hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Eelco Lempsink <lemps...@paragin.nl>
Subject Re: Running on multiple CPU's
Date Tue, 17 Apr 2007 22:27:10 GMT
On 17-apr-2007, at 20:09, Doug Cutting wrote:
> Misconfiguration would be my first guess, since lots of folks  
> (including the unit tests) run multiple daemons on a single host.   
> This message looks like the reduce tasks are starting, but no map  
> tasks are yet complete.  That could be because the map tasks are  
> somehow stuck, or for some other reason.  It's hard to say.

I can confirm the map tasks were stuck.  In fact, I found out the  
problem was caused by the firewall.  Narrowing it down, I came to the  
conclusion that the tasktracker somehow needs a hole in the firewall  
(port 50050).  This is strange, since in my configuration I bound it  
to 'lo' (mapred.tasktracker.dns.interface).  I don't think this is  
correct behaviour, so I'm filing a bug.


Eelco Lempsink

View raw message