Week 6 Report and Some Questions

[Sheriffo Ceesay <sn...@gmail.com>]

Week six report now available at [1]

I have added CouchDB to the benchmark module. Adding any of the Gora
implemented data store would require auto-creating the mapping and Avro
files. It also requires setting up the datastore locally for testing. I
think for this work and given the timeline, it would be much better to
focus only a few databases and set up a standard benchmarking process. This
process can be used to benchmark other data stores in future.

I will need some suggestions on the way forward for benchmarking these data
stores. With limited resources, these are my plans:

   1. Use a single node on Google Cloud (I got $500 Google credit, thanks
   to a tip from Kevin)
   2. Setup MongoDB, HBase and CouchDB
   3. Setup gora-benchmark to connect to the Google Cloud Instance
   4. Benchmark these three datastores base on the workloads in YCSB
   5. Standardise the process
   6. Add more datastores if time permits. (Would be good to know which
   ones are a priority.)

Next Week: I hope to complete setting up an environment on Google Cloud and
probably have some preliminary numbers to present.

Any suggestions are highly welcomed.

Thank you.

[1]
https://cwiki.apache.org/confluence/display/GORA/%5BGORA-532%5D+Apache+Gora+Benchmark+Module+Weekly+Report

**Sheriffo Ceesay**