From the
archive.

File Stubs provide info about the file but are not the file, we put that part somewhere else.

A cost-per-gigabyte breakdown comparing LTO-9 and cloud storage.

A deep look at how Linux's fanotify API makes this possible.

The Cloud is just one big cartrige. This is how we get the most out of every MB and minimize cost in the cloud.

Files are easy to find if we make some solid tags. This is how we do it.

SMR drives can save money and power if you treat them right.

The file system no longer provides the location of your files.

Moving from a testing environment to deployment means tuning up the configuration file to get the most out of your set up.

There is a difference! You can archive your backup and you can backup your archive, but there are a lot of things that separate the two.

Weather it is lots of small files, writing super big files, maxing capacity, giant unserchable directiories, this is how we address problems file systems have.

Metadata isn't nothing and when you get to one million files of metadata, your file system will still battle with the large amount of files. How do we overcome this ?

CMR drives, there is some room for improvement.

HuskHoard tells you what your data looked like at the exact moment before the first file was touched.

The Real 2026 Cost Math for Cold Archive