Bucket Connection Configuration

This page explains how to connect the Bucket Storage Support services to the two different kinds of supported buckets. The guide is equivalent for all the services belonging to the Bucket Storage Support feature.

Connect with Google Bucket

In order to connect to a Google bucket, after setting the environment variable BUCKET_TYPE to google, a secret will have to be created, containing the credentials of the service account necessary for connecting to the Google bucket. Once done, the environment variable GOOGLE_APPLICATION_CREDENTIALS has to be set with the same value of the mount path defined in the secret. The document must be a json file, with the following structure

{
  "type": "<type>",
  "project_id": "<projectId>",
  "private_key_id": "<privateKeyId>",
  "private_key": "<privateKey>",
  "client_email": "<clientEmail>",
  "client_id": "<clientId>",
  "auth_uri": "<authUri>",
  "token_uri": "<tokenUri>",
  "auth_provider_x509_cert_url": "<authProviderCertUri>",
  "client_x509_cert_url": "<clientCertUrl>"
}

Connect with S3 Bucket

In order to connect to an S3 bucket, after setting the environment variable `BUCKET_TYPE` to `s3`, an additional configuration has to be written. Inside the microservice, in the ConfigMaps section, a new configMap has to be created, with the mount path `/app/config`. Inside the new config map, a new file has to be added, named `application.yaml`. The file will contain the credentials for interacting with the bucket, with the following structure:

bss:
  bucket:
    s3-key-id: {{S3_KEY_ID}}
    s3-key: {{S3_KEY}}
    region: {{S3_REGION}}
    endpoint: {{S3_ENDPOINT}}

caution

When setting the bucket endpoint, it is important to specify the URI scheme (e.g. https://) before the address. Otherwise the service may fail to start with the following error:

FD_ARC_E0003: provided S3 Cloud Storage configuration is invalid: The URI scheme of endpointOverride must not be null.