Thank you for all the info. I am able to rerun the failed job (CDH3u3)
-DR -----Original Message----- From: Mayank Bansal [mailto:[email protected]] Sent: Tuesday, March 13, 2012 10:26 PM To: [email protected] Subject: Re: rerun of failed workflow jobs I think you are running cdh3u2 which is Oozie 2.3 and does not contain this property oozie.wf.rerun.failnodes You just need to give "oozie.wf.rerun.skip.nodes=," and it should work Thanks, Mayank On Tue, Mar 13, 2012 at 10:11 PM, Michelle Chiang <[email protected]>wrote: > > Hi DR, > > (1) To rerun workflow job from failed nodes, add this property to > job.properties: > oozie.wf.rerun.failnodes=true > > then, rerun workflow as: > oozie job -rerun $jid -config job.properties > > (2) or rerun workflow job from since there is 1 node, add this > property to > job.properties: > oozie.wf.rerun.skip.nodes=, > > then > oozie job -rerun $jid -config job.properties > > > > please note, for workflow rerun, job properties file cannot have both > "oozie.wf.rerun.failnodes=true" and "oozie.wf.rerun.skip.nodes=," at > the same time. > > > regards > michelle > > -----Original Message----- > From: Deepak Reddy [mailto:[email protected]] > Sent: Tuesday, March 13, 2012 8:04 PM > To: [email protected] > Subject: RE: rerun of failed workflow jobs > > Yes I added those but now I am stuck at this variable > oozie.wf.rerun.failnodes > > I only have one action node in my XML so what should I enter for > oozie.wf.rerun.failnodes > > Thanks, > DR > > -----Original Message----- > From: Mayank Bansal [mailto:[email protected]] > Sent: Tuesday, March 13, 2012 7:58 PM > To: [email protected] > Subject: Re: rerun of failed workflow jobs > > Did you try with the changes which i told you earlier, you have to > provide those properties in your property file while submitting the job. > > Thanks, > Mayank > > On Tue, Mar 13, 2012 at 7:25 PM, Deepak Reddy <[email protected]> > wrote: > > > I have only one action node in my workflow which I want to rerun. > > > > What would be the value for oozie.wf.rerun.failnodes in that case. > > > > Thanks, > > DR > > > > -----Original Message----- > > From: Mayank Bansal [mailto:[email protected]] > > Sent: Tuesday, March 13, 2012 5:03 PM > > To: [email protected] > > Subject: Re: rerun of failed workflow jobs > > > > Are you using secure hadoop or non secure hadoop? > > try adding following properties to your property file > > > > mapreduce.jobtracker.kerberos.principal=mapred/_HOST@LOCALREALM > > > > dfs.namenode.kerberos.principal=hdfs/_HOST@LOCALREALM > > > > Thanks, > > Mayank > > > > On Tue, Mar 13, 2012 at 4:51 PM, Deepak Reddy > > <[email protected]> > > wrote: > > > > > I tried with oozie.wf.rerun.failnodes=false > > > > > > It didn't work. > > > > > > -DR > > > > > > -----Original Message----- > > > From: Mona Chitnis [mailto:[email protected]] > > > Sent: Tuesday, March 13, 2012 4:47 PM > > > To: [email protected] > > > Subject: Re: rerun of failed workflow jobs > > > > > > Hello, > > > > > > Have you specified either of these properties correctly in your > > > properties file? > > > > > > oozie.wf.rerun.skip.nodes (comma-separated list of action nodes. > > > Should be completed in previous run) oozie.wf.rerun.failnodes ( > > > can be {true | false} > > > ) > > > > > > > > > --Mona > > > > > > > > > On 3/13/12 4:40 PM, "Deepak Reddy" <[email protected]> wrote: > > > > > > Hi > > > > > > When I try to rerun a failed oozie workflow job I am getting the > > > following error > > > > > > oozie job -config /tmp/loadlogs.properties -rerun > > > 0000048-120312221512410-oozie-oozi-W > > > > > > its giving me Error: HTTP error code: 500 : Internal Server Error > > > > > > I am using oozie 2.3.0 from CDH3u3 distribution. > > > > > > Thanks, > > > DR > > > > > > > > > -- Thanks and Regards, Mayank Cell: 408-718-9370
