incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Simon Guindon <simon.guin...@jsitelecom.com>
Subject RE: Concurrency and secondary indexes
Date Fri, 23 Nov 2012 17:34:08 GMT
Aaron,

Nodetool reports: 1.1.6-dse-p1

What about inserted data using Hadoop? Will the secondary index execute concurrently or will
this also be single threaded (on the same machine I mean)?

From: aaron morton [mailto:aaron@thelastpickle.com]
Sent: November-22-12 8:04 PM
To: user@cassandra.apache.org
Subject: Re: Concurrency and secondary indexes

What version are you on ?


 but we are finding a secondary index is performing slow
Not sure what you mean here.

Are secondary indexes concurrent or single threaded?
Rebuilding a secondary index (via node tool) is a single threaded operation, but *all* indexes
specified on the command line are built at the same time. Rebuilding any one index requires
reading all the rows in the CF.

Cheers

-----------------
Aaron Morton
Freelance Cassandra Developer
New Zealand

@aaronmorton
http://www.thelastpickle.com

On 23/11/2012, at 8:20 AM, Simon Guindon <simon.guindon@jsitelecom.com<mailto:simon.guindon@jsitelecom.com>>
wrote:


We are importing data from one column family into a second column family via "nodetool refresh"
but we are finding a secondary index is performing slow and the machine CPU is pretty much
idle. We are trying to bulk load data as fast as possible.

Are secondary indexes concurrent or single threaded?


Mime
View raw message