Okay, this is leaning more towards getting Brisk into our environment and making sure we can get it all working.
We plan on deploying to production Cassandra 0.8.5/6, however, Brisk only works on 0.8.1 (in the 0.8.x release)
Can we have a Brisk node operating as part of the ring to still do our data warehousing somewhat in a non-accessed state?
For our configuration, we have a ring of 2 DC's 3 nodes per DC. Our applications will be configured using Hector to connect to all 3 nodes in the local DC
Could we have:
* 3 Cassandra 0.8.5 Nodes
* 1 Brisk node
in one DC with the applications still connecting only to the 0.8.5 nodes? The other DC will have only the 3 Cassandra 0.8.5 nodes.
Are there any risks to setting up this way?
My understanding is that DataStax will be contributing the Hive Cassandra Driver to the Apache project, however, my concern is that with Brisk we had very little config involved in getting started. I don't want to have to install Hadoop and any dependencies separately when it can come prepackaged and configured out of the box - yes I know this is the goal with DataStax Enterprise.