search

baskard / thrudb

Automatically exported from code.google.com/p/thrudb
0 stars 0 forks source link

Which implementation to use? #16

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago 
Which of the two implementations (the one hosted here and the other at
github) should be considered more or less stable? What are the differences
between them (apart from language) and how should users choose one now or
for future use? What are the prospects?

Original issue reported on code.google.com by neithere on 26 Jun 2009 at 7:03

GoogleCodeExporter commented 8 years ago 
I'm definitely focused on the java implementation now. Thou the code isn't as 
far along as the current thrudb.

The main reasons I've moved to java is for ease of deployment and adoption of 
java lucene over clucene.

The main differences:

  Thrudoc:  java implementation is based on a tokyocabinet or bdb backend.  This allows support for lists and 
queues along with get/set operations.  No plans for a S3 backend other than for 
backup/replication purposes.

  Thrudex:  java implementation uses java lucene which I've found is basically as fast as clucene. plus it 
crashes a lot less :)

  Thruqueue:  This goes away since the new version of thrudoc supports queues.

  Throxy:  in Java impl this is built into thrudoc and thrudex as an optional backend. based on zookeeper.

I think in the end the functionality for clients will be mainly the same but it 
should be much simpler and 
robust.

Original comment by jake%3.r...@gtempaccount.com on 28 Jun 2009 at 1:32