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
HBase >> mail # user >> HBase client connection in application server


Copy link to this message
-
Re: HBase client connection in application server
That what we are doing at Salesforce and the reason why I added this API.
Why also create our thread pool, though, since the default one is too small.

I blogged about this here: http://hadoop-hbase.blogspot.com/2011/12/long-running-hbase-clients.html

-- Lars

________________________________
 From: Amit Sela <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Sent: Sunday, January 19, 2014 2:10 AM
Subject: HBase client connection in application server
 

I need to connect to HBase (0.94.12) from an application server - Jetty
(Felix).

Documentation says the way to go is to get HConnection with
HConnectionManager.createConnection(Configuration) use it to get table,
close table and close connection.
I also see that when creating an HConnection, a batchpool is created (lazy).

Can I use one instance of HConnection to serve all HBase client requests on
the server and close it only when the server goes down ? will it affect
performance ? go stale ? how about regions cache, if new regions are
created, do I need to create a new HConnection ?

I think it might be a good idea to add a detailed explanation about HBase
client management in the manual, maybe some best-practice advice or
something like that.

Thanks,

Amit.
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