Skip to content

Latest commit

 

History

History
58 lines (47 loc) · 1.75 KB

README.md

File metadata and controls

58 lines (47 loc) · 1.75 KB
Raw

introduction

Ansible is the new tool we used for network operations. All the host info used by ansible is in /etc/ansible/hosts Please make sure you have the ssh agent running with the right mainnet keys before any ansible operation.

ssh-add -l
2048 SHA256:K9D3flNNlwei50Hz78PXubKNacmSQqxiTaQfHf92bP8 leochen@MBP15.local (RSA)

install ansible

https://docs.ansible.com/ansible/latest/installation_guide/intro_installation.html

install ansible roles

ansible-galaxy install charliemaiors.rclone_ansible
ansible-galaxy install ryandaniels.create_users

role: create_users

https://github.com/ryandaniels/ansible-role-create-users

what is ansible vault

https://docs.ansible.com/ansible/latest/user_guide/vault.html

create users on users / test

  • edit inventory/test.hosts file to use your own host IP
  • make sure you have proper ssh setup to login to your own host

on ankr host

cd ~/experiment-deploy/ansible
ansible-inventory --list -i inventory/test.hosts
ansible-playbook playbooks/create-users.yml --ask-vault-pass --extra-vars "inventory=h2 user=ec2-user" -i inventory/test.hosts

on aws instance

cd ~/experiment-deploy/ansible
ansible-inventory --list -i inventory/hmy.hosts
ansible-playbook playbooks/users.yml --ask-vault-pass --extra-vars "inventory=devop" -i inventory/hmy.hosts

mainnet upgrade operation

cd ~/experiment-deploy/ansible
./scripts/upgrade-mainnet.sh

1) rolling_upgrade
2) restart
3) force_upgrade
4) quit

All the actions can operation on either a shard, a group, or a single host. By default, the rolling_upgrade action won't upgrade the current leader node. You need to use force_upgrade action to upgrade leader. restart can restart a node, or a shard, ex, s0, s1, s2, s3.