You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@oozie.apache.org by dejace sr <de...@gmail.com> on 2012/01/27 02:45:49 UTC
oozie rerun job command in 2.3.0
Hi,
we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
move to 3.x.
One of the important oozie command is re-running of the job from the point
of its failure, but I get error code 500.
oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
Error: HTTP error code: 500 : Internal Server Error
I did check the below, but wanted to know if it has been fixed.
https://issues.apache.org/jira/browse/OOZIE-256
Thanks,
DJ
Re: oozie rerun job command in 2.3.0
Posted by Mohammad Islam <mi...@yahoo.com>.
What are the contents of your config file?
I want to know only the key names.
I'm looking for a workaround where you could send the appropriate config.
Regards,
Mohammad
----- Original Message -----
From: dejace sr <de...@gmail.com>
To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>
Cc: "mayank@apache.org" <ma...@apache.org>
Sent: Friday, January 27, 2012 11:19 AM
Subject: Re: oozie rerun job command in 2.3.0
Hi Mohammad,
I think it was a typo from my end while drafting the mail.
I had provided the job config in the command and I get error code 500.
So do we have any fix for this in 2.3.0? Bcaz, I cant switch
to 3.x until the project switches to 3.x?
(Parallel, I am also filing a bug against cloudera which is supporting our
hadoop cluster)
Thanks for the reply.
Regards,
DJ
On Thu, Jan 26, 2012 at 11:37 PM, Mohammad Islam <mi...@yahoo.com> wrote:
>
>
> Hi DJ,
> In addition, 3.x has feature where you don't need to specify the
> skipnodes, if you want to rerun from the failed node.
>
>
>
> Most importantly:
> Just came to my notice . why are you not passing -config in your rerun
> command.
> Please see the link
>
>
> http://yahoo.github.com/oozie/releases/2.3.0/DG_CommandLineTool.html#Rerunning_a_Workflow_Job
>
>
> It is very important.
>
>
> Regards,
> Mohammad
>
>
>
> ________________________________
> From: dejace sr <de...@gmail.com>
> To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>;
> mayank@apache.org
> Sent: Thursday, January 26, 2012 6:27 PM
> Subject: Re: oozie rerun job command in 2.3.0
>
> Thanks Mohammad.
>
> Our current objective is to go to production with oozie 2.3.0
> and after some months we will be switching to 3.x
>
> so in 2.3.0, do we have any patch/fix or workaround for this?
> because, we don't want to rerun the passed tasks of the job, but
> instead, would like to rerun from the point of failed task in the job;
> this way, it saves time and resources.
>
> ( I am trying to document all the operational scenarios / failures
> and how to recover from them. This document will act as a handbook
> for ops )
>
> Thanks,
> DJ
>
> On Thu, Jan 26, 2012 at 6:19 PM, Mohammad Islam <mi...@yahoo.com>
> wrote:
>
> > Hi DJ,
> > Sorry for the inconvenience.
> >
> > I think the issue is resolved.
> > Did you try with the latest 3.x version?
> >
> > Regards,
> > Mohammad
> >
> >
> > ----- Original Message -----
> > From: dejace sr <de...@gmail.com>
> > To: oozie-users@incubator.apache.org; mayank@apache.org
> > Cc:
> > Sent: Thursday, January 26, 2012 5:45 PM
> > Subject: oozie rerun job command in 2.3.0
> >
> > Hi,
> >
> > we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
> > move to 3.x.
> >
> > One of the important oozie command is re-running of the job from the
> point
> > of its failure, but I get error code 500.
> >
> > oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
> > Error: HTTP error code: 500 : Internal Server Error
> >
> > I did check the below, but wanted to know if it has been fixed.
> > https://issues.apache.org/jira/browse/OOZIE-256
> >
> > Thanks,
> > DJ
> >
> >
>
Re: oozie rerun job command in 2.3.0
Posted by Mohammad Islam <mi...@yahoo.com>.
Forgot to add.
Please send me the exact exception trace that you see in your oozie log.
----- Original Message -----
From: dejace sr <de...@gmail.com>
To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>
Cc: "mayank@apache.org" <ma...@apache.org>
Sent: Friday, January 27, 2012 11:19 AM
Subject: Re: oozie rerun job command in 2.3.0
Hi Mohammad,
I think it was a typo from my end while drafting the mail.
I had provided the job config in the command and I get error code 500.
So do we have any fix for this in 2.3.0? Bcaz, I cant switch
to 3.x until the project switches to 3.x?
(Parallel, I am also filing a bug against cloudera which is supporting our
hadoop cluster)
Thanks for the reply.
Regards,
DJ
On Thu, Jan 26, 2012 at 11:37 PM, Mohammad Islam <mi...@yahoo.com> wrote:
>
>
> Hi DJ,
> In addition, 3.x has feature where you don't need to specify the
> skipnodes, if you want to rerun from the failed node.
