Hello,
 
   I am trying to do a breadth first crawl using nutch. Currently, the
generator lets me specify the max # of URLs to be generated for
fetching. If the # of unfetched URLs in the current crawldb is too
large, I would like to create multiple fetchlists using one generate
task, by specifying a batchSize parameter. For e.g., topN=10M and
batchSize=1M will create at most 10 segments each 1M URLs in its
fetchlist (the last segment could have less).
 
  I went through the discussion here:
http://issues.apache.org/jira/browse/NUTCH-171. I have to agree with Rod
and Stefan here that a larger job is much more likely to fail than
several smaller jobs. That is why I was wondering if I can modify the
generator to create multiple output segments somehow. Any idea as to how
this can be accomplished?
 
Thank you,
 
Vishal.
 
 

Reply via email to