Finale: btrfs send/receive to replicate server images

• Last year I gave a talk on how google does live server upgrades by doing file level updates http://goo.gl/CxedK

• We use a custom rsync like system that was tricky to write and generates a lot of I/O

• Instead, you can use btrfs send/receive to replicate a golden image in a subvolume to all your servers

• Local changes are symlinked to a local FS or bind mounted on a file by file basis.

• Once the new subvolume has the new image, you can reboot to it and/or use kexec for a faster reboot cycle.

• btrfs-diff shows most files changed, use this to sync changes with cp –reflink instead of reboot http://goo.gl/fkLxAu

• Improve btrfs send/receive to only show files changed.