Consumer tw tot Deploy

STREAM DEPLOY:

To deploy consumer-tw-tot-stream

• with start-all.sh script: just check that on sda/scripts/module tw-tot-stream key is uncommented.

• Using consumer-tw-tot/start-tw-tot-stream.sh:

  Provide the following options to the script or set the corrisponding environment variables:

SCRIPT ARGUMENT	ENV VARIABLE	DESCRIPTION
sda-home	SDA_HOME	The path of social-data-aggregator folder
with-master	MASTER	master name (eg local,spark://xxx.xxx)
spark-home	SPARK_HOME	The path to spark folder

BATCH DEPLOY:

In order to run consumer-tw-tot batch analytics start the shell script under the folder sda/scripts/consumer-tw-tot/start-tw-tot-batch.sh after providing the following settings:

• on sda/scripts/consumer-tw-tot/consumer-tw-tot-confs.cfg set the following properties:

KEY NAME	DESCRIPTION
MASTER	Spark master address (spark://MASTER_IP:MASTER_PORT) or local
SPARK_HOME	absolute path to spark home
INPUT_DATA_PATH	Default input data path (where raw data, on which analysis have to be done, are stored)

• start-tw-tot-batch.sh script:

./start-tw-tot-batch.sh --help

OPTIONS:

OPTION NAME	DESCRIPTION
from	time from which you want to start the analysis (ISO8601 format) e.g 2015-02-18T17:00:00+01
to	time to which you want to stop the analysis (ISO8601 format) e.g 2015-02-28T17:00:00+01
roundMode	define the round mode on the creation time. Possible options are: min:round on minute hour:round on hour day:round on day
granMin	valid only if round mode is min. Granularity,if you want to group in minute intervals (e.g gran=5 will group by 5 minutes -> the number of tweets in 5 minutes)