>
>
>
> Most importantly:
> Just came to my notice . why are you not passing -config in your rerun
> command.
> Please see the link
>
>
> http://yahoo.github.com/oozie/releases/2.3.0/DG_CommandLineTool.html#Rerunning_a_Workflow_Job
>
>
> It is very important.
>
>
> Regards,
> Mohammad
>
>
>
> ________________________________
> From: dejace sr <de...@gmail.com>
> To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>;
> mayank@apache.org
> Sent: Thursday, January 26, 2012 6:27 PM
> Subject: Re: oozie rerun job command in 2.3.0
>
> Thanks Mohammad.
>
> Our current objective is to go to production with oozie 2.3.0
> and after some months we will be switching to 3.x
>
> so in 2.3.0, do we have any patch/fix or workaround for this?
> because, we don't want to rerun the passed tasks of the job, but
> instead, would like to rerun from the point of failed task in the job;
> this way, it saves time and resources.
>
> ( I am trying to document all the operational scenarios / failures
> and how to recover from them. This document will act as a handbook
> for ops )
>
> Thanks,
> DJ
>
> On Thu, Jan 26, 2012 at 6:19 PM, Mohammad Islam <mi...@yahoo.com>
> wrote:
>
> > Hi DJ,
> > Sorry for the inconvenience.
> >
> > I think the issue is resolved.
> > Did you try with the latest 3.x version?
> >
> > Regards,
> > Mohammad
> >
> >
> > ----- Original Message -----
> > From: dejace sr <de...@gmail.com>
> > To: oozie-users@incubator.apache.org; mayank@apache.org
> > Cc:
> > Sent: Thursday, January 26, 2012 5:45 PM
> > Subject: oozie rerun job command in 2.3.0
> >
> > Hi,
> >
> > we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
> > move to 3.x.
> >
> > One of the important oozie command is re-running of the job from the
> point
> > of its failure, but I get error code 500.
> >
> > oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
> > Error: HTTP error code: 500 : Internal Server Error
> >
> > I did check the below, but wanted to know if it has been fixed.
> > https://issues.apache.org/jira/browse/OOZIE-256
> >
> > Thanks,
> > DJ
> >
> >
>
Re: oozie rerun job command in 2.3.0
Posted by dejace sr <de...@gmail.com>.
Hi Mohammad,
I think it was a typo from my end while drafting the mail.
I had provided the job config in the command and I get error code 500.
So do we have any fix for this in 2.3.0? Bcaz, I cant switch
to 3.x until the project switches to 3.x?
(Parallel, I am also filing a bug against cloudera which is supporting our
hadoop cluster)
Thanks for the reply.
Regards,
DJ
On Thu, Jan 26, 2012 at 11:37 PM, Mohammad Islam <mi...@yahoo.com> wrote:
>
>
> Hi DJ,
> In addition, 3.x has feature where you don't need to specify the
> skipnodes, if you want to rerun from the failed node.
>
>
>
> Most importantly:
> Just came to my notice . why are you not passing -config in your rerun
> command.
> Please see the link
>
>
> http://yahoo.github.com/oozie/releases/2.3.0/DG_CommandLineTool.html#Rerunning_a_Workflow_Job
>
>
> It is very important.
>
>
> Regards,
> Mohammad
>
>
>
> ________________________________
> From: dejace sr <de...@gmail.com>
> To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>;
> mayank@apache.org
> Sent: Thursday, January 26, 2012 6:27 PM
> Subject: Re: oozie rerun job command in 2.3.0
>
> Thanks Mohammad.
>
> Our current objective is to go to production with oozie 2.3.0
> and after some months we will be switching to 3.x
>
> so in 2.3.0, do we have any patch/fix or workaround for this?
> because, we don't want to rerun the passed tasks of the job, but
> instead, would like to rerun from the point of failed task in the job;
> this way, it saves time and resources.
>
> ( I am trying to document all the operational scenarios / failures
> and how to recover from them. This document will act as a handbook
> for ops )
>
> Thanks,
> DJ
>
> On Thu, Jan 26, 2012 at 6:19 PM, Mohammad Islam <mi...@yahoo.com>
> wrote:
>
> > Hi DJ,
> > Sorry for the inconvenience.
> >
> > I think the issue is resolved.
> > Did you try with the latest 3.x version?
> >
> > Regards,
> > Mohammad
> >
> >
> > ----- Original Message -----
> > From: dejace sr <de...@gmail.com>
> > To: oozie-users@incubator.apache.org; mayank@apache.org
> > Cc:
> > Sent: Thursday, January 26, 2012 5:45 PM
> > Subject: oozie rerun job command in 2.3.0
> >
> > Hi,
> >
> > we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
> > move to 3.x.
> >
> > One of the important oozie command is re-running of the job from the
> point
> > of its failure, but I get error code 500.
> >
> > oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
> > Error: HTTP error code: 500 : Internal Server Error
> >
> > I did check the below, but wanted to know if it has been fixed.
