OverlayFS File-System Proposed For Linux 3.10 Kernel
Phoronix: OverlayFS File-System Proposed For Linux 3.10 Kernel
The Overlay File-System has been in development for several years and is used by some notable Linux distributions, but has yet to be merged into the mainline Linux kernel after having to be pulled a few times in the past. The new plan is to merge OverlayFS for the Linux 3.10 kernel...
Wow, that really took long. I would have liked when aufs was added as that's what i prefer to use but if overlayfs is mainline this is also fine. Currently no mainline kernel can be used to create fast live systems. There is a workaround userspace tool called unionfs-fuse but it is so slow that nobody can really consider it. Fedora used another way to implement the live mode but that seemed a bit more restrictive as the amount of free space did not seem to rely on the ram installed.
What implementations of union overlay are there out there?
Does not Linux already have support for union overlay?
Plan 9 from Bell Labs have support for union mounts, I think it originated from there, it was the first OS with support for union mount and was invented and introduced there.
Informer many live systems used unionfs and later aufs and Ubuntu is now using overlayfs. Of course I can not test all other live systems but at least i know the most important ones.
Actually union mounts were the favored solution of the kernel VFS maintainers. But the union mounts project was abandoned by its developer, which lead to the aufs maintainer to request inclusion of aufs and overlayfs in the kernel:
iirc, Fedora uses LVM writable snapshot with memory backed VG. All other distros use overlayfs, aufs or unionfs for their live media.
Originally Posted by J. R. Okajima
Hehe... looks like Michael has found a way to defeat ad blockers .
If you use AdBlockPlus then it automatically allows in non-intrusive advertising-- text adds, static images, things like that. It only really blocks audio/video/gif/flash ads which is fine by me because those are the worst.
Originally Posted by VinzC
On my home server, I use mhddfs to pool together various disparate hard drives into one combined volume. It's based on FUSE so it's not the most efficient solution out there, but it works very well for me. Plus, it has the advantage of automatically choosing the filesystem where a file will be written, depending on the free space available.