Such a mix won't work, and hence my (1) point of check. Please use a
jar that matches the server version, and you'll have it working.
On Thu, Nov 29, 2012 at 12:53 AM, Pedro Sá da Costa wrote:
> And for this error, after all maybe I'm running hadoop.jar with
> different versions. I'm running hado
And for this error, after all maybe I'm running hadoop.jar with
different versions. I'm running hadoop-0.20 and trying to run
JobClient in with hadoo-1.0
On 28 November 2012 19:20, Pedro Sá da Costa wrote:
> I've this error in jobtracker log. Maybe this is the reason. What this
> error means?
>
>
I've this error in jobtracker log. Maybe this is the reason. What this
error means?
2012-11-28 19:19:17,697 WARN org.apache.hadoop.ipc.Server: Incorrect
header or version mismatch from 127.0.0.1:60089 got version 4 expected
version 3
On 28 November 2012 18:28, Pedro Sá da Costa wrote:
> On 28
On 28 November 2012 18:12, Harsh J wrote:
> nt application's hadoop jar same version as the server?
Yes it is.
> 2. Is the port 54311 the proper JobTracker port?
This jobtracker port is set to:
mapred.job.tracker
localhost:54311
The host and port that the MapReduce job tracker runs
a
Two things to check and address if not true:
1. Is your client application's hadoop jar same version as the server?
2. Is the port 54311 the proper JobTracker port?
On Wed, Nov 28, 2012 at 11:37 PM, Pedro Sá da Costa wrote:
> Yes, I already did like you tell, but I get an error in the
> connecti
Yes, I already did like you tell, but I get an error in the
connection, and this doesn't make any sense.
java.io.IOException: Call to localhost/127.0.0.1:54311 failed on local
exception: java.io.EOFException
at org.apache.hadoop.ipc.Client.wrapException(Client.java:1107)
at org.apa
Mahesh,
The JobTracker is not a user-end API, it is a service. You can't query
it for its private objects as simply as that, and hence what you've
suggested would not work.
Pedro,
To get a RunningJob instance from a String representing the JobID you
need the JobClient class, and you can do like
I have the jobId as a String, and from that I want to access the
RunningJob API for that jobId. I think that it is only possible to
access this API through the JobInProgress class, but maybe I'm wrong.
Is this true?
On 28 November 2012 17:24, Mahesh Balija wrote:
> Hi Pedro,
>
> You ca
The JobInProgress object is a private one and can't be fetched unless
you are running a service within the JobTracker environment.
What are you looking for, from the JobInProgress' information,
specifically, that is not found in the RunningJob interface provided
to clients?
On Wed, Nov 28, 2012 a
I'm building a Java class and given a JobID, how can I get the
JobInProgress? Can anyone give me an example?
--
Best regards,
I am not sure of an existing tool to do that, but there are APIs under
the RunningJob interface that gives you access to such data, so you
could write your own.
On Wed, Nov 28, 2012 at 5:14 PM, Pedro Sá da Costa wrote:
> Yes I can, but I want more details about the tasks, like the time that
> the
Yes I can, but I want more details about the tasks, like the time that
they start, ended, the duration of the shuffle. I want as much
information as the "hadoop job history all" command can give it, but I
want as the job progress.
On 28 November 2012 11:32, Harsh J wrote:
> hadoop job -status
You can repeatedly query with `hadoop job -status `.
On Wed, Nov 28, 2012 at 4:45 PM, Pedro Sá da Costa wrote:
> Hadoop Mapreduce has an web interface that shows the progress of
> running jobs. Can I get the same information about the job progress in
> bash? There's a program to print the progres
Hadoop Mapreduce has an web interface that shows the progress of
running jobs. Can I get the same information about the job progress in
bash? There's a program to print the progress in the terminal?
Thanks,
--
Best regards,
14 matches
Mail list logo
