No, you have to restart the job and task trackers for that setting to
take effect.
On Sat, Aug 13, 2011 at 3:36 PM, Xiangping Bu wrote:
> Hi, All
>
> Is that possible to reconfigure the number of map/red slots per
> machine without restarting the Hadoop cluster?
>
> Thanks
>
> Xiangping
>
:
>> This looks very useful and looks like nice work.
>>
>> I note that the methods used are prone to race conditions, but if you are
>> just thinking about shared maps, this probably isn't important.
>>
>> On Wed, Dec 8, 2010 at 12:40 PM, Anthony Urso wrote:
&
oks very useful and looks like nice work.
>
> I note that the methods used are prone to race conditions, but if you are
> just thinking about shared maps, this probably isn't important.
>
> On Wed, Dec 8, 2010 at 12:40 PM, Anthony Urso wrote:
>
>> I am pleased to announce the
I am pleased to announce the initial release of KeptCollections, a
library of drop-in replacements for standard Java Collections that use
Apache ZooKeeper as a backing store.
KeptCollections are designed to make it easy for anyone to write
distributed applications without having to learn the intri
I am pleased to announce the v0.0 release of Sizzle, a compiler and
runtime for the Sawzall language. Sizzle targets Hadoop directly, by
compiling Sawzall programs into Hadoop job jars that can be run
anywhere Hadoop is installed, without requiring a Sawzall interpreter
to also be present.
Althoug
Try the fair scheduler, it will seem more simultaneous than the
default scheduler.
http://hadoop.apache.org/mapreduce/docs/r0.21.0/fair_scheduler.html
On Mon, Sep 13, 2010 at 2:33 PM, Rahul Malviya wrote:
> Hi,
>
> I am running Pig jobs on Hadoop cluster.
>
> I just wanted to know whether I can
The new Mapper class passes a Map.Context object to the map() method.
>From this you can get a StatusReporter object which can produce a
named Counter object and increment it.
http://hadoop.apache.org/common/docs/current/api/org/apache/hadoop/mapreduce/Mapper.Context.html
http://hadoop.apache.or
The dirty way to do this is to have the reducer throw an exception if
it receives a key that was not intended for the node it is running on.
It will be rescheduled on another node, and eventually it will land
on the correct one.
Depending on the total number of nodes and reducers in the job, you
