{
  "schema_version": 2,
  "id": "integrate/redis-data-integration/data-pipelines/prepare-dbs/spanner",
  "title": "Prepare Spanner for RDI",
  "url": "https://redis.io/docs/latest/integrate/redis-data-integration/2.0.0/data-pipelines/prepare-dbs/spanner/",
  "summary": "Prepare Google Cloud Spanner databases to work with RDI",
  "content": "\nGoogle Cloud Spanner requires specific configuration to enable change data capture (CDC) with RDI.\nRDI operates in two phases with Spanner: snapshot (initial sync) and streaming. During the snapshot\nphase, RDI uses the JDBC driver to connect directly to Spanner and read the current state of the\ndatabase. In the streaming phase, RDI uses [Spanner's Change Streams](https://cloud.google.com/spanner/docs/change-streams) to capture changes related to\nthe monitored schemas and tables.\n\n\u003e [!NOTE]\n\u003e Spanner is only supported with RDI deployed on Kubernetes/Helm. RDI VM mode does not support Spanner as a source database.\n\nThe following checklist summarizes the steps to prepare a Spanner\ndatabase for RDI, with links to the sections that explain the steps in\nfull detail. You may find it helpful to track your progress with the\nchecklist as you complete each step.\n\n```checklist {id=\"spannerlist\"}\n- [ ] [Prepare for snapshot](#1-prepare-for-snapshot)\n- [ ] [Prepare for streaming](#2-prepare-for-streaming)\n- [ ] [Create a service account](#3-create-a-service-account)\n- [ ] [Set up secrets (optional)](#4-set-up-secrets-optional)\n- [ ] [Configure RDI for Spanner](#5-configure-rdi-for-spanner)\n- [ ] [Additional Kubernetes configuration](#6-additional-kubernetes-configuration)\n```\n\n## 1. Prepare for snapshot\n\nDuring the snapshot phase, RDI executes multiple transactions to capture data at an exact point \nin time that remains consistent across all queries. This is achieved using a Spanner feature called \n[Timestamp bounds with exact staleness](https://cloud.google.com/spanner/docs/timestamp-bounds#exact_staleness). \n\nThis feature relies on the \n[version_retention_period](https://cloud.google.com/spanner/docs/reference/rest/v1/projects.instances.databases#Database.FIELDS.version_retention_period), \nwhich is set to one hour by default. Depending on the database tier, the volume of data to be \ningested into RDI, and the load on the database, this setting may need to be increased. You can \nupdate it using [this method](https://cloud.google.com/spanner/docs/use-pitr#set-period).\n\n## 2. Prepare for streaming\n\nTo enable streaming, you must create a change stream in Spanner at the database level. Use the \noption `value_capture_type = 'NEW_ROW_AND_OLD_VALUES'` to capture both the previous and updated \nrow values.\n\nBe sure to specify only the tables you want to ingest from and, optionally, the specific columns \nyou're interested in. Here's an example using Google SQL syntax:\n\n```sql\nCREATE CHANGE STREAM change_stream_table1_and_table2\n  FOR table1, table2\n  OPTIONS (\n    value_capture_type = 'NEW_ROW_AND_OLD_VALUES'\n  );\n```\n\nRefer to the [official documentation](https://cloud.google.com/spanner/docs/change-streams/manage#googlesql) \nfor more details, including additional configuration options and dialect-specific syntax.\n\n## 3. Create a service account\n\nTo allow RDI to access the Spanner instance, you'll need to create a service account with the\nappropriate permissions. By default, RDI uses Google Cloud Workload Identity authentication. In this case RDI will assume the [service account is assigned to the GKE cluster](https://cloud.google.com/kubernetes-engine/docs/how-to/workload-identity#enable_on_clusters_and_node_pools). Alternatively, you can provide the\nservice account credentials as a Kubernetes secret (see step 4 for details).\n\n```checklist {id=\"spanner-service-account\" nointeractive=\"true\" }\n- [ ] [Create the service account](#create-the-service-account)\n- [ ] [Grant required roles](#grant-required-roles)\n- [ ] [Download the service account key](#download-the-service-account-key)\n```\n\n1. \u003ca id=\"create-the-service-account\"\u003e\u003c/a\u003e\n   Create the service account\n\n    ```bash\n    gcloud iam service-accounts create spanner-reader-account \\\n        --display-name=\"Spanner Reader Service Account\" \\\n        --description=\"Service account for reading from Spanner databases\" \\\n        --project=YOUR_PROJECT_ID\n    ```\n\n1. \u003ca id=\"grant-required-roles\"\u003e\u003c/a\u003e\n   Grant required roles:\n\n    **Database Reader** (read access to Spanner data):\n\n    ```bash\n    gcloud projects add-iam-policy-binding YOUR_PROJECT_ID \\\n        --member=\"serviceAccount:spanner-reader-account@YOUR_PROJECT_ID.iam.gserviceaccount.com\" \\\n        --role=\"roles/spanner.databaseReader\"\n    ```\n\n    **Database User** (query execution and metadata access):\n\n    ```bash\n    gcloud projects add-iam-policy-binding YOUR_PROJECT_ID \\\n        --member=\"serviceAccount:spanner-reader-account@YOUR_PROJECT_ID.iam.gserviceaccount.com\" \\\n        --role=\"roles/spanner.databaseUser\"\n    ```\n\n    **Viewer** (viewing instance and database configuration):\n\n    ```bash\n    gcloud projects add-iam-policy-binding YOUR_PROJECT_ID \\\n        --member=\"serviceAccount:spanner-reader-account@YOUR_PROJECT_ID.iam.gserviceaccount.com\" \\\n        --role=\"roles/spanner.viewer\"\n    ```\n\n1. \u003ca id=\"download-the-service-account-key\"\u003e\u003c/a\u003e\n   Download the service account key:\n\n    Save the credentials locally so they can be used later by RDI:\n\n    ```bash\n    gcloud iam service-accounts keys create ~/spanner-reader-account.json \\\n        --iam-account=spanner-reader-account@YOUR_PROJECT_ID.iam.gserviceaccount.com \\\n        --project=YOUR_PROJECT_ID\n    ```\n\n### Authentication methods\n\nRDI supports two authentication methods for accessing Spanner:\n\n1. **Workload Identity (default)**: The service account is assigned to the GKE cluster, and RDI\n   automatically uses the cluster's identity to authenticate. This is the recommended approach\n   as it's more secure and doesn't require managing credential files.