Mailing-List: contact common-dev-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: common-dev@hadoop.apache.org
Received-SPF: pass (nike.apache.org: domain of carp84@gmail.com designates
 209.85.216.176 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=tcTenpHsy17+Elck+/hVQQKS/crrpKi9Rw4sg1zEEHJMFbnTtz85nJyEb9a+Kl0Dy7
         noie2s0AUEki0iajunZvHXwCUjqnQBIv5bFJgi3+LCfnkgrTU7MeiZ2GiYnbMNzc3jMt
         8EzV2QCgj74XbhprIQZN3hZnzFSl+PPRgL5T0=
MIME-Version: 1.0
In-Reply-To: <AANLkTinXsAPCKyV1XXxhsODfUgQ4u_s-0RtOeXd2GNVQ@mail.gmail.com>
References: <AANLkTinVGEFPxmsKy5NDB3yEsamexwRK97i6mF3BAlyW@mail.gmail.com>
	<46A377B1A3A3074D8B989BF96663C10DF78EEA4A97@EGL-EX07VS01.ds.corp.yahoo.com>
	<AANLkTimlgHLtm99l5zm5kV93ntrcgbGbbMRB_v9d78hl@mail.gmail.com>
	<AANLkTikHjz7-xCjRSDsAyU0lWom2GCpcgWA2XmqEzUKx@mail.gmail.com>
	<AANLkTinNxAmv6cf9W5YQJ8LOCeOxjuAJzJpyTIo-dHoC@mail.gmail.com>
	<AANLkTinXsAPCKyV1XXxhsODfUgQ4u_s-0RtOeXd2GNVQ@mail.gmail.com>
Date: Thu, 24 Jun 2010 08:42:46 +0800
Message-ID: <AANLkTilLAs_B60pm0UoBiwkfxB6P0g8vo_N40oCE-dpx@mail.gmail.com>
Subject: Re: Questions about recommendation value of the "io.sort.mb"
	parameter
From: Yu Li <carp84@gmail.com>
To: common-dev@hadoop.apache.org
Content-Type: multipart/alternative; boundary=00c09fa543e9ea4ba00489bbee84

--00c09fa543e9ea4ba00489bbee84
Content-Type: text/plain; charset=GB2312
Content-Transfer-Encoding: quoted-printable

Hi Todd,

Thanks a lot for your further explanation! It makes me more clear about thi=
s
parameter.

BTW, please allow me to express my thankfulness to everyone helps.

Best Regards,
Carp

=D4=DA 2010=C4=EA6=D4=C224=C8=D5 =C9=CF=CE=E71:49=A3=ACTodd Lipcon <todd@cl=
oudera.com>=D0=B4=B5=C0=A3=BA

