Running on the Accumulo 1.5 line will be easier to manage as it
eliminates an extra Accumulo process (that was necessitated due to
limitations in pre-Hadoop2).
In short, if you're running Hadoop2, 1.5.1 is a good idea :)
As for serializing objects, I haven't done too much in that realm. There
are lots of different objects of serialization frameworks (avro,
protobuf, etc), but I'm not familiar with too many ORM solutions for
Accumulo outside of Gora.
On 3/27/14, 10:09 AM, Geoffry Roberts wrote:
I'm just getting started with Accumulo and have two questions:
1. I have hadoop 2.3.0 and zookeeper 3.4.6 running. Should I go with
Accumulo 1.5.1 or stick with 1.4.4?
2. Is the use of Gora/Avro advised? Gora seems to require quite a bit
of configuration.
I need to persist objects based on EFM (Eclipse Modeling Framework).
I'll be generating the EMF classes and I suppose I could generate the
JSON and XML required by Gora/Avro but I don't want to take on too much.
Thanks All.
--
There are ways and there are ways,
Geoffry Roberts
