cursedfs icon indicating copy to clipboard operation
cursedfs copied to clipboard

BTRFS

Open allo- opened this issue 5 years ago • 2 comments

Btrfs's first metadata is at 64 KiB, which leaves plenty of space at the start for other FS's metadata (though not ZFS, the two couldn't coexist). But I don't know of a mechanism by which it can reserve space later in the disk where the other filesystems could store their data

Did you consider using a read-only snapshot for this?

allo- avatar Jan 21 '20 08:01 allo-

Would that allow us to create a contiguous chunk of space that btrfs wouldn't try to read to or write from? I assumed that they behaved similarly to ZFS's snapshots, where they simply preserve the current block structure of an existing filesystem/volume in their current location on disk, and which may need to be read in order to satisfy reads to the fs/vol that they're a snapshot of.

pcd1193182 avatar Jan 21 '20 17:01 pcd1193182

I think you can allocate a block that cannot be written to (except you delete the snapshot) by creating a file and then a read-only snapshot containing that file, but you will be able to read it.

I am not sure if there are any tricks to allocate blocks that do not belong to a file or hide existing snapshots.

allo- avatar Jan 21 '20 17:01 allo-