Database Access with AWS RDS and Aurora for PostgreSQL, MySQL and MariaDB

Teleport can provide secure access to Amazon RDS or Aurora via the Teleport Database Service. This allows for fine-grained access control through Teleport's RBAC.

In this guide, you will:

1. Configure your Amazon RDS or Aurora database with IAM authentication.
2. Add the database to your Teleport cluster.
3. Connect to the database via Teleport.

How it works

The Teleport Database Service uses IAM authentication to communicate with RDS. When a user connects to the database via Teleport, the Teleport Database Service obtains AWS credentials and authenticates to AWS as an IAM principal with permissions to access the database.

• Self-Hosted
• Cloud-Hosted

This guide shows how to register a single RDS with your Teleport cluster. For a more scalable approach, learn how to set up Database Auto-Discovery to automatically enroll all AWS databases in your infrastructure.

Prerequisites

• A running Teleport cluster version 17.0.2 or above. If you want to get started with Teleport, sign up for a free trial or set up a demo environment.

• The tctl admin tool and tsh client tool.

  Visit Installation for instructions on downloading tctl and tsh.

• AWS account with RDS and Aurora databases and permissions to create and attach IAM policies.

  IAM authentication

  Your RDS and Aurora databases must have password and IAM authentication enabled.

  If IAM authentication is not enabled on the target RDS and Aurora databases, the Database Service will attempt to enable IAM authentication by modifying them using respective APIs.

• A Linux host or Amazon Elastic Kubernetes Service cluster where you will run the Teleport Database Service, which proxies connections to your RDS databases.

• To check that you can connect to your Teleport cluster, sign in with tsh login, then verify that you can run tctl commands using your current credentials.

  For example:

  tsh login --proxy=teleport.example.com --user=email@example.com
  tctl status
  Cluster  teleport.example.com
  Version  17.0.2
  CA pin   sha256:abdc1245efgh5678abdc1245efgh5678abdc1245efgh5678abdc1245efgh5678

  If you can connect to the cluster and run the tctl status command, you can use your current credentials to run subsequent tctl commands from your workstation. If you host your own Teleport cluster, you can also run tctl commands on the computer that hosts the Teleport Auth Service for full permissions.

If you plan to run the Teleport Database Service on Kubernetes, you will need the following:

• The aws CLI in your PATH. Install it by following the AWS documentation.

• An IAM OIDC provider running in your Kubernetes cluster. See the AWS documentation for how to create an IAM OIDC provider.

  To check whether you have an IAM OIDC provider running in your cluster, run the following aws command, assigning eks-region to the region where your EKS cluster is running and cluster-name to the name of your Kubernetes cluster:

  aws --region=eks-region eks describe-cluster --name cluster-name --query "cluster.identity.oidc.issuer" --output text

  If you have an IAM OIDC provider associated with your cluster, this command will print its ID.

• The jq CLI tool, which we use to process JSON data in this guide.

Step 1/6. Create a Teleport user

tctl users add \  --roles=access \  --db-users="*" \  --db-names="*" \  alice

tctl users add \  --roles=access,requester \  --db-users="*" \  --db-names="*" \  alice

For more detailed information about database access controls and how to restrict access see RBAC documentation.

Step 2/6. Create a Database Service configuration

In this section, you will configure the Teleport Database Service. To do so, you will:

• Create a join token for the service to demonstrate trust with your Teleport cluster.
• Set up your package manager so you can install and run the Database Service.
• Generate a configuration for the Database Service.

Create a join token

Establish trust between the Teleport Database Service and your Teleport cluster by creating a join token.

Generate a join token by running the following command on your workstation:

tctl tokens add --type=db

The next step depends on how you plan to run the Teleport Database Service:

Prepare your environment

Next, get your environment ready to run the Teleport Database Service:

sudo teleport db configure create \   -o file \   --name=rds-example \   --proxy=example.teleport.sh:443  \   --protocol=protocol \   --uri=endpoint:port \   --labels=env=dev \   --token=/tmp/token

helm repo add teleport https://charts.releases.teleport.dev

helm repo update

Step 3/6. Create IAM policies for Teleport

The Teleport Database Service needs AWS IAM permissions to provide access to RDS instances and Aurora clusters.

Create an IAM role for Teleport

Grant the Database Service access to credentials that it can use to authenticate to AWS.

• If you are running the Database Service on an EC2 instance, you may use the EC2 Instance Metadata Service method
• If you are running the Database Service in Kubernetes, you can use IAM Roles for Service Accounts (IRSA)
• Otherwise, you must use environment variables

AWS_ACCESS_KEY_ID=00000000000000000000
AWS_SECRET_ACCESS_KEY=0000000000000000000000000000000000000000
AWS_DEFAULT_REGION=<YOUR_REGION>

Grant permissions

Attach the following AWS IAM permissions to the Database Service IAM role:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "RDSAutoEnableIAMAuth",
            "Effect": "Allow",
            "Action": [
                "rds:ModifyDBCluster",
                "rds:ModifyDBInstance"
            ],
            "Resource": "*"
        },
        {
            "Sid": "RDSConnect",
            "Effect": "Allow",
            "Action": "rds-db:connect",
            "Resource": "*"
        },
        {
            "Sid": "RDSFetchMetadata",
            "Effect": "Allow",
            "Action": [
                "rds:DescribeDBClusters",
                "rds:DescribeDBInstances"
            ],
            "Resource": "*"
        }
    ]
}

The Teleport Database Service uses rds:ModifyDBInstance and rds:ModifyDBCluster to automatically enable IAM authentication on RDS instances and Aurora clusters, respectively. You can omit the RDSAutoEnableIAMAuth permissions if IAM authentication is already enabled on your databases.

The rds-db:connect permission is required to connect to databases. You can reduce the scope of the permission to only allow specific databases, regions, or users. The resource ARN has the following format:

arn:aws:rds-db:{Region}:{AccountID}:dbuser:{ResourceID}/{UserName}

Refer to Creating and using an IAM policy for IAM database access for more information about the rds-db:connect permission grant syntax.

Databases discovered by the Teleport Discovery Service should be registered with complete metadata, so you can also omit the RDSFetchMetadata permissions if all of your AWS databases are being auto-discovered.

Step 4/6. Start the Database Service

Start the Teleport Database Service in your environment:

sudo systemctl enable teleport
sudo systemctl start teleport

sudo teleport install systemd -o /etc/systemd/system/teleport.service
sudo systemctl enable teleport
sudo systemctl start teleport

tctl tokens ls
Token                            Type Labels Expiry Time (UTC)-------------------------------- ---- ------ ----------------------------abcd123-insecure-do-not-use-this Db          14 Jun 23 21:21 UTC (20m15s)

authToken: token
proxyAddr: example.teleport.sh:443
roles: db
databases:
- name: example
  uri: "endpoint:port"
  protocol: protocol
  static_labels:
    env: dev
annotations:
  serviceAccount:
    eks.amazonaws.com/role-arn: arn:aws:iam::aws-account:role/teleport-rds-role

helm -n teleport-agent install teleport-kube-agent teleport/teleport-kube-agent \  --values values.yaml --create-namespace

kubectl -n teleport-agent get pods
NAME                    READY   STATUS    RESTARTS   AGEteleport-kube-agent-0   1/1     Running   0          32s

Step 5/6. Create a database IAM user

Database users must allow IAM authentication in order to be used with Database Access for RDS. See below how to enable it for the user alice on your database engine. In the next step, we will authenticate to the database as the alice user via the user's Teleport account.

CREATE USER alice;
GRANT rds_iam TO alice;

CREATE USER alice IDENTIFIED WITH AWSAuthenticationPlugin AS 'RDS';

GRANT ALL ON `%`.* TO 'alice'@'%';

See Creating a database account using IAM authentication for more information.

Step 6/6. Connect

Once the Database Service has started and joined the cluster, log in as the alice user you created earlier to see the registered databases:

tsh login --proxy=example.teleport.sh:443 --user=alice
tsh db ls
Name        Description Labels
----------- ----------- --------
rds-example             env=dev

Retrieve credentials for the database and connect to it as the alice user:

tsh db connect --db-user=alice --db-name=postgres rds-example

Log out of the database and remove credentials:

tsh db logout rds-example

Troubleshooting

Certificate error

If your tsh db connect error includes the following text, you likely have an RDS or DocumentDB database created before July 28, 2020, which presents an X.509 certificate that is incompatible with Teleport:

x509: certificate relies on legacy Common Name field, use SANs instead

AWS provides instructions to rotate your SSL/TLS certificate.

No credential providers error

If you see the error NoCredentialProviders: no valid providers in chain in Database Service logs then Teleport is not detecting the required credentials to connect via AWS IAM permissions. Check whether the credentials or security role has been applied in the machine running the Teleport Database Service.

When running on EKS, this error may occur if the Teleport Database Service cannot access IMDSv2 when the PUT requests hop limit on the worker node instance is set to 1. You can use the following commands to check the hop limit:

aws ec2 describe-instances --instance-ids <node-instance-id> | grep HttpPutResponseHopLimit
                        "HttpPutResponseHopLimit": 1,

See IMDSv2 support for EKS and EKS best practices for more details.

Timeout errors

The Teleport Database Service needs connectivity to your database endpoints. That may require enabling inbound traffic on the database from the Database Service on the same VPC or routing rules from another VPC. Using the nc program you can verify connections to databases:

nc -zv postgres-instance-1.sadas.us-east-1.rds.amazonaws.com 5432
Connection to postgres-instance-1.sadas.us-east-1.rds.amazonaws.com (172.31.24.172) 5432 port [tcp/postgresql] succeeded!

Not authorized to perform sts:AssumeRole

The Database Service assumes an IAM role in one of following situations:

• A Teleport user specifies an IAM role as the database user they wish to use when accessing AWS services that require IAM roles as database users. Databases that support using an IAM role as a database user include: DynamoDB, Keyspaces, Opensearch, Redshift, and Redshift Serverless.
• The assume_role_arn field is specified for the database resources or dynamic resource matchers.

You may encounter the following error if the trust relationship is not configured properly between the IAM roles:

AccessDenied: User: arn:aws:sts::111111111111:assumed-role/teleport-db-service-role/i-* is not authorized to perform: sts:AssumeRole on resource: arn:aws:iam::111111111111:role/db-user-role

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "AWS": "arn:aws:iam::aws-account-id:role/teleport-db-service-role"
      },
      "Action": "sts:AssumeRole"
    }
  ]
}

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "AWS": "arn:aws:iam::aws-account-id:root"
      },
      "Action": "sts:AssumeRole"
    }
  ]
}

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "AWS": "arn:aws:iam::external-aws-account-id:role/teleport-db-service-role"
      },
      "Action": "sts:AssumeRole",
      "Condition": {
        "StringEquals": {
          "sts:ExternalId": "example-external-id"
        }
      }
    }
  ]
}

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Action": "sts:AssumeRole",
            "Effect": "Allow",
            "Resource": "arn:aws:iam::aws-account-id:role/db-user-role"
        }
    ]
}

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Action": "sts:AssumeRole",
            "Effect": "Allow",
            "Resource": "*"
        }
    ]
}

You can test the trust relationship by running this AWS CLI command as teleport-db-service-role:

aws sts assume-role --role-arn arn:aws:iam::111111111111:role/db-user-role --role-session-name test-trust-relationship

Learn more on how to use trust policies with IAM roles.

Maximum policy size exceeded errors

Due to IAM and STS character limits, you may encounter one of the following errors in the Database Service logs when large numbers of databases are registered:

• LimitExceeded: Maximum policy size of 2048 bytes exceeded for user <iam-user>
• LimitExceeded: Maximum policy size of 10240 bytes exceeded for role <iam-role>

For reference, a user policy can maintain permissions for approximately 6 Redshift databases, or 20 RDS databases due to the IAM policy character limits. A role policy can maintain permissions for approximately 30 Redshift databases, or 100 RDS databases.

To get around this limit, try using one or a combination of the following methods:

discovery_service:
  discovery_group: "prod"
  enabled: "yes"
  aws:
    - types: ["rds"]
      regions: ["us-west-1", "us-west-2"]
      assume_role_arn: "arn:aws:iam::123456789012:role/example-role-rds-env-prod-discovery"
      tags:
        "env": "prod"

    - types: ["redshift", "redshift-serverless"]
      regions: ["us-west-2"]
      assume_role_arn: "arn:aws:iam::123456789012:role/example-role-redshift-env-dev"
      tags:
        "env": "dev"

db_service:
  enabled: "yes"
  resources:
    # Matches us-west-1 env=prod RDS databases from Discovery Service, and
    # overwrites assume_role_arn.
    - labels:
        "env": "prod"
        "region": "us-west-1"
      aws:
        assume_role_arn: "arn:aws:iam::123456789012:role/example-role-rds-env-prod-us-west-1-access"

    # Matches us-west-2 env=prod RDS databases from Discovery Service, and
    # overwrites assume_role_arn.
    - labels:
        "env": "prod"
        "region": "us-west-2"
      aws:
        assume_role_arn: "arn:aws:iam::123456789012:role/example-role-rds-env-prod-us-west-2-access"

