Google Cloud Storage (GCS)¶

Pre-requisites¶

To create an Instance Secret for securing GCS credentials, you must have the following information:

Access Permissions in DataOS¶

To create an Instance Secret in DataOS, at least one of the following role tags must be assigned:

roles:id:data-dev
roles:id:system-dev

roles:id:user

    NAME     │     ID      │  TYPE  │        EMAIL         │              TAGS               
─────────────┼─────────────┼────────┼──────────────────────┼─────────────────────────────────
Iamgroot     │   iamgroot  │ person │   iamgroot@tmdc.io   │ roles:id:data-dev,              
             │             │        │                      │  roles:id:user,                  
             │             │        │                      │ users:id:iamgroot

Checking Assigned Roles

Use the following command to verify assigned roles:

dataos-ctl user get

If any required roles are missing, contact a DataOS Operator or submit a Grant Request for role assignment.

Alternatively, if access is managed through use cases, ensure the following use case is assigned:

Manage All Instance-level Resources of DataOS in User Layer

To validate assigned use cases, refer to the Bifrost Application Use Cases section.

Bifrost Governance

Source System Requirements¶

Project ID for the Google Cloud Project: The unique identifier of the Google Cloud project where your GCS bucket resides. You can retrieve this from the Google Cloud Console by navigating to your project dashboard and locating the *Project ID- under project information.
Email Address Associated with the Google Cloud Service Account: The email address of the service account used to authenticate and access your GCS bucket. This can be found in the IAM & Admin > Service Accounts- section of the Google Cloud Console by selecting the service account and viewing its details.
JSON Key File for the Google Cloud Service Account: The path to the JSON key file used to authenticate the service account. To obtain this, navigate to IAM & Admin > Service Accounts- in the Google Cloud Console, select the relevant service account, click Keys, and download the JSON key file by selecting Add Key > JSON. Ensure the file is stored securely, and note its path.

Ensure you have these credentials ready before proceeding with the Instance Secret creation process.

Create an Instance Secret for securing GCS credentials¶

Google Cloud Storage (GCS) is an object storage system. Object stores are distributed storage systems designed to store and manage large amounts of unstructured data.

To create a Google Cloud Storage (GCS) Instance Secret in DataOS, ensure you have access to the DataOS Command Line Interface (CLI) and the required permissions. Follow the steps below to complete the creation process efficiently and securely.

Step 1: Create a manifest file¶

Begin by creating a manifest file to hold the configuration details for your GCS Instance Secret. Depending on your access needs (read-only or read-write), start with the corresponding YAML template provided below.

🗣️ Note that for read-write access of an Instance Secret the user has to create two Instance Secrets one with `acl:r` and other with `acl:rw` with similar names such as `testdepot-r `and` testdepot-rw`. If a user creates an Instance Secret with only read-write access and does not create a separate read-only Instance Secret, an error will be triggered while applying the Depot manifest file, as shown below.

dataos-ctl apply -f /home/office/Depots/sf_depot.yaml
INFO[0000] 🛠 apply...                                   
INFO[0000] 🔧 applying testdepot:v2alpha:depot...        
WARN[0000] 🔧 applying testdepot:v2alpha:depot...error   
⚠️ Invalid Parameter - failure validating instance resource : invalid depot, missing a secret reference with acl: r
WARN[0000] 🛠 apply...error                              
ERRO[0000] failure applying resources

Read-only Instance SecretRead-write Instance Secret

name: ${gcs-depot-name}-r  # Name of the instance-secret, indicating it's for read-only access.
version: v1  # Version of the instance-secret.
type: instance-secret  # Specifies that this is an instance-secret.
description: ${description}  # Optional: Brief description of the instance-secret's purpose.
layer: user  # DataOS layer where the secret is assigned.
instance-secret:
type: key-value-properties  # Type of Instance-secret, stored as key-value pairs.
acl: r  # Access control level, set to 'r' for read-only access.
data:
    projectid: ${project-id}  # Unique identifier of the Google Cloud project that your GCS bucket resides in.
    email: ${client-email}  # Email address associated with the Google Cloud service account.
files:
    gcskey_json: ${path-to-gcskey-json}  # Path to the JSON key file for the Google Cloud service account.

name: ${gcs-depot-name}-rw  # Name of the instance-secret, indicating it's for read-write access.
version: v1  # Version of the instance-secret.
type: instance-secret  # Specifies that this is an instance-secret.
description: ${description}  # Optional: Brief description of the instance-secret's purpose.
layer: user  # DataOS layer where the secret is assigned.
instance-secret:
type: key-value-properties  # Type of Instance-secret, stored as key-value pairs.
acl: rw  # Access control level, set to 'rw' for read-write access.
data:
    projectid: ${project-id}  # Unique identifier of the Google Cloud project that your GCS bucket resides in.
    email: ${client-email}  # Email address associated with the Google Cloud service account.
files:
    gcskey_json: ${path-to-gcskey-json}  # Path to the JSON key file for the Google Cloud service account.

Resource meta section

The GCS manifest includes a Resource meta section with essential metadata attributes common to all resource types. Some attributes in this section are optional, while others are mandatory. For more details, refer to the configurations section.

