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: local policy)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.com;
  h=X-YMail-OSG:Received:X-Rocket-MIMEInfo:X-RocketYMMF:X-Mailer:References:Message-ID:Date:From:Reply-To:Subject:To:In-Reply-To:MIME-Version:Content-Type;
  b=U2+I9JiVaKqr+GjaOpHyHjinwL6DCWkrnRelALDCYUsmaW3xiGPj/Mb8+quqDQs9DmXUALLxzBkAFzvDaqfqbkupeCm63SbrbniIOBosSK8cEQXfBrli4XBKP5UW4ht7Oni6qytA83U2HgwdYDH7tilb6VlNr31DMh2up3o0oLI=;
References: 
 <CANNBgPLCRtPi33htgT5qXscp1W5HBqXv7AeyJJ+qSNuNJYzEDw@mail.gmail.com>
Message-ID: <1361261530.94194.YahooMailNeo@web140601.mail.bf1.yahoo.com>
Date: Tue, 19 Feb 2013 00:12:10 -0800 (PST)
From: lars hofhansl <larsh@apache.org>
Reply-To: lars hofhansl <larsh@apache.org>
Subject: Re: HBase without compactions?
To: "user@hbase.apache.org" <user@hbase.apache.org>
In-Reply-To: 
 <CANNBgPLCRtPi33htgT5qXscp1W5HBqXv7AeyJJ+qSNuNJYzEDw@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/alternative;
 boundary="969045052-1116646268-1361261530=:94194"

--969045052-1116646268-1361261530=:94194
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

If you store data in LSM trees you need compactions.=0AThe advantage is tha=
t your data files are immutable.=0AMapR has a mutable file system and they =
probably store their data in something more akin to B-Trees...?=0AOr maybe =
they somehow avoid the expensive merge sorting of many small files. It seem=
s that is has to be one or the other.=0A=0A(Maybe somebody from MapR reads =
this and can explain how it actually works.)=0A=0ACompations let you trade =
random IO for sequential IO (just to state the obvious). It seems that you =
can't have it both ways.=0A=0A-- Lars=0A=0A=0A=0A__________________________=
______=0A From: Otis Gospodnetic <otis.gospodnetic@gmail.com>=0ATo: user@hb=
ase.apache.org =0ASent: Monday, February 18, 2013 7:30 PM=0ASubject: HBase =
without compactions?=0A =0AHello,=0A=0AIt's kind of funny, we run SPM, whic=
h includes SPM for HBase (performance=0Amonitoring service/tool for HBase e=
ssentially) and we currently store all=0Aperformance metrics in HBase.=0A=
=0AI see a ton of HBase development activity, which is great, but it just=
=0Aoccurred to me that I don't think I recall seeing anything about getting=
=0Arid of compactions.=A0 Yet, compactions are one thing that I know hurt u=
s the=0Amost and is one thing that MapR somehow got rid of in their impleme=
ntation.=0A=0AHave there been any discussions,attempts, or thoughts about f=
inding a way=0Ato avoid compactions?=0A=0AThanks,=0AOtis=0A--=0AHBASE Perfo=
rmance Monitoring - http://sematext.com/spm/index.html
--969045052-1116646268-1361261530=:94194--