Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Threaded View
Hive >> mail # user >> How does hive decide to launch how many map tasks?


Copy link to this message
-
Re: How does hive decide to launch how many map tasks?
Thank you so much :)

On Fri, Nov 16, 2012 at 5:49 PM, Bejoy KS <[EMAIL PROTECTED]> wrote:
> Hi Chen
>
> The computation on the number of Input Splits/ map tasks is totally determined by the InputFormat used as well as the split size.
>
> Hive used CombineHiveInput format so you may not be having one mapper per file if your files are small. You can control the number of maps by controlling the split sizes.
> Mapred.min.split.size
> Mapred.max.split.size
>
> Regards
> Bejoy KS
>
> Sent from handheld, please excuse typos.
>
> -----Original Message-----
> From: Cheng Su <[EMAIL PROTECTED]>
> Date: Fri, 16 Nov 2012 14:39:57
> To: <[EMAIL PROTECTED]>
> Reply-To: [EMAIL PROTECTED]
> Subject: How does hive decide to launch how many map tasks?
>
> Hi, all
>
> How does hive decide to launch how many map tasks?
> I know there are some configs to help hive to decide how many reduce
> task to launch?
> But how about map tasks?
>
> I thought that number of map tasks equals to the number of the store files.
> I have a table now with 2 partitions, and one has 4 files in it, the
> other has 2,
> when I execute "select count(*) from table", only one map is launched.
>
> How can I increase the number of map tasks to improve the performance?
>
> Thanks.
>
> --
>
> Regards,
> Cheng Su

--

Regards,
Cheng Su
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB