AgileLogging

AgileLogging is a practical streaming data infrastructure case.

About

This repo is a stack to have running :

1. Fluentd : An open source data collector for unified logging layer. I used a connector (@type kafka2) to send logs in Kafka.
2. Kafka : Distributed streaming platform. Fluentd produces data in Kafka.
3. Zookeeper : Kafka run with Zookeeper. Zookeeper is a centralized service for maintaining configuration information, naming, providing distributed synchronization, and providing group services.
4. Kafka Connect : an component of Kafka, is a framework for connecting Kafka with external systems such as databases, key-value stores, search indexes, and file systems. It is used here to easily connect to kafka and send data in Elasticsearch. I used the Kafka Connect ElasticSearch
5. kafka-connect-ui : Used to have access to Kafka Connect UI. I used it to set up connector without go on command line.
6. Elasticsearch : It is an distributed platform and a RESTful search and analytics engine. Used here to index and easily search data.
7. Kibana : UI to connect and view data or logs stored in Elasticsearch.
8. python-app-test : A Python application to test the logs streaming.

How do you get set up?

• Dependencies

  • Docker

• Deployment instructions

  • Open your terminal

  • Go in the stack folder: $ cd agile-logging

  • Run the docker-compose and wait until services started: $ docker-compose -f docker-compose-dist-logging.yaml up --build --d

  • After all services created, check services running: $ docker ps You should see the following services unordered running: . python-app-test . fluentd-agile . zookeeper-agile-node1 . kafka-agile-node1 . kafka_connect . kafka-connect-ui . elasticsearch-agile-one . kibana-agile

  • Start a connector on kafka connector. It will be responsible to consume data from kafka and sync it in elasticsearch You have 2 possibilities to start connector:

    1. Using the command : Enter in the docker $ docker exec -it kafka_connect bash Inside the docker, add the the connector by running the following command $ curl -X POST -H "Content-Type: application/json" --data @/opt/connector_conf/connector_elasticsearch.json http://localhost:8083/connectors

       You will see an output like the following. It means that your connector has been added  :
       `{
           {
               "name":"elasticsearch-sink",
               "config":{"connector.class":"io.confluent.connect.elasticsearch.ElasticsearchSinkConnector",
               "tasks.max":"1",
               "topics":"log-messages",
               "key.ignore":"true",
               "schema.ignore":"true",
               "connection.url":"http://elasticsearch-agile-one:9200",
               "type.name":"kafka-connect",
               "name":"elasticsearch-sink"
           },
           "tasks":[],
           "type":null
       }`
       

    2. Using the Kafka-connect UI Open your browser and go on http://localhost:8003 Click on the button "New" to add new connector Choose ElasticSearch Connector and fill json like in the file located here connector_conf/connector_elasticsearch.json { "name":"elasticsearch-sink", "config":{ "connector.class":"io.confluent.connect.elasticsearch.ElasticsearchSinkConnector", "tasks.max":"1", "topics":"log-messages", "key.ignore":"true", "schema.ignore": "true", "connection.url": "http://elasticsearch-agile-one:9200", "type.name": "kafka-connect" }

       Click on the button "create" when you finish.

  • Open your browser and go on http://localhost:8003. You should see the count of SINK CONNECTORS at 1. It means connectors has been added.

  • Let's check now with Kibana if the connector has created our indice in Elasticsearch. . Open your browser and go on http://localhost:5601/ . Find "Management" menu and click on it . Under Elasticsearch section, click on "Index Management" and you should see your indice created with the name "log-messages" . To visualize data with Kibana, we should create an index pattern. So under Kibana section, click on "Index Patterns" and click on "Create index pattern". Fill the index pattern name "log-messages*" and click on "next step" and finish . Indexpattern created now, go in the left menu and click on "Discover" link to already view some data after selecting the index "log-messages".

  • Open your browser and go on http://localhost:7500/v1/organization and you will see the application output:

    {"result": "Wow! !:)"}

    Go on Kibana and refresh data to see your logs data there. Hit again http://localhost:7500/v1/organization and go to view changes in Kibana

Possible performance optimizations for the Code.

- This is a proof of concept. To scale the system we can add more Zookeeper and Kafka nodes.

Who do you talk to?

• Repo owner or admin Full name: Emmanuel HODONOU Mail: [email protected] Github: https://github.com/emmano3h/