Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (athena.apache.org: domain of yuzhihong@gmail.com
 designates 209.85.220.52 as permitted sender)
From: Ted Yu <yuzhihong@gmail.com>
Content-Type: text/plain;
	charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (1.0)
Subject: Re: Multiple scans for mapreduce over snapshots
Message-Id: <8507952B-F082-44AF-BDCF-1CE7F5FAD353@gmail.com>
Date: Fri, 27 Mar 2015 23:02:32 -0700
References: <5515A729.5030309@upsight.com>
In-Reply-To: <5515A729.5030309@upsight.com>
To: "user@hbase.apache.org" <user@hbase.apache.org>

I think adding support for multi table scan over snapshots would be a nice f=
eature to have.=20


> On Mar 27, 2015, at 11:53 AM, Andrew Mains <andrew.mains@kontagent.com> wr=
ote:
>=20
> Hi all,
>=20
> We're looking into using TableSnapshotInputFormat on a salted table, and w=
e need to push down conditions on the rest of the rowkey to each bucket (usi=
ng hive with my patch for https://issues.apache.org/jira/browse/HIVE-7805). M=
ultiTableInputFormat allows us to do this on HBase proper, but it seems like=
 this isn't yet supported on snapshots. There's nothing in either google or J=
IRA discussing such a feature afaict, so I thought I'd ask here:
>=20
> Would it be reasonable for HBase to support an equivalent of `MultiTableIn=
putFormat` over snapshots? Is there a better alternative that we should be u=
sing instead?
>=20
> Thanks!
>=20
> Andrew