\n\n2. **Service account credentials file**: You provide the service account key file as a Kubernetes\n   secret. This method requires setting `use_credentials_file: true` in your RDI configuration.\n\n## 4. Set up secrets (optional)\n\nBefore deploying the RDI pipeline, you need to configure the necessary secrets for the target\ndatabase. Set them with\n[`redis-di set-secret`](https://redis.io/docs/latest/integrate/redis-data-integration/reference/cli/redis-di-set-secret),\nas described in\n[Set secrets](https://redis.io/docs/latest/integrate/redis-data-integration/data-pipelines/deploy#set-secrets).\nA Spanner source needs no credentials secret of its own, because it authenticates through\nWorkload Identity or a service account credentials file.\n\n**Optional**: If you prefer to use a service account credentials file instead of Workload Identity\nauthentication, you'll need to create a Spanner-specific secret named\n`\u003csource\u003e-db-credentials`, where `\u003csource\u003e` is the name of the source in `config.yaml`.\nThe example below assumes a source named `spanner`, so its corresponding secret is\n`spanner-db-credentials`. See\n[Multiple sources in one pipeline](https://redis.io/docs/latest/integrate/redis-data-integration/data-pipelines/multiple-sources)\nfor the source naming rules.\n\nThis secret should contain the service account key file generated during the Spanner setup phase.\nUse the command below to create it:\n\n```bash\nkubectl create secret generic spanner-db-credentials --namespace=rdi \\\n--from-file=gcp-service-account.json=~/spanner-reader-account.json \\\n--save-config --dry-run=client -o yaml | kubectl apply -f -\n```\n\nBe sure to adjust the file path (`~/spanner-reader-account.json`) if your service account key is\nstored elsewhere.\n\n\u003e [!NOTE]\n\u003e `redis-di set-secret` has no key for this secret, so you must\n\u003e create it with `kubectl`. A secret created that way must also be labeled so that the RDI operator\n\u003e discovers it as a pipeline secret. See\n\u003e [Set secrets for K8s/Helm deployment using Kubectl command](https://redis.io/docs/latest/integrate/redis-data-integration/data-pipelines/deploy#set-secrets-for-k8shelm-deployment-using-kubectl-command) for more information.\n\n\u003e [!NOTE]\n\u003e If you create the `spanner-db-credentials` secret, you must also set `use_credentials_file: true`\n\u003e in your RDI configuration to use the credentials file instead of Workload Identity authentication.\n\n## 5. Configure RDI for Spanner\n\nWhen configuring your RDI pipeline for Spanner, use the following example configuration in your \n`config.yaml` file:\n\n```yaml\nsources:\n  spanner:\n    type: flink\n    connection:\n      type: spanner\n      project_id: your-project-id\n      instance_id: your-spanner-instance\n      database_id: your-spanner-database\n      # use_credentials_file: false  # Default: uses Workload Identity. Set to true to use service account credentials file instead\n      change_streams:\n        change_stream_all:\n          {}\n          # retention_hours: 24\n    # schemas:\n    #  - DEFAULT\n    # tables:\n    #   products: {}\n    #   orders: {}\n    #   order_items: {}\n    # logging:\n    #   level: debug\n    # advanced:\n    #   source:\n    #     spanner.change.stream.retention.hours: 24\n    #     spanner.fetch.timeout.milliseconds: 20000\n    #     spanner.dialect: POSTGRESQL\n    #   flink:\n    #     jobmanager.rpc.port: 7123\n    #     jobmanager.memory.process.size: 1024m\n    #     taskmanager.numberOfTaskSlots: 3\n    #     taskmanager.rpc.port: 7122\n    #     taskmanager.memory.process.size: 2g\n    #     blob.server.port: 7124\n    #     rest.port: 8082\n    #     parallelism.default: 4\n    #     restart-strategy.type: fixed-delay\n    #     restart-strategy.fixed-delay.attempts: 3\ntargets:\n  target:\n    connection:\n      type: redis\n      host: ${HOST_IP}\n      port: 12000\n      user: ${TARGET_DB_USERNAME}\n      password: ${TARGET_DB_PASSWORD}\nprocessors:\n  target_data_type: hash\n```\n\nMake sure to replace the relevant connection details with your own for both the Spanner and target \nRedis databases.\n\n## 6. Additional Kubernetes configuration\n\nIn your `rdi-values.yaml` file for Kubernetes deployment, make sure to configure the `dataPlane` \nsection like this:\n\n```yaml\noperator:\n  dataPlane:\n    flinkCollector:\n      enabled: true\n      jobManager:\n        ingress:\n          enabled: true\n          className: traefik # Replace with your ingress controller\n          hosts:\n            - hostname # Replace with your desired ingress hostname\n```\n\n## 7. Configuration is complete\n\nOnce you have followed the steps above, your Google Spanner database is ready for RDI to use.\n",
  "tags": ["docs","integrate","rs","rdi"],
  "last_updated": "2026-10-30T16:31:14Z"
}