Public | Automated Build

Last pushed: 4 days ago
Short Description
This application reads concepts from s3 and posts them to kafka
Full Description

Aggregate Concept Transformer (aggregate-concept-transformer)

A service which gets notified via SQS of updates to source concepts in an Amazon S3 bucket. It then returns all UUIDs with concordance to said concept, requests each in turn from S3, builds the concorded JSON model and sends the updated concept JSON to both Neo4j and Elasticsearch. Finally it sends a notification of all updated concepts to a kinesis stream and removes the SNS message from the queue


For the first time:

go get

or update:

go get -u


Usage: aggregate-concept-service [OPTIONS]

Aggregating and concording concepts in UPP.

  --app-system-code="aggregate-concept-transformer"       System Code of the application ($APP_SYSTEM_CODE)
  --app-name="Aggregate Concept Transformer"              Application name ($APP_NAME)
  --port="8080"                                           Port to listen on ($APP_PORT)
  --bucketRegion="eu-west-1"                              AWS Region in which the S3 bucket is located ($BUCKET_REGION)
  --sqsRegion=""                                          AWS Region in which the SQS queue is located ($SQS_REGION)
  --bucketName=""                                         Bucket to read concepts from. ($BUCKET_NAME)
  --queueUrl=""                                           Url of AWS sqs queue to listen to ($QUEUE_URL)
  --messagesToProcess=10                                  Maximum number or messages to concurrently read off of queue and process ($MAX_MESSAGES)
  --visibilityTimeout=30                                  Duration(seconds) that messages will be ignored by subsequent requests after initial response ($VISIBILITY_TIMEOUT)
  --waitTime=20                                           Duration(seconds) to wait on queue for messages until returning. Will be shorter if messages arrive ($WAIT_TIME)
  --neo4jWriterAddress="http://localhost:8080/"           Address for the Neo4J Concept Writer ($NEO_WRITER_ADDRESS)
  --concordancesReaderAddress="http://localhost:8080/"    Address for the Neo4J Concept Writer ($CONCORDANCES_RW_ADDRESS)
  --elasticsearchWriterAddress="http://localhost:8080/"   Address for the Elasticsearch Concept Writer ($ES_WRITER_ADDRESS)
  --crossAccountRoleARN                                   ARN for cross account role ($CROSS_ACCOUNT_ARN)
  --kinesisStreamName=""                                  AWS Kinesis stream name ($KINESIS_STREAM_NAME)
  --kinesisRegion="eu-west-1"                             AWS region the kinesis stream is located ($KINESIS_REGION)
  --requestLoggingOn=true                                 Whether to log http requests or not ($REQUEST_LOGGING_ON)
  --logLevel="info"                                       App log level ($LOG_LEVEL)

The app assumes that you have correctly set up your AWS credentials by either using the ~/.aws/credentials file:

aws_access_key_id = AKID1234567890
aws_ secret_access_key = MY-SECRET-KEY

or the default AWS environment variables otherwise requests will return 401 Unauthorised


Build and deployment


This service aggregates a number of source concepts into a single canonical view. At present, the logic is as follows:

  • All concorded/secondary concepts are merged together without any ordering. These will always be from TME at the moment.
  • The primary concept is then merged, overwriting the fields from the secondary concepts. This is a Smartlogic concept.
  • Aliases are the exception - they are merged between all concepts and de-duplicated.


See swagger.yml

Admin Endpoints

  • Health checks: http://localhost:8080/__health
  • Good to go: http://localhost:8080/__gtg
  • Build info: http://localhost:8080/__build-info


Docker Pull Command