https://github.com/tmulc18/googlecloudsetup

Getting started with google cloud setup
https://github.com/tmulc18/googlecloudsetup

Last synced: about 2 months ago
JSON representation

Getting started with google cloud setup

• Host: GitHub
• URL: https://github.com/tmulc18/googlecloudsetup
• Owner: tmulc18
• License: mit
• Created: 2017-07-18T08:10:06.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-08-22T21:11:04.000Z (over 7 years ago)
• Last Synced: 2024-10-29T21:10:10.812Z (3 months ago)
• Language: Python
• Size: 23.4 KB
• Stars: 2
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# GoogleCloudSetup

Getting started with [Google Cloud](https://cloud.google.com).

# [Quotas](https://docs.google.com/forms/d/e/1FAIpQLSe291Fkdz1BuSe33-lSWXe5L_WmVhdeTq0WIE-wlREGz9zkDA/viewform)

Need to have quota > 0 in the region you want to compute.

# Create instance

1. Google cloud console

2. Main menu (top left)

3. VM Instances

4. Custom settings

5. Choose instance with GPU and pick Ubuntu 16.04 LTS as OS

[VM Instance link](https://console.cloud.google.com/compute/instances)

# [SSH Keys](https://cloud.google.com/compute/docs/instances/adding-removing-ssh-keys)

1. Generate the key `ssh-keygen -t rsa -f ~/.ssh/[KEY_FILE_NAME] -C [USERNAME]`

2. Set permissions `chmod 400 ~/.ssh/[KEY_FILE_NAME]`

3. In local terminal run `cat ~/.ssh/[KEY_FILE_NAME].pub` and copy contents

4. Add key to the project [here](https://console.cloud.google.com/compute/metadata/sshKeys)

# Logging into the server

1. Navigate to the vm instances page and find your ip address

2. `ssh -i ~/.ssh/[KEY_FILE_NAME] [IP_ADDRESS]`

# Install GPU drivers, Python packages, etc. 

`sh instance_setup.sh`

# Refresh

`source activate ~/.bashrc`

# Test Install

It's important to test your intall.  Do this with 

`python -c "import tensorflow"` which should run without errors.