> > https://issues.apache.org/jira/browse/OOZIE-256
> >
> > Thanks,
> > DJ
> >
> >
>
Re: oozie rerun job command in 2.3.0
Posted by Mohammad Islam <mi...@yahoo.com>.
Hi DJ,
In addition, 3.x has feature where you don't need to specify the skipnodes, if you want to rerun from the failed node.
Most importantly:
Just came to my notice . why are you not passing -config in your rerun command.
Please see the link
http://yahoo.github.com/oozie/releases/2.3.0/DG_CommandLineTool.html#Rerunning_a_Workflow_Job
It is very important.
Regards,
Mohammad
________________________________
From: dejace sr <de...@gmail.com>
To: oozie-users@incubator.apache.org; Mohammad Islam <mi...@yahoo.com>; mayank@apache.org
Sent: Thursday, January 26, 2012 6:27 PM
Subject: Re: oozie rerun job command in 2.3.0
Thanks Mohammad.
Our current objective is to go to production with oozie 2.3.0
and after some months we will be switching to 3.x
so in 2.3.0, do we have any patch/fix or workaround for this?
because, we don't want to rerun the passed tasks of the job, but
instead, would like to rerun from the point of failed task in the job;
this way, it saves time and resources.
( I am trying to document all the operational scenarios / failures
and how to recover from them. This document will act as a handbook
for ops )
Thanks,
DJ
On Thu, Jan 26, 2012 at 6:19 PM, Mohammad Islam <mi...@yahoo.com> wrote:
> Hi DJ,
> Sorry for the inconvenience.
>
> I think the issue is resolved.
> Did you try with the latest 3.x version?
>
> Regards,
> Mohammad
>
>
> ----- Original Message -----
> From: dejace sr <de...@gmail.com>
> To: oozie-users@incubator.apache.org; mayank@apache.org
> Cc:
> Sent: Thursday, January 26, 2012 5:45 PM
> Subject: oozie rerun job command in 2.3.0
>
> Hi,
>
> we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
> move to 3.x.
>
> One of the important oozie command is re-running of the job from the point
> of its failure, but I get error code 500.
>
> oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
> Error: HTTP error code: 500 : Internal Server Error
>
> I did check the below, but wanted to know if it has been fixed.
> https://issues.apache.org/jira/browse/OOZIE-256
>
> Thanks,
> DJ
>
>
Re: oozie rerun job command in 2.3.0
Posted by dejace sr <de...@gmail.com>.
Thanks Mohammad.
Our current objective is to go to production with oozie 2.3.0
and after some months we will be switching to 3.x
so in 2.3.0, do we have any patch/fix or workaround for this?
because, we don't want to rerun the passed tasks of the job, but
instead, would like to rerun from the point of failed task in the job;
this way, it saves time and resources.
( I am trying to document all the operational scenarios / failures
and how to recover from them. This document will act as a handbook
for ops )
Thanks,
DJ
On Thu, Jan 26, 2012 at 6:19 PM, Mohammad Islam <mi...@yahoo.com> wrote:
> Hi DJ,
> Sorry for the inconvenience.
>
> I think the issue is resolved.
> Did you try with the latest 3.x version?
>
> Regards,
> Mohammad
>
>
> ----- Original Message -----
> From: dejace sr <de...@gmail.com>
> To: oozie-users@incubator.apache.org; mayank@apache.org
> Cc:
> Sent: Thursday, January 26, 2012 5:45 PM
> Subject: oozie rerun job command in 2.3.0
>
> Hi,
>
> we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
> move to 3.x.
>
> One of the important oozie command is re-running of the job from the point
> of its failure, but I get error code 500.
>
> oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
> Error: HTTP error code: 500 : Internal Server Error
>
> I did check the below, but wanted to know if it has been fixed.
> https://issues.apache.org/jira/browse/OOZIE-256
>
> Thanks,
> DJ
>
>
Re: oozie rerun job command in 2.3.0
Posted by Mohammad Islam <mi...@yahoo.com>.
Hi DJ,
Sorry for the inconvenience.
I think the issue is resolved.
Did you try with the latest 3.x version?
Regards,
Mohammad
----- Original Message -----
From: dejace sr <de...@gmail.com>
To: oozie-users@incubator.apache.org; mayank@apache.org
Cc:
Sent: Thursday, January 26, 2012 5:45 PM
Subject: oozie rerun job command in 2.3.0
Hi,
we are currently using Oozie server build version: 2.3.0-cdh3u1 until we
move to 3.x.
One of the important oozie command is re-running of the job from the point
of its failure, but I get error code 500.
oozie job -rerun <<jobid>> oozie.wf.rerun.skip.nodes=node1,node2
Error: HTTP error code: 500 : Internal Server Error
I did check the below, but wanted to know if it has been fixed.
https://issues.apache.org/jira/browse/OOZIE-256
Thanks,
DJ