Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
MIME-Version: 1.0
In-Reply-To: 
 <CA+z1Z6ec5-ii42a-O7sCxfRgTF-3KpeD96huun88cbsQNRJvLg@mail.gmail.com>
References: 
 <CA+z1Z6ec5-ii42a-O7sCxfRgTF-3KpeD96huun88cbsQNRJvLg@mail.gmail.com>
From: Carl Yeksigian <carl@yeksigian.com>
Date: Mon, 4 Jan 2016 16:14:34 -0500
Message-ID: 
 <CAGTwgn4XTRxOGdcE_nQac6dwq0VrJFKL6Of9YS5jQCMDDrBSJA@mail.gmail.com>
Subject: Re: compaction_throughput_mb_per_sec
To: user@cassandra.apache.org
Content-Type: multipart/alternative; boundary=001a113f82645c3e6c0528889dea

--001a113f82645c3e6c0528889dea
Content-Type: text/plain; charset=UTF-8

This is set in the cassandra.yaml on each node independently; it doesn't
have to be same cluster-wide.

On Mon, Jan 4, 2016 at 3:59 PM, Ken Hancock <ken.hancock@schange.com> wrote:

> I was surprised the other day to discover that this was a cluster-wide
> setting.   Why does that make sense?
>
> In a heterogeneous cassandra deployment, say I have some old servers
> running spinning disks and I'm bringing on more nodes that perhaps utilize
> SSD.  I want to have different compaction throttling  on different nodes to
> minimize read impact times.
>
> I can already balance data ownership through either token allocation or
> vnode counts.
>
> Also, as I increase my node count, I technically also have to increase my
> compaction_throughput which would require a rolling restart across the
> cluster.
>
>
>

--001a113f82645c3e6c0528889dea
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>This is set in the cassandra.yaml on each node indepe=
ndently; it doesn&#39;t have to be same cluster-wide.<br></div></div><div c=
lass=3D"gmail_extra"><br><div class=3D"gmail_quote">On Mon, Jan 4, 2016 at =
3:59 PM, Ken Hancock <span dir=3D"ltr">&lt;<a href=3D"mailto:ken.hancock@sc=
hange.com" target=3D"_blank">ken.hancock@schange.com</a>&gt;</span> wrote:<=
br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left=
:1px #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div><div><div>I was sur=
prised the other day to discover that this was a cluster-wide setting.=C2=
=A0=C2=A0 Why does that make sense?<br><br></div>In a heterogeneous cassand=
ra deployment, say I have some old servers running spinning disks and I&#39=
;m bringing on more nodes that perhaps utilize SSD.=C2=A0 I want to have di=
fferent compaction throttling=C2=A0 on different nodes to minimize read imp=
act times.<br><br></div>I can already balance data ownership through either=
 token allocation or vnode counts. <br><br></div>Also, as I increase my nod=
e count, I technically also have to increase my compaction_throughput which=
 would require a rolling restart across the cluster.<br><br><br></div>
</blockquote></div><br></div>

--001a113f82645c3e6c0528889dea--