How do I improve Level DB performance?
rzezeski at basho.com
Fri May 11 12:13:41 EDT 2012
On Thu, May 10, 2012 at 11:14 PM, Tim Haines <tmhaines at gmail.com> wrote:
> With the adjusted ring size and settings, and adjusted to only do puts (so
> no missed reads), my cluster is doing about 400 puts per second:
Actually, every put (put from a riak API level) does a read on the backend
. This is needed to merge contents from the two objects .
Like Dave already mentioned the key generation strategy along with
leveldb's degrading performance on not-found means your benchmark will just
get worse the longer it runs.
Are you testing an actual use case here? Do you envision 100M objects
being written in a constant stream? Will your objects have a median size
of 1000 bytes? Basho bench also provides a pareto key generator which uses
a fraction of the key space most of the time. I'm not sure it matches your
use case but thought I'd mention it is there.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the riak-users