[
https://issues.apache.org/jira/browse/NUTCH-2911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17453074#comment-17453074
]
Sebastian Nagel commented on NUTCH-2911:
----------------------------------------
Hi [~prakharchaube], thanks. I've also seen this and triggered a [new
build|https://ci-builds.apache.org/job/Nutch/job/Nutch-trunk/54/] which
succeeded. Actually, it was a protocol plugin test which failed:
{noformat}
[junit] Test org.apache.nutch.protocol.okhttp.TestBadServerResponses FAILED
{noformat}
This happens from time to time because these tests launch a server listening on
a fixed port and then send requests to that port. If the port is already
occupied, the test fails. We should at least fix the tests that all plugins use
different ports or not run multiple tests in parallel (see {{<parallel
threadCount="2">}} in src/plugin/build.xml).
> Add cleanup call in Fetcher.java
> --------------------------------
>
> Key: NUTCH-2911
> URL: https://issues.apache.org/jira/browse/NUTCH-2911
> Project: Nutch
> Issue Type: Improvement
> Components: fetcher
> Affects Versions: 1.18
> Reporter: Prakhar Chaube
> Priority: Minor
> Labels: improvement
> Fix For: 1.19
>
>
> Fetcher's inner class, FetcherRun overrides Hadoop Mapper's run().
> Even though Nutch's FetcherRun doesn't need an explicit call the Mapper's
> cleanup() (Which is a blank function), it would increase the readability and
> completeness of the run Method to do so.
> Ideally, every implementation of Mapper is supposed to do the following tasks:
> 1. Perform Setup
> 2. Call map on the data set
> 3. Perform cleanups.
> Moreover, in case a custom Fetcher is written extending Fetcher.java cleanup
> could get easily missed out.
> PR for Fix: [here|https://github.com/apache/nutch/pull/708]
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
