Raw RegionScanner?

[Vladimir Rodionov <vr...@carrieriq.com>]

Question:

I have a M/R job which needs to scan HBase table in a most efficient way (as fast as possible)

1. Data locality matters. Each map task must connect to local RS and fetch ONLY data which are local to this particular node.
2. Order of rows does nor matters, actually - only requirement is : data is between startRow, endRow

Taking into account that with some new patches to HDFS we can get direct access to local FS, ideally I see this as following:

A. Map Task open scanner to HTable REGION which collocates on the same node physically
B. Data flow bypasses DN entirely and we can read directly from local FS or from OS page cache
C. We do not do any ordering during scan, which imposes additional overhead.

The question is: can I do this? All? or at least, partially?



Best regards,
Vladimir Rodionov
Principal Platform Engineer
Carrier IQ, www.carrieriq.com
e-mail: vrodionov@carrieriq.com

________________________________________
From: shaposhnik@gmail.com [shaposhnik@gmail.com] On Behalf Of Roman Shaposhnik [roman@shaposhnik.org]
Sent: Wednesday, December 21, 2011 11:59 AM
To: dev@hbase.apache.org
Cc: ramkrishna.vasudevan@huawei.com
Subject: Re: hadoop-core-1.0.0rc3.jar not able to open the zip file error

On Wed, Dec 21, 2011 at 7:18 AM, Stack <st...@duboce.net> wrote:
> Please paste maven error Ram.  It seems to work here for different
> machines.  Does it work for others?

It worked for me when I was doing BT build:
   http://bigtop01.cloudera.org:8080/view/RCs/job/Bigtop-trunk-rc-hadoop-1.0.0-hbase-0.92.0/4/label=centos5/

Thanks,
Roman.

Confidentiality Notice:  The information contained in this message, including any attachments hereto, may be confidential and is intended to be read only by the individual or entity to whom this message is addressed. If the reader of this message is not the intended recipient or an agent or designee of the intended recipient, please note that any review, use, disclosure or distribution of this message or its attachments, in any form, is strictly prohibited.  If you have received this message in error, please immediately notify the sender and/or Notifications@carrieriq.com and delete or destroy any copy of this message and its attachments.