Eaton-Feeder

README

Eaton-Feeder

Application(s) for reading information from external feeds, initially the Indeed API for job descriptions

Installation

You'll need to follow the instructions here for installing go: https://golang.org/doc/install

You will also need kafka servers running for the program to interact with. You can get kafka quickly started by following the instructions here: http://kafka.apache.org/documentation.html#quickstart

Git is also required to pull down the code.

Once you have everything installed (with your GOPATH set), just run:

$ go get github.com/ECLabs/Eaton-Feeder

Here's an example script to run the program as a producer:

#!/bin/sh

# Override the following to change which topic to produce messages to,
# which servers to send messages to, and which publisher id to use.
# The following are the default values:
#   KAFKA_TOPIC = eaton-feeder
#       This is the location of where each job result from the IndeedAPI is sent to in XML format.
#   KAFKA_SERVERS = 127.0.0.1:9092
#       This is the comma delmited listing of all kafka servers to send messages to.
#   INDEED_PUBLISHER_ID = 
#       This defaults to an empty string and is REQUIRED for the application to work properly.
#   KAFKA_LOGGER_TOPIC = logs
#       This is the topic that all logs are sent to to be consumed by the application in the http directory.

KAFKA_TOPIC=myTopic
KAFKA_SERVERS=127.0.0.1:9092
INDEED_PUBLISHER_ID=123456789
KAFKA_LOGGER_TOPIC=logs-producer

export KAFKA_TOPIC
export KAFKA_SERVERS
export INDEED_PUBLISHER_ID
# a -1 interval disables the builtin polling capability and once the 
# program has finished pulling all results from the indeed API it
# will terminate.
$GOPATH/bin/Eaton-Feeder -produce=true -interval=-1

The program operates sightly differently when it's a consumer. It will stay running indefinitely since the client needs to always be ready for new messages. To ensure that it is always running, you can install monit and have a script similar to the following to keep track of the process id:

#!/bin/sh
# Override the following to change which topic to produce messages to,
# which servers to send messages to, and which publisher id to use.
# The following are the default values:
#   KAFKA_TOPIC = eaton-feeder
#       This is the location of where each job result from the IndeedAPI is consumed from and should match what the producer is using.
#   KAFKA_SERVERS = 127.0.0.1:9092
#       This is the comma delmited listing of all kafka servers to consumer messages from.
#   KAFKA_LOGGER_TOPIC = logs
#       This is the topic that all logs are sent to to be consumed by the application in the http directory.
#   AWS_SECRET_KEY_ID =
#       This defaults to an empty string and is REQUIRED for the application to work properly.
#   AWS_SECRET_ACCESS_KEY =
#       This defaults to an empty string and is REQUIRED for the application to work properly.

AWS_ACCESS_KEY_ID=my_access_key
AWS_SECRET_ACCESS_KEY=my_secret_key
KAFKA_TOPIC=myTopic
KAFKA_SERVERS=127.0.0.1:9092
KAFKA_LOGGER_TOPIC=logs-consumer

export AWS_SECRET_ACCESS_KEY
export AWS_ACCESS_KEY_ID
export KAFKA_SERVERS
export KAFKA_TOPIC

$GOPATH/bin/Eaton-Feeder -offset=newest -consume=true  & echo "$!" > eatonfeeder.pid