Hi, Constantin
You can try to use Apache Spark to complete the mapreduce bulkload job. As far as I know, bulkloading into
phoenix or hbase may be affected by several conditions, like wal enabled or numbers of split regions. And your hbase
or phoenix configuration parameter may also influence the bulkloading performance. 

You can share more about your specific data loading information and I can help you do some tuning work.

Thanks,
Sun.



CertusNet 


From: Ciureanu, Constantin (GfK)
Date: 2015-01-13 17:12
To: user@phoenix.apache.org
Subject: MapReduce bulk load into Phoenix table

Hello all,

 

(Due to the slow speed of Phoenix JDBC – single machine ~ 1000-1500 rows /sec) I am also documenting myself about loading data into Phoenix via MapReduce.

 

So far I understood that the Key + List<[Key,Value]> to be inserted into HBase table is obtained via a “dummy” Phoenix connection – then those rows are stored into HFiles (then after the MR job finishes it is Bulk loading those HFiles normally into HBase).

 

My question: Is there any better / faster approach? I assume this cannot reach the maximum speed to load data into Phoenix / HBase table.

  

Also I would like to find a better / newer sample code than this one:

http://grepcode.com/file/repo1.maven.org/maven2/org.apache.phoenix/phoenix/4.0.0-incubating/org/apache/phoenix/mapreduce/CsvToKeyValueMapper.java#CsvToKeyValueMapper.loadPreUpsertProcessor%28org.apache.hadoop.conf.Configuration%29

 

Thank you,

   Constantin