zipper-s3

Introduction

UPP Golang App to download files from S3, zip them and upload the newly created zip file back to S3

Installation

Download the source code, dependencies and test dependencies:

    go get -u github.com/Financial-Times/zipper-s3
    cd $GOPATH/src/github.com/Financial-Times/zipper-s3
    go build .

Running locally

1. Run the tests and install the binary:

    go test ./... -v -race
    go install
   

2. Run the binary (using the help flag to see the available optional arguments): Environment variables:

   • IS_ENABLED flag which if it is true, the app will run the zip creation process, otherwise will stop immediately after start.
   • MAX_NO_OF_GOROUTINES the maximum number of goroutines which is used to zip files
   • YEAR_TO_START the app will create yearly zips starting from provided year. Defaults to 1995, when the first FT article has been published.
   • BUCKET_NAME bucket name of content
   • BUCKET_REGION bucket-name's region
   • S3_DOMAIN S3 domain of content
   • S3_CONTENT_FOLDER name of the folder that json files with the content are stored in
   • S3_CONCEPT_FOLDER name of the folder that json files with the concept are stored in
   • LOG_DEBUG flag which if it is set to true, the app will also output debug logs

   AWS related envvars.

   • AWS_ACCESS_KEY_ID S3 access key
   • AWS_SECRET_ACCESS_KEY S3 secret key
   • AWS_REGION S3 region

Running in Kubernetes

When the app is running in kubernetes, AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY envvars are not being used, instead AWS_ROLE_ARN and AWS_WEB_IDENTITY_TOKEN_FILE are used. The aws-sdk-go uses whichever envvars are present behind the scenes(in our code base there isn't logic for this).

Testing

This service is executed as a daily Kubernetes cronjob in the UPP Publishing clusters. It takes around 2 hours to complete its execution. If you don't want to wait until the next cron execution time to check the logs and the files in S3, you can manually trigger an execution using the following command (wait 2 hours to see if all archives are uploaded successfully):

kubectl create job --from=cronjob/zipper-s3 zipper-s3-manual

Build and deployment

• Built by Docker Hub on merge to master: coco/zipper-s3
• CI provided by CircleCI: zipper-s3