50-100 ops/sec on a 5 Node cluster
yousuffauzan at gmail.com
Tue Jun 26 12:36:03 EDT 2012
I used ephemeral storage without any improvement.
Also iotop and iostat did not show any saturation.
After talking with Sean Cribbs on IRC, I tried disabling nagle in riak
config, changed IO scheduler to noop, decreased cache size to 1 MB from 372
MB, removed load balancer.
After doing all this too, there was absolutely no change in throughput.
For Basho bench, I used examples/riakc_pb.config.
On Tuesday, June 26, 2012, Reid Draper <reiddraper at gmail.com> wrote:
> On Jun 26, 2012, at 11:48 AM, David Smith wrote:
>> What does your basho_bench config file look like?
>> Seeing a low number of ops/sec is usually an indicator that you're
>> hitting a bottleneck on the I/O path -- typically due to random reads.
>> LevelDB has a number of characteristics that can make this worse --
>> what happens if you use bitcask?
>> On Tue, Jun 26, 2012 at 4:57 AM, Yousuf Fauzan <yousuffauzan at gmail.com>
>>> Here is my setup of Riak 1.1.4
>>> 5 Nodes
>>> ami-a29943cb (EC2 AMI)
>>> m1.large (7.5GB Ram, 4 EC2 compute instances)
>>> EBS storage
> I'd recommend giving ephemeral storage a shot as well. It tends to have
significantly less variable performance than EBS.
>>> LevelDB backend
>>> Python Client with Protobuf (on a separate machine)
>>> While loading data I am getting around 50 ops/sec.
>>> I tried running basho benchmark on the setup and it showed 100 ops/sec
>>> After talking to people or Riak IRC, I realized that upwards of 1.5K
>>> is what I should be getting.
>>> Could anyone help me in figuring out what I am doing wrong?
>>> riak-users mailing list
>>> riak-users at lists.basho.com
>> Dave Smith
>> VP, Engineering
>> Basho Technologies, Inc.
>> dizzyd at 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