Instance-secret specific section

This section focuses on attributes specific to GCS Instance Secrets. It includes details like:

type: Specifies the Instance Secret type (key-value-properties).
acl: Access control level (read-only or read-write).
data: Contains sensitive information such as Azure endpoint suffix, storage account key, and storage account name.

For more information, refer to the configurations section.

Step 2: Apply the manifest¶

To create the GCS Instance Secret within DataOS, use the apply command. Since GCS Instance Secrets are Instance-level resources, do not specify a workspace while applying the manifest.

CommandAlternative CommandExample Usage

dataos-ctl resource apply -f ${manifest-file-path}

dataos-ctl apply -f ${manifest-file-path}

dataos-ctl resource apply -f depot_secret.yaml
Example usage:
$ dataos-ctl apply -f depot_secret.yaml
INFO[0000] 🛠 apply...                                   
INFO[0000] 🔧 applying depotsecret-r:v1:instance-secret... 
INFO[0004] 🔧 applying depotsecret-r:v1:instance-secret...created 
INFO[0004] 🛠 apply...complete

Step 3: Validate the Instance Secret¶

To validate the proper creation of the Instance Secret in DataOS, use the get command.

CommandAlternative Command

dataos-ctl resource get -t instance-secret

INFO[0000] 🔍 get...                                     
INFO[0000] 🔍 get...complete                             

        NAME     | VERSION |      TYPE       | WORKSPACE | STATUS |  RUNTIME  |  OWNER             
-----------------|---------|-----------------|-----------|--------|-----------|------------------------------
     depotsecret | v1      | instance-secret |           | active |           | iamgroot

To get the list of all the Instance Secret within the Dataos environment execute the following command.

CommandExample Usage

dataos-ctl resource get -t instance-secret -a

dataos-ctl resource get -t instance-secret -a
INFO[0000] 🔍 get...                                     
INFO[0000] 🔍 get...complete                             

            NAME            | VERSION |      TYPE       | WORKSPACE | STATUS | RUNTIME |         OWNER          
-----------------------------|---------|-----------------|-----------|--------|---------|------------------------
abfssv2alpha-r             | v1      | instance-secret |           | active |         | iamgroot       
abfssv2alpha-rw            | v1      | instance-secret |           | active |         | iamgroot       
abfsswithoutmetastore-r    | v1      | instance-secret |           | active |         | thisisthor              
abfsswithoutmetastore-rw   | v1      | instance-secret |           | active |         | thisisthor

Alternatively, you can also check on Metis UI by searching the Instance Secret by name.

Metis UI

Delete the Instance Secret¶

🗣️ Before you can delete an Instance Secret, you need to make sure there are no other Resources dependent on it. For example, if a Depot has a dependency on an Instance Secret, trying to delete that Instance Secret will cause an error. So, you'll need to remove the Depot first, and then you can delete the Instance Secret. This rule applies to not just Depot, but all dependent Resources, such as Workflow, Service, Worker, etc. The following error will be thrown if any Resource has a dependency on Instance Secret as shown below. Example usage:

dataos-ctl delete -t instance-secret -n sampleinstsecret
INFO[0000] 🗑 delete...                                  
INFO[0000] 🗑 deleting sampleinstsecret:v1:instance-secret... 
INFO[0001] 🗑 deleting sampleinstsecret:v1:instance-secret...error 
WARN[0001] 🗑 delete...error                             
ERRO[0001] Invalid Parameter - failure deleting instance resource : cannot delete resource, it is a dependency of 'depot:v2alpha:sampleinstdepot'

To delete an Instance Secret, use one of the following methods:

Method 1¶

Specify the Resource type and Instance Secret name in the delete command.

CommandAlternative CommandExample Usage

dataos-ctl resource delete -t ${resource-type} -n ${resource-name}

dataos-ctl delete -t ${resource-type} -n ${resource-name}

dataos-ctl resource delete -t instance-secret -n sampleinstsecret
Expected output:
dataos-ctl delete -t instance-secret -n sampleinstsecret
INFO[0000] 🗑 delete...                                  
INFO[0000] 🗑 deleting sampleinstsecret:instance-secret...deleted
INFO[0000] 🗑 delete...complete

Method 2¶

Copy the Instance Secret name, version, and Resource-type from the output of the get command separated by '|' enclosed within quotes and use it as a string in the delete command.

CommandAlternative CommandExample Usage

dataos-ctl resource delete -i "${resource-name|version|resource-type}"

dataos-ctl delete -i "${resource-name|version|resource-type}"

dataos-ctl delete -i "sfdepot01-r | v1      | instance-secret | public   "
INFO[0000] 🗑 delete...                                  
INFO[0000] 🗑 deleting sfdepot01-r:v1:instance-secret... 
INFO[0000] 🗑 deleting sfdepot01-r:v1:instance-secret...deleted 
INFO[0000] 🗑 delete...complete

Method 3¶

Specify the path of the manifest file and use the delete command.

CommandAlternative CommandExample Usage

dataos-ctl resource delete -f ${manifest-file-path}

dataos-ctl delete -f ${manifest-file-path}

dataos-ctl delete -f /home/desktop/connect-city/instance_secret.yaml
Expected output:
INFO[0000] 🗑 delete...                                  
INFO[0000] 🗑 deleting sampleinstsecret:instance-secret...deleted
INFO[0000] 🗑 delete...complete