How random is the lookup for your data? Is it completely random or can
you rely on a fixed number of users to come in?
In the completely random case, random disk performance will be
essential. Depending on how fast your disk storage is, you can max out
the number of entries in your BDB storage engine until performance is
unacceptable( > 300 ms in your case.)
If your data is pretty hot and you can rely on a cache, you can
probably size your cache using this utility? You will want to make
sure that your data can mostly fit in cache, and perhaps misses are
not a big deal.
Take a look at:
http://sna-projects.com/blog/2009/08/introducing-the-nio-socketserver-implementation/
Note as the size of your data gets bigger, the throughput decreases.
Tom
> --
> You received this message because you are subscribed to the Google Groups "project-voldemort" group.
> To post to this group, send email to project-...@googlegroups.com.
> To unsubscribe from this group, send email to project-voldem...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/project-voldemort?hl=en.
>
>
--
Tom Chen
Software Architect
GOGII, Inc
t...@gogii.net
650-468-6318