    # Matches env=dev Redshift databases from Discovery Service and inherits
    # "arn:aws:iam::123456789012:role/example-role-redshift-env-dev"
    - labels:
        "env": "dev"

teleport db configure aws create-iam --types redshift,redshift-serverless --name teleport-redshift-access
teleport db configure aws print-iam --types redshift,redshift-serverless

db_service:
  enabled: "yes"
  aws:
    - types: ["rds"]
      regions: ["us-west-1", "us-west-2"]
      assume_role_arn: "arn:aws:iam::123456789012:role/example-role-rds-env-prod"
      tags:
        "env": "prod"

    - types: ["redshift", "redshift-serverless"]
      regions: ["us-west-2"]
      assume_role_arn: "arn:aws:iam::123456789012:role/example-role-redshift-env-dev"
      tags:
        "env": "dev"

teleport db configure bootstrap \    -c /etc/teleport.yaml \    --policy-name teleport-policy-rds-env-prod \    --attach-to-role "arn:aws:iam::123456789012:role/example-role-rds-env-prod"

db_service:
  enabled: "yes"
  databases:
  - name: "rds-postgres"
    protocol: "postgres"
    uri: "rds-postgres.abcdef012345.us-west-1.rds.amazonaws.com:5432"
    aws:
        assume_role_arn: "arn:aws:iam::123456789012:role/example-rds-access-role"

teleport db configure bootstrap \    -c /etc/teleport.yaml \    --policy-name teleport-policy-rds-access \    --attach-to-role "arn:aws:iam::123456789012:role/example-rds-access-role"

kind: db
version: v3
metadata:
  name: "rds-postgres"
  labels:
    env: "dev"
spec:
  protocol: "postgres"
  uri: "rds-postgres.abcdef012345.us-west-1.rds.amazonaws.com:5432"
  aws:
    assume_role_arn: "arn:aws:iam::123456789012:role/example-rds-access-role"

db_service:
  enabled: "yes"
  resources:
    # Matches env=dev databases and overwrites assume_role_arn.
    - labels:
        "env": "dev"
      aws:
        assume_role_arn: "arn:aws:iam::123456789012:role/example-env-dev-access"

    # Matches env=prod database, and use the assume_role_arn in the database's
    # definition or use the host IAM identity if assume_role_arn is empty.
    - labels:
        "env": "prod"

teleport db configure aws create-iam --types rds --name teleport-rds-access
teleport db configure aws print-iam --types rds

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "rds-db:connect",
            "Resource": "*"
        }
    ]
}

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "redshift:GetClusterCredentials",
            "Resource": "*"
        }
    ]
}

Unable to cancel a query

If you use a PostgreSQL cli client like psql, and you try to cancel a query with ctrl+c, but it doesn't cancel the query, then you need to connect using a tsh local proxy instead. When psql cancels a query, it establishes a new connection without TLS certificates, however Teleport requires TLS certificates not only for authentication, but also to route database connections.

If you enable TLS Routing in Teleport then tsh db connect will automatically start a local proxy for every connection. Alternatively, you can connect via Teleport Connect which also uses a local proxy. Otherwise, you need to start a tsh local proxy manually using tsh proxy db and connect via the local proxy.

If you have already started a long-running query in a psql session that you cannot cancel with ctrl+c, you can start a new client session to cancel that query manually:

First, find the query's process identifier (PID):

SELECT pid,usename,backend_start,query FROM pg_stat_activity WHERE state = 'active';

Next, gracefully cancel the query using its PID. This will send a SIGINT signal to the postgres backend process for that query:

SELECT pg_cancel_backend(<PID>);

You should always try to gracefully terminate a query first, but if graceful cancellation is taking too long, then you can forcefully terminate the query instead. This will send a SIGTERM signal to the postgres backend process for that query:

SELECT pg_terminate_backend(<PID>);

See the PostgreSQL documentation on admin functions for more information about the pg_cancel_backend and pg_terminate_backend functions.

SSL SYSCALL error

You may encounter the following error when your local psql is not compatible with newer versions of OpenSSL:

tsh db connect --db-user postgres --db-name postgres postgres
psql: error: connection to server at "localhost" (::1), port 12345 failed: Connection refused    Is the server running on that host and accepting TCP/IP connections?connection to server at "localhost" (127.0.0.1), port 12345 failed: SSL SYSCALL error: Undefined error: 0

Please upgrade your local psql to the latest version.

Next steps

• Learn how to restrict access to certain users and databases.

• View the High Availability (HA) guide.

• Take a look at the YAML configuration reference.

• See the full CLI reference.

• Set up automatic database user provisioning.