RiakSearch Benchmark Test Invitation.
prometheus.willsurvive at gmail.com
Sat Oct 30 05:42:57 EDT 2010
I have first benchmark results :
Test Data : http://rapidshare.com/files/427591191/wikipedia350.tar.gz 350K doc from wikipedia. Ready to index by riaksearch (search-cmd solr index ..xml)
Cluster : 1x2core , 1x2core , 1x6 core each has 8 gig ram. Gig Network
Nginx reverse proxy in a sperate machine.
Indexing time around : 10 min for 20K doc
Avarage qps : 3700 ms . and min qps is around 2200 ms
Is there anyone that make this test ? I expected better results from riaksearch but may be I make some mistakes in the test i dont know.
In our projects it is important to have qps below the 900 ms .
Waiting your test results guy.. come on.
On Oct 28, 2010, at 7:02 PM, Rusty Klophaus wrote:
> Hi folks,
> Very happy to see the excitement around Riak Search. Just a quick note on benchmarking approach. For best results, make sure to spread the indexing load across multiple machines in the cluster, rather than firing all requests on a single node. Otherwise, you will become CPU bound on that node. Load balancing in a round-robin fashion is fine.
> To make this easier, you may want to bypass the command line interface and post to Solr directly. In curl, it looks like this:
> curl -X POST -H text/xml --data-binary @datafile.xml http://hostname:8098/solr/myindex/update
> (Change the name of the datafile, hostname, and index appropriately.)
> On Thu, Oct 28, 2010 at 6:46 AM, Prometheus WillSurvive <prometheus.willsurvive at gmail.com> wrote:
> Hi Guys,
> I just put the wikipedia riaksearch solr index ready XMLs to the:
> you can download from there.
> there is also a small keyword list for benchmark test.
> We can put bigger documents later ie 3 million wikipedia doc.
> Let us know your test results. I used Apache Jmeter to send 10 clients queering to the clusters (3 machine)
> Best Regards
> On Oct 28, 2010, at 12:28 PM, Neville Burnell wrote:
>> Put it on S3
>> On 28 October 2010 20:20, francisco treacy <francisco.treacy at gmail.com> wrote:
>> Very good idea!
>> 2010/10/28 Prometheus WillSurvive <prometheus.willsurvive at gmail.com>:
>> > Hi All,
>> > We have prepare wikipedia database output ready to submit RiakSearch. It is
>> > XML and described format for solr submit. Each file has 20.000 Document and
>> > totaly 15 xml files. Each file around 44 MB.
>> > You can submit all XML 's = bin/search-cmd solr wikipedia
>> > /wikipedia/content-xml-out/wikipedia_1.xml
>> > So you only need to submit this files to the riaksearch and than make a
>> > benchmark test/tune and share your experience.
>> > I would like to ask Riak Admin guys is there any place that I can share
>> > these files for public access to start collaborative tests ?
>> > Second phase I can put 3 million wikipedia XML sets to ready to submit
>> > riaksearch. So All we have some common benchmark and tuning parameters.
>> > I hope this will help the riaksearch community to better understanding its
>> > capability.
>> > Best Regards
>> > _______________________________________________
>> > riak-users mailing list
>> > riak-users at lists.basho.com
>> > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
>> riak-users mailing list
>> riak-users at lists.basho.com
> riak-users mailing list
> riak-users at lists.basho.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the riak-users