[regarding mapred ver 0.8]
Anton Potehin wrote:
I tried to launch mapred on 2 machines: 192.168.0.250 and 192.168.0.111.
051123 053136 task_m_xaynqo -14885.741% /user/root/seeds/urls:31+31
Please help me to find out what the problem is? And what I did wrong?
Is the problem the negative progress percentages (perhaps it also does not
finish)?
I am going to write up a new wiki page about usage scenarios and it looks like
you are using an unusual setup by having different configuration settings on
each node. Use the same settings (that's why it is called nutch-site.xml) on
both (all) nodes.
The bin/start_all.sh provides a perfect example of how to start up the
ensemble. The sh scripts are extemely elegant, but be careful to notice the
difference between nutch-daemon.sh and nutch-daemons.sh; in your case,
192.168.0.250 is the master node, so you need to put the following into
$HOME/.slaves
192.168.0.111
192.168.0.250
(notice your master is also a slave; this should work, but the normal case is
to have only namenode and jobtracker running on the master (and, of course,
namenode and jobtracker are not run on any other nodes) and you need to allow
ssh work from the master to the slave without a password (either that or
customize bin/slaves.sh for rsh, etc. or unwind it and it by hand)
Paul
[EMAIL PROTECTED]
