Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (nike.apache.org: domain of jgray@facebook.com designates
 69.63.184.105 as permitted sender)
From: Jonathan Gray <jgray@facebook.com>
To: "user@hbase.apache.org" <user@hbase.apache.org>
Subject: RE: row lost after add-delete-add operation
Thread-Topic: row lost after add-delete-add operation
Thread-Index: AQHLJAq/RUTpned83kqhnhOwktsrDZKyGy3Q
Date: Thu, 15 Jul 2010 15:24:24 +0000
Message-ID: 
 <5A76F6CE309AD049AAF9A039A3924282071A21AE@sc-mbx04.TheFacebook.com>
References: <AANLkTikoeJ4HqoI2FXPFujnEkxVHhyWWJOn8m0Bf8D8_@mail.gmail.com>
In-Reply-To: <AANLkTikoeJ4HqoI2FXPFujnEkxVHhyWWJOn8m0Bf8D8_@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

This is not a problem with the KeyValueSkipListSet but rather it is the del=
ete tombstone continuing to override any Puts with a timestamp <=3D to the =
deleted stamp.

Check out this blog post for details on this:

http://outerthought.org/blog/blog/417-OTC.html

(man this thing is coming in handy, great work bruno)

JG

> -----Original Message-----
> From: Angus He [mailto:angushe@gmail.com]
> Sent: Thursday, July 15, 2010 3:43 AM
> To: user@hbase.apache.org
> Subject: row lost after add-delete-add operation
>=20
> Suppose we have a row : row1, cf1:q1, value1, timestamp1.
>=20
> After the add-delete-add operation sequence on the same row,  the row
> lost.
>=20
> It seems that it was KeyValueSkipListSet in class MemStore ignored the
> last add operation.
> I would like to know what is the purpose of KeyValueSkipListSet,
> is it possible to replace KeyValueSkipListSet with a multi-value set
> implementation?
>=20
>=20
>=20
> --
> Regards
> Angus