> Plus there's some overhead for each record of map output. Specifically, 2=
4
> bytes. So if you output 64MB worth of data, but each of your objects is
> only
> 24 bytes long itself, you need more than 128MB worth of spill space for i=
t.
> Last, the map output buffer begins spilling when it is partially full so
> that more records can be collected while spill proceeds.
>
> 200MB io.sort.mb has enough headroom for most 64M input splits that don't
> blow up the data a lot. Expanding much above 200M for most jobs doesn't b=
uy
> you much. Good news is it's easy to tell by looking at the logs to see ho=
w
> many times the map tasks are spilling. If you're only spilling once, more
> io.sort.mb will not help.
>
> -Todd
>
> 2010/6/23 =C0=EE=EE=DA <carp84@gmail.com>
>
> > Hi Jeff,
> >
> > Thanks for your quick reply. Seems my thinking is stuck on the job styl=
e
> > I'm
> > running. Now I'm much clearer about it.
> >
> > Best Regards,
> > Carp
> >
> > 2010/6/23 Jeff Zhang <zjffdu@gmail.com>
> >
> > > Hi =C0=EE=EE=DA
> > >
> > > The size of map output depends on your Mapper class. The Mapper class
> > > will do processing on the input data.
> > >
> > >
> > >
> > > 2010/6/23 =C0=EE=EE=DA <carp84@gmail.com>:
> > >  > Hi Sriguru,
> > > >
> > > > Thanks a lot for your comments and suggestions!
> > > > Here I still have some questions: since map mainly do data
> preparation,
> > > > say split input data into KVPs, sort and partition before spill,
> would
> > > the
> > > > size of map output KVPs be much larger than the input data size? If
> > not,
> > > > since one map task deals with one input split, and one input split =
is
> > > > usually 64M, the map KVPs size would be proximately 64M. Could you
> > please
> > > > give me some example on map output much larger than the input split=
?
> It
> > > > really confuse me for some time, thanks.
> > > >
> > > > Others,
> > > >
> > > > Also badly need your help if you know about this, thanks.
> > > >
> > > > Best Regards,
> > > > Carp
> > > >
> > > > =D4=DA 2010=C4=EA6=D4=C223=C8=D5 =CF=C2=CE=E75:11=A3=ACSrigurunath =
Chakravarthi <sriguru@yahoo-inc.com
> > >=D0=B4=B5=C0=A3=BA
> > > >
> > > >> Hi Carp,
> > > >>  Your assumption is right that this is a per-map-task setting.
> > > >> However, this buffer stores map output KVPs, not input. Therefore
> the
> > > >> optimal value depends on how much data your map task is generating=
.
> > > >>
> > > >> If your output per map is greater than io.sort.mb, these rules of
> > thumb
> > > >> that could work for you:
> > > >>
> > > >> 1) Increase max heap of map tasks to use RAM better, but not hit
> swap.
> > > >> 2) Set io.sort.mb to ~70% of heap.
> > > >>
> > > >> Overall, causing extra "spills" (because of insufficient io.sort.m=
b)
> > is
> > > >> much better than risking swapping (by setting io.sort.mb and heap
> too
> > > >> large), in terms of relative performance penalty you will pay.
> > > >>
> > > >> Cheers,
> > > >> Sriguru
> > > >>
> > > >> >-----Original Message-----
> > > >> >From: =C0=EE=EE=DA [mailto:carp84@gmail.com]
> > > >> >Sent: Wednesday, June 23, 2010 12:27 PM
> > > >> >To: common-dev@hadoop.apache.org
> > > >> >Subject: Questions about recommendation value of the "io.sort.mb"
> > > >> >parameter
> > > >> >
> > > >> >Dear all,
> > > >> >
> > > >> >Here I've got a question about the "io.sort.mb" parameter. We can
> > find
> > > >> >material from Yahoo! or Cloudera which recommend setting this val=
ue
> > to
> > > >> >200
> > > >> >if the job scale is large, but I'm confused about this. As I know=
,
> > > >> >the tasktracker will launch a child-JVM for each task, and
> > > >> >=A1=B0*io.sort.mb*=A1=B1
> > > >> >presents the buffer size in memory inside *one map task child-JVM=
*,
> > the
> > > >> >default value 100MB should be large enough because the input spli=
t
> of
> > > >> >one
> > > >> >map task is usually 64MB, as large as the block size we usually
> set.
> > > >> >Then
> > > >> >why the recommendation of =A1=B0*io.sort.mb*=A1=B1 is 200MB for l=
arge jobs
> (and
> > > >> >it
> > > >> >really works)? How could the job size affect the procedure?
> > > >> >Is there any fault here of my understanding? Any comment/suggesti=
on
> > > >> >will be
> > > >> >highly valued, thanks in advance.
> > > >> >
> > > >> >Best Regards,
> > > >> >Carp
> > > >>
> > > >
> > >
> > >
> > >
> > > --
> > > Best Regards
> > >
> > > Jeff Zhang
> > >
> >
>
>
>
> --
> Todd Lipcon
> Software Engineer, Cloudera
>

--00c09fa543e9ea4ba00489bbee84--