I am working on a project upstart that is using the YaCy ([url removed, login to view]) peer-to-peer search engine which is written in Java.
This search engine webcrawls various URL's and does a local cache of the indexes so that when a user query comes in, it check the cache first and then goes out to the online peers to see what information they contain relative to the query.
I am looking for a Java developer, preferable with YaCy development experience to make some additional modifications to the codebase such that an option can be configured in the online administrative section such that YaCy will not only read/write from the local cache, but also read/write to a Riak KV Cluster of nodes.
As the YaCy engine already uses a DHT (Distributed Hash Table) approach, and a Riak KV cluster is also a DHT, I want to combine the 2 components YaCy & Riak Cluster such that the "local" cluster can new be a distributed cluster which should speed up results when there are many YaCy nodes reading and writing their local indexes out to a Riak cluster that all nodes can see as a clustered index.
I would estimate that there should not require a lot of Java coding as the main patchwork would probably be in adding the Java Riak connector classes.
If this first set of modifications as an initial job are successful then there may be additional modifications for another job or two to follow for the reasonably priced and adequately skilled Java programmer.