ahmza
1ef8f861c1
'docker compose down' removes caddy_cf and the default network, which intermittently severs the SSH session's own network path before 'up' restores it (remote command exited without exit status). Replace with 'pull && up -d --force-recreate --remove-orphans' so the stack is reconciled in place without tearing the network out from under the connection, while still picking up the new :latest image.
blog-hugo
I am hoping to migrate my dynamic blog currently hosted using GHOST CMS to a static site using HUGO.
Pros of static sites
- Security
- Reliability
- Speed
- Scalability
- Low Resource requirement
Why
The reasons above all fit well with my current goal of trying to containerise all of the services running on my home lab. Currently I have 2 servers running Proxmox VE with a combined 48 threads and 256GB of RAM which as you can imagine take a lot of power.
The plan is to slowly move the majority of my services to a cluster of 4 RaspberryPis. If all the services required support the ARM CPU architecture then I can completely decommission the power hungry servers.
How
I am still coming up with the deployment method of this site but from what I have seen the Github CI/CD pipeline makes a lot of sense as the site will be hosted on Github anyway.
- Update Github with new code
- Github generates a docker image
- Kubernetes cluster uses that image to host the site
When Pulling / Cloning
git clone --recurse-submodules https://github.com/4hmz4/blog
Update the theme
git submodule update --remote --merge