You can create an API token by navigating to API tokensin Settings , as shown below. You can also provide an expiration date for your token upon creation. You can create multiple tokens and revoke them any time.

You can change any information in your profile from the Profile tab in your Settings:

When you create UDFs, notebooks, dashboards, groups, and ML models, TileDB Cloud stores each asset as a TileDB object (array or group). In order to do so, it requires you to provide a default storage path for storing those assets. You can do so from your profile settings as follows:

In order to be able to create, register, and access arrays through the TileDB Cloud service, you need to set up access credentials. For S3 compatible object stores, TileDB Cloud supports both IAM Roles and Access Credential key pairs. TileDB Cloud securely stores all keys in an encrypted database and never grants your keys to any other user. TileDB Cloud uses your keys in containarized stateless workers, which are under TileDB's full control and inaccessible by any other user's code (e.g., SQL or UDFs).

Note: You can add multiple AWS keys to TileDB Cloud, register different arrays with different keys, select a key to be your default key, and revoke any key at any time.

AWS Access Keys

You can add your AWS keys from the AWS credentials tab of Settings as follows:

AWS Assume Role

With an AWS AssumeRole policy we are solving the very same issue we used keys before: Enable AWS cross-account access, so that a role in one account can access a bucket in a separate account.

When using AWS AssumeRole, temporary keys are created through the Service Token Service (STS), and used from the deployment party (TileDB Cloud Console). This means that for organisation purposes there is no need to create an AWS IAM User for every user logging into TileDB Cloud Console and generate key pairs. Instead, after a User is authenticated, the AssumeRole functionality enables TileDB Cloud Console to access the bucket on behalf of a User and the credentials used in that case can be reused by multiple Users in the same organisation that need to access the same S3 buckets.

As an example, let 's consider the account (Account A) we are signing up with TileDB Cloud to access bucket(s) in User's AWS account (Account B). For that purpose, Account B has a bucket created. The most common setup is to create an IAM role for TileDB Cloud to use and then allow it to access a specific bucket with an AWS S3 bucket policy. Requests for access to the bucket will only be granted coming from our AWS account with our external ID.

Steps:

In TileDB Cloud Console navigate to Settings then select the tab Cloud Credentials

Click Add credentials, then select ARN Role and click Next and Next in the following step which is just a short description

Select tab Existing Role that presents the Account A ID as well as the External ID

Select tab New Role that proposes the JSON Account B can use to create the role. Please note Account A Account ID as well as the External ID

In Account B, User (or Admin) can create the bucket policy

In Account B User (or Admin) can create the role, using Account A ID and External ID

In Account B User (or Admin) has to attach the policy to the role

Obtain ARN for the new role

Then press Next in TileDB Console Add Credentials modal dialog and enter a name for the new AssumeRole Credentials and the ARN obtained in previous step

Test the connection

Example configurations have been detailed below:

AWS IAM Role:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Principal": {
                "AWS": "<PROVIDED IN UI>"
            },
            "Action": "sts:AssumeRole",
            "Condition": {
                "StringEquals": {
                    "sts:ExternalId": "<PROVIDED IN UI>"
                }
            }
        }
    ]
}

Note: Both the AWS Principal and External ID will be provided when attempting to register the ARN role

AWS S3 Bucket Policy:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "ListObjectsInBucket",
            "Effect": "Allow",
            "Action": ["s3:ListBucket", "s3:ListBucketMultipartUploads"],
            "Principal": {
                "AWS": "<IAM ROLE ARN>"
            },
            "Resource": ["arn:aws:s3:::<BUCKET NAME>"]
        },
        {
            "Sid": "AllObjectActions",
            "Effect": "Allow",
            "Action": ["s3:GetObject",
                       "s3:PutObject",
                       "s3:AbortMultipartUpload",
                       "s3:ListMultipartUploadParts",
                       "s3:DeleteObject"],
            "Principal": {
                "AWS": "<IAM ROLE ARN>"
            },
            "Resource": ["arn:aws:s3:::<BUCKET NAME>/*"]
        }
    ]
}

Setup KMS for the target bucket

It is possible that encryption is needed for the target bucket

To enable KMS usage for the target bucket, it is needed to edit the policy for the KMS key and add a statement that gives access to the role used previously

Example configuration is provided below

    "Version": "2012-10-17",
    "Statement": [
        ...
        {
            "Sid": "Example KMS Role",
            "Effect": "Allow",
            "Principal": {
                "AWS": "<IAM ROLE ARN>"
            },
            "Action": [
                "kms:Encrypt",
                "kms:Decrypt",
                "kms:ReEncrypt*",
                "kms:GenerateDataKey*",
                "kms:DescribeKey"
            ],
            "Resource": "*"
        },
        ...
    ]
}

You can add your payment method and see your billing information, including your current balance breakdown and all invoices, by navigating to the Billing tab of your Settings :

This page group contains simple recipes for managing your account. You can find the contents below: