With 1.4.0 almost behind us, the server team will shift to modeling some of our 
more important domains. Connections and activities are almost finished, we just 
need to make sure that we have made the storage layer pluggable. The next two 
domains we want to work on are People and Pooled Content. I'm tentatively 
including Worlds in this list as well, because we need to get deletes working 
properly, and if it starts to seem like changing the storage technology is the 
most straightforward way to do that, we'll try to implement it that way.

These implementations give us a test platform for both a relational database 
and Cassandra. I think the advocacy for MongoDB is almost totally dried up, so 
our decision is made that much simpler. Our engagement with OmniTI has begun, 
which means they'll be able to give us advice as we work.

By the 1.5.0 release, our team will have explored these options enough to make 
a recommendation. We'll have hard numbers from load tests. We'll have advice 
from OmniTI. I think the goal at that point will be to achieve a consensus with 
all the institutions who are currently deploying OAE. I certainly hope we don't 
put it to a vote.

I've learned a lot about our options over the last couple of months, and I have 
come to believe a few things:

  1. that we can succeed or fail regardless of the outcome of this particular 
decision
  2. that no matter what, we will be deploying a combination of complex 
technologies
  3. that the combination may ultimately include both a relational db and 
Apache Cassandra
  4. that we will experience pain one way or another, just different pain

Zach
On Jul 20, 2012, at 3:50 PM, Lance Speelmon wrote:

> With such an important decision ahead of us, could someone please articulate 
> how the decision will be made?  And in what time frame?
> 
> Thanks, L

_______________________________________________
oae-dev mailing list
[email protected]
http://collab.sakaiproject.org/mailman/listinfo/oae-dev

Reply via email to