DESCRIPTION

WIP sample script for creating n buckets in a cluster

this script will create n buckets in the cluster and then starts
loading data into these buckets. for each bucket we kick off a new
thread that starts loading data into that bucket

Change-Id: Ib658fe8ccf68b560b56b3a3d8e5f025849e4fb31


COMMENTS

author: Zackery Barr
date: 2011-07-08 16:29:11.134000000

Patch Set 1: I would prefer that you didn't submit this

still testing

-------------------------------------
author: Nora Conway
date: 2011-07-08 16:38:11.157000000

Patch Set 1: (3 inline comments)

cool!  some feedback on parametrization.

Line:28, scripts/bucketloader.py -> Besides calculated from available_ram, can we also have this ini-file or cmd-line parametrized?

Line:40, scripts/bucketloader.py -> Can this be more parametrized?

Line:53, scripts/bucketloader.py -> Can we have bucket loading be optionally turned off?  That is, a way where this program just creates a delta of buckets?

-------------------------------------
author: Zackery Barr
date: 2011-07-08 16:48:18.424000000

Patch Set 1: (3 inline comments)



Line:28, scripts/bucketloader.py -> parametrize  from the command line bucket_ram=xyz

Line:40, scripts/bucketloader.py -> as of now 
1- we can specify whether to do a mixture of set/get
2- use moxi or do not use moxi
3- what value size we want to use 
the default is 50% use 500 bytes and 50% 1K

we can paramterize above options through command line.
let me know if this makes sense ?
moxi=true(or false),set=true,get=true,size=500,1024,2048


in few weeks we will replace this load_bucket with another data generator called load_runner which is more customized but is WIP now.

Line:53, scripts/bucketloader.py -> by default we can turn off the loading part and if user wants we can pass in load_true ? 

or I can write another script that simply loads data

-------------------------------------
