type: Bug
Components: indexer, linkdb
Affects Versions: 1.10
Reporter: Sebastian Nagel
Fix For: 1.11
If the linkdb is missing the indexer fails with
{noformat}
2015-06-17 12:52:10,621 ERROR
...cause:org.apache.hadoop.mapred.InvalidInputException: Input path does not
exist: .../linkdb/current
{noformat}
If both db.ignore.internal.links and db.ignore.external.links there will be no
LinkDb even if "invertlinks" is run (as consequence of NUTCH-1913). The script
"bin/crawl" does not know about the values of these two properties and calls
indexer with "-linkdb .../linkdb" which will then fail.
Since "bin/crawl" is agnostic to properties defined in nutch-site.xml we
solution similar to NUTCH-1854: make the tool/job more tolerant and log a
warning instead of raising an error.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
