Mailing-List: contact user-help@curator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@curator.apache.org
Received-SPF: pass (nike.apache.org: domain of rkanter@cloudera.com designates
 209.85.213.173 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <etPan.54caa632.189a769b.13d@Jordans-MacBook-Pro.local>
References: 
 <CAHz+ZFfiByu6dOZpEXDa3zGePc9iE-gC26KWYYVhVk4snOubPw@mail.gmail.com>
	<etPan.54caa632.189a769b.13d@Jordans-MacBook-Pro.local>
Date: Fri, 30 Jan 2015 09:47:49 -0800
Message-ID: 
 <CAHz+ZFc2aP2As=1S7rizDs5XCuMZFERo_tvUoyhKSid_XcLTug@mail.gmail.com>
Subject: Re: InterProcessSemaphoreMutex Scalability
From: Robert Kanter <rkanter@cloudera.com>
To: Jordan Zimmerman <jordan@jordanzimmerman.com>
Cc: user@curator.apache.org
Content-Type: multipart/alternative; boundary=001a113ed99a04e5a2050de2346d

--001a113ed99a04e5a2050de2346d
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Great, thanks!

On Thu, Jan 29, 2015 at 1:29 PM, Jordan Zimmerman <
jordan@jordanzimmerman.com> wrote:

> The major limitation is that, by default, ZooKeeper API calls are limited
> to 1MB. The lock recipe calls getChildren() and the concatenated list of
> children can=E2=80=99t be more than 1MB. The node names contain the prefi=
x =E2=80=9Clease-=E2=80=9C
> (6 chars), a UUID (36 chars) and the sequence number (10 chars). So, that=
=E2=80=99s
> 52 chars per lock. I don=E2=80=99t know if there=E2=80=99s other overhead=
. But, anyway,
> that gives you room for something like 19K lock participants. So, 600 is =
no
> problem. Additionally, each lock participant only watches it=E2=80=99s pr=
eceding
> participant so you won=E2=80=99t get herding as the number of clients gro=
w.
>
> -Jordan
>
>
>
> On January 29, 2015 at 1:01:56 PM, Robert Kanter (rkanter@cloudera.com)
> wrote:
>
> Hi all,
>
> Is there any information on how scalable the InterProcessSemaphoreMutex
> is?  I'm working on somethere where, in some cases, there could be ~600
> clients trying to acquire the same lock.  Will that be a problem?
> I'd go and test this directly, but I don't currently have enough machines
> to try it with so I was hoping someone has already done some testing in
> this area.
>
> thanks
> - Robert
>
>

--001a113ed99a04e5a2050de2346d
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Great, thanks!</div><div class=3D"gmail_extra"><br><div cl=
ass=3D"gmail_quote">On Thu, Jan 29, 2015 at 1:29 PM, Jordan Zimmerman <span=
 dir=3D"ltr">&lt;<a href=3D"mailto:jordan@jordanzimmerman.com" target=3D"_b=
lank">jordan@jordanzimmerman.com</a>&gt;</span> wrote:<br><blockquote class=
=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padd=
ing-left:1ex"><div style=3D"word-wrap:break-word"><div style=3D"font-family=
:Helvetica,Arial;font-size:13px;color:rgba(0,0,0,1.0);margin:0px;line-heigh=
t:auto">The major limitation is that, by default, ZooKeeper API calls are l=
imited to 1MB. The lock recipe calls getChildren() and the concatenated lis=
t of children can=E2=80=99t be more than 1MB. The node names contain the pr=
efix =E2=80=9Clease-=E2=80=9C (6 chars), a UUID (36 chars) and the sequence=
 number (10 chars). So, that=E2=80=99s 52 chars per lock. I don=E2=80=99t k=
now if there=E2=80=99s other overhead. But, anyway, that gives you room for=
 something like 19K lock participants. So, 600 is no problem. Additionally,=
 each lock participant only watches it=E2=80=99s preceding participant so y=
ou won=E2=80=99t get herding as the number of clients grow.</div><span clas=
s=3D"HOEnZb"><font color=3D"#888888"><div style=3D"font-family:Helvetica,Ar=
ial;font-size:13px;color:rgba(0,0,0,1.0);margin:0px;line-height:auto"><br><=
/div><div style=3D"font-family:Helvetica,Arial;font-size:13px;color:rgba(0,=
0,0,1.0);margin:0px;line-height:auto">-Jordan</div></font></span><div><div =
class=3D"h5"> <br> <div><div style=3D"font-family:helvetica,arial;font-size=
:13px"><br></div></div> <br><p style=3D"color:#000">On January 29, 2015 at =
1:01:56 PM, Robert Kanter (<a href=3D"mailto:rkanter@cloudera.com" target=
=3D"_blank">rkanter@cloudera.com</a>) wrote:</p> <blockquote type=3D"cite">=
<span><div><div></div><div>


<div dir=3D"ltr">Hi all,
<div><br></div>
<div>Is there any information on how scalable
the=C2=A0InterProcessSemaphoreMutex is?=C2=A0 I&#39;m working on
somethere where, in some cases, there could be ~600 clients trying
to acquire the same lock.=C2=A0 Will that be a problem?
=C2=A0</div>
<div>I&#39;d go and test this directly, but I don&#39;t currently have
enough machines to try it with so I was hoping someone has already
done some testing in this area. =C2=A0</div>
<div><br></div>
<div>thanks</div>
<div>- Robert</div>
</div>


</div></div></span></blockquote></div></div></div></blockquote></div><br></=
div>

--001a113ed99a04e5a2050de2346d--