Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (nike.apache.org: domain of jamesgolick@gmail.com
 designates 209.85.217.209 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:date:message-id:subject:from:to:content-type;
        b=xFoP0bpMPwe4t06VFOj50CUYWOEkL9mZsvbgaYl/Lsn930ite3LBDX7Uq4c6kOivsF
         I+EThmiTx/8SRSXTg6x5uTouFznAbaSsbkDNr66foMMVkQ+YkGIoaJ36YJW+hfb7waZn
         3QmTy9s/K1YXPZAYIYNpUjMxUscn60YZAPqag=
MIME-Version: 1.0
Date: Tue, 30 Mar 2010 20:47:48 -0700
Message-ID: <1ab2da821003302047h128aaed9g1a65f3972836b086@mail.gmail.com>
Subject: Read Performance
From: James Golick <jamesgolick@gmail.com>
To: cassandra-user@incubator.apache.org
Content-Type: multipart/alternative; boundary=0016e6d447d41402e00483109cb6

--0016e6d447d41402e00483109cb6
Content-Type: text/plain; charset=ISO-8859-1

We are starting to use cassandra to power our activity feed. The way we
organize our data is simple. "Event"s live in a CF called Events and are
keyed by a UUID. The timelines themselves live in a CF called Timelines,
which is keyed by user id (i.e. "1229") and contains a event uuids as column
names (sorted by TimeUUIDType).

To load a feed, we get a slice of the timeline CF for that user, then
multiget all of the corresponding events.

Loading the slice of the timeline is reasonably fast at 4-6ms. But,
multigetting the events is terribly slow - on the order of 35-100ms.

To alleviate the problem, we write events through to memcached and use a
memcached multiget in front of the cassandra multiget. We have enough cache
space to get upwards of a 99% hit rate, which makes loading the events
extremely fast, but it would be nice to make use of the 24GB of memory in
our cassandra nodes.

We're on 0.6, and I've enabled the row cache. It seems to have data in it,
but it's still slow.

So, am I doing something wrong, or is this the expected perf?

- James

--0016e6d447d41402e00483109cb6
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

We are starting to use cassandra to power our activity feed. The way we org=
anize our data is simple. &quot;Event&quot;s live in a CF called Events and=
 are keyed by a UUID. The timelines themselves live in a CF called Timeline=
s, which is keyed by user id (i.e. &quot;1229&quot;) and contains a event u=
uids as column names (sorted by TimeUUIDType).<div>
<br></div><div>To load a feed, we get a slice of the timeline CF for that u=
ser, then multiget all of the corresponding events.</div><div><br></div><di=
v>Loading the slice of the timeline is reasonably fast at 4-6ms. But, multi=
getting the events is terribly slow - on the order of 35-100ms.</div>
<div><br></div><div>To alleviate the problem, we write events through to me=
mcached and use a memcached multiget in front of the cassandra multiget. We=
 have enough cache space to get upwards of a 99% hit rate, which makes load=
ing the events extremely fast, but it would be nice to make use of the 24GB=
 of memory in our cassandra nodes.</div>
<div><br></div><div>We&#39;re on 0.6, and I&#39;ve enabled the row cache. I=
t seems to have data in it, but it&#39;s still slow.</div><div><br></div><d=
iv>So, am I doing something wrong, or is this the expected perf?</div><div>
<br></div><div>- James</div>

--0016e6d447d41402e00483109cb6--