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
MapReduce >> mail # user >> HBase MR: run more map tasks than regions


Copy link to this message
-
HBase MR: run more map tasks than regions
Hello,

As far as I know, the number of map tasks for "scan-based" mapreduce job is
equal (not more than) number of underlying regions (for scan). Of course, if
the max map task capacity is big enough.
I have a situation, when map-side processing is very heavy but uses quite
small amount of records from the HBase table. It may occur that those
records belongs to one or several regions and this results in running just
several map tasks. This may make processing very slow without utilising all
of the cluster resources :(. Is there a way to set the minimal number (or
just particular number) of map tasks in this situation? Is the only way is
to enhance TableInputFormat for me?

Thank you,

Alex Baranau
---
http://sematext.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