[jira] [Commented] (NUTCH-2420) Bug in variable generate.max.count and fetcher.server.delay

["Sebastian Nagel (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NUTCH-2420?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16240416#comment-16240416 ] 

Sebastian Nagel commented on NUTCH-2420:
----------------------------------------

That's really ugly! It will work in local mode or on a small cluster with not too many hosts, but on a larger cluster a HostDb with many partitions and millions of hosts, it will slow down of Generator significantly. The correct solution would be to use <host,score> pairs as keys in the Selector job, with a partitioner and secondary sorting so that all keys with same host end up in the same call of the reducer. If values can also hold a HostDb entry and the sort comparator guarantees that the HostDb entry (entries if partitioned by domain or IP) comes in front of all CrawlDb entries. But that would be a substantial improvement...

+1 to commit this fix and to open a new one to speed up the merging of HostDb entries

> Bug in variable generate.max.count and fetcher.server.delay
> -----------------------------------------------------------
>
>                 Key: NUTCH-2420
>                 URL: https://issues.apache.org/jira/browse/NUTCH-2420
>             Project: Nutch
>          Issue Type: Bug
>          Components: generator
>            Reporter: Markus Jelsma
>            Assignee: Markus Jelsma
>             Fix For: 1.14
>
>         Attachments: NUTCH-2420.patch
>
>
> Feature added by NUTCH-2368 does not work for multiple hosts. Once a HostDatum has been read by getHostDatum(), the next host cannot be read. Apparantly i need to open and close the SequenceFile.Readers for every HostDatum it needs. Reader has no reset() method or whatsoever.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)