If you want to interact with google cloud through CLI or through code, you need to setup google cloud SDK on your workstation.
For example, if you want to write a python script that interacts with google cloud, you should have the sdk installed so that all the required packages for python will be available for your script.
This tutorial will guide you through the steps for setting up Google Cloud SDK on your workstation.
Also, I have added some of the tips and tricks that I have learned while using
gcloud CLI utility in my workstation.
Setup Google Cloud SDK
Note: If you’re using a google cloud instance, SDK is installed by default.
Step 1: Head over to https://cloud.google.com/sdk/ and copy the link to the appropriate installation package. This package is common for all Linux distributions.
For Windows users: Download the installation package from https://cloud.google.com/sdk/windows and install it directly on your systems. You will have the gcloud command available in the terminal.
Step 2: Download the package.
Step 3: Untar the sdk package.
tar -xvf google-cloud-sdk*
Step 4: Install the SDK using the
install.sh script. It will detect the appropriate package for your Linux distribution and installs it.
Step 5: Follow through the installation prompts and select the required options. It is pretty self-explanatory.
Step 6: Now, to validate the installation, open a new terminal session and execute the
If you dont open a new terminal session, you might see the following error.
-bash: gcloud: command not found
Configuring Google Cloud CLI With Google Project
Now that we have the SDK installed, we need to configure the Google cloud SDK to talk to a specific project.
We use the
gcloud utility from the installed SDK for CLI based google cloud access.
Follow the steps given below to configure
gcloud with a specific google project.
Step 1: Initialise the project configuration using the following command.
gcloud init --console-only
Step 2: You will see the following prompt. Select “Y”.
To continue, you must log in. Would you like to log in (Y/n)? Y
Copy and paste the URL in the browser. Then log in to your google cloud account and grant permissions to access google cloud resources.
You will see a page with a token. Copy the token and enter in the terminal.
Now you will be logged in, and you can select the project that you want to set as default, as shown in the image below.
Like this, you can add multiple gcloud configurations if you work with multiple google cloud accounts. You can switch between different accounts. I have explained this in the next section.
Google Cloud SDK Auth Config Location
To know all the configuration location of google cloud SDK, execute the following command.
All the SDK configs get stored in the
.config folder in your workstation home location.
If I run a tree from the
.config folder, this is what I have. The JSON files under legacy_credentials have the
client id and
client secret to connect to the google cloud account.
Validate Google Cloud CLI – gcloud
Now that we have a valid cloud configuration, let’s run some basic
gcloud CLI commands to verify the installation.
Step 1: List the credential account.
gcloud auth list
In my case, I have two google cloud accounts configured in my workstation. So it lists two accounts, and there is a
* in the active account.
When you have multiple
gcloud accounts, you can switch between accounts using the following command
gcloud config set account [email protected]
Step 2: To view your default
gcloud configs of the active configuration, use the following command.
gcloud config list
You can also manually change the configuration by editing the config file from the
Sometimes when you have multiple accounts configured, there could be a mismatch between activated account and active configuration. To solve this, first list the configruations.
gcloud config configurations list
If there is a mismatch, you can activate the correct configuration using the following command, assuming you activate the default configuration.
gcloud config configurations activate default
Step 3: To list all the
gcloud commands, use the following command.
gcloud Cheat Sheet
Well, I don’t have any cheat sheet for
However, there is a cool
gcloud cheat sheet command which displays the gcloud cheat sheet.
Also you can use the cheat sheet from google cloud documentation.
Other gcloud Components
When you install the google SDK, only the core google cloud libraries get installed.
But there are other components that are part of google cloud sdk.
kubectl. A utility required to interact with Google Kubernetes cluster.
To list all the available components execute the following command.
gcloud components list
Updating gcloud CLI
Updating gcloud CLI is pretty simple.
Use the following one-line to update gcloud components.
gcloud components update
Create a Google Compute Instance Using CLI
This section is only optional.
If you want to have a little hands-on with the gcloud command-line utility, you can follow this section to create a compute instance, access it, and then delete it.
To start with, we will create a instance using the CLI.
Step 1: Get the list of images using the following command.
gcloud compute images list
Step 2: The following command will create an f1 micro ubuntu instance with the ubuntu image from the highlighted image family.
gcloud compute instances create devopscube-demo-instance \ --image centos-7-v20170523 \ --image-project centos-cloud --set-machine-type f1-micro --zone us-central1-a
There are so many flags associated with instance creations. Refer to this official documentation for more information on the flags.
Step 3: To connect the instance via ssh, just execute the following command. The
gcloud command will automatically create the ssh key in your ~/.ssh folder if it doesn’t exist and connects to your instance.
Security Note: Do not deploy any public images in enterprise accounts even if you have permissions. Always use security-approved images in enterprise accounts to stay out of trouble.
gcloud compute ssh (instance-name)
gcloud compute ssh devopscube-demo-instance
Step 4: Delete the created instance using the following command.
gcloud compute instances delete (instance name)
gcloud compute instances delete devopscube-demo-instance
In this blog I have covered the steps to setup google cloud SDK and using it with gcloud CLI
Hope it helped or you might have more tips that can add value to the existing content.
Either way, please drop a comment.