Steve, The patch applies to the source files. The "ant clean jar" command then recompiles the system from its source. You'll need to restart the daemons after recompiling for this to take effect.
That having been said, HADOOP-1700 is a notoriously huge patch that touches a very large spread of the HDFS code. It will assuredly fail to apply cleanly to the Hadoop 0.17.0 source code; you'll need to merge any collisions in by hand, and there will be an awful lot of them. You'll probably have better luck by just upgrading to Hadoop 0.19.0, where HADOOP-1700 was enabled. (It has since been disabled in 0.19.1, because it can cause data loss; so you shouldn't use this on a production cluster.) Good luck with whatever route you go with. Cheers, - Aaron On Tue, Mar 17, 2009 at 4:56 PM, Steve Gao <steve....@yahoo.com> wrote: > Thank you, Ravi and Amandeep. > Can I apply a patch on the fly? Or I need to shut down HTFS & job-tracker > beforehand? > > > --- On Tue, 3/17/09, Ravi Phulari <rphul...@yahoo-inc.com> wrote: > From: Ravi Phulari <rphul...@yahoo-inc.com> > Subject: Re: How to apply a patch to my hadoop? > To: "core-user@hadoop.apache.org" <core-user@hadoop.apache.org>, " > steve....@yahoo.com" <steve....@yahoo.com> > Date: Tuesday, March 17, 2009, 7:52 PM > > Hello Steve. > > Assuming you are using *nix. > > To Apply patch > patch -p0 -E < HADOOP-X.patch > > To remove Patch > patch -p0 --reverse -E < HADOOP-X.patch > > > Hope this helps. > > Regards, > Ravi > > > > On 3/17/09 4:48 PM, "Steve Gao" <steve....@yahoo.com> wrote: > > I want to apply this patch > https://issues.apache.org/jira/browse/HADOOP-1700 > to my hadoop 0.17.0 . > > Would anybody tell me how to do it? Thanks! > > > > > > > > > > >
