You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-user@hadoop.apache.org by Forrest Aldrich <fo...@gmail.com> on 2014/01/23 01:13:07 UTC

Hadoop status pages

We are running Hadoop 1.x (due to API requirements).   I've been asked a 
question by one of our developers, relevant to clicking through the job 
information in various pages hadoop presents.

http://our-machine.com:50030/machines.jsp?type=active

Some of the URLs point to the private, backend network that we use for 
Hadoop traffic.  Others point to the front-end IP.   Because our office 
is not connected to the non-routed backend network, we can't click 
through some of the URLs to see reports and thus have to parse logfiles.

Is there a way around this?   We want the Hadoop slaves to listen only 
on the backend network -- but I wonder if there's a configuration 
directive I've missed that would solve the above problem.

And finally, as for the HTML presented in these various links, is it 
possible to insert custom layout (ie: CSS) to make the presentation 
better -- or is this dynamically generated from within the Hadoop code.


Thank you.