> From: Michael Torrie <[email protected]> > On 05/14/2013 01:44 AM, Gabriel Gunderson wrote: >> There are real reasons why this might not *actually* work. Be aware of >> caches, buffers, journals, snapshots, flushes, and RAIDs, oh my!
> And on copy-on-write filesystems such as BtrFS and ZFS, I imagine it's > not possible to overwrite an existing block; each shred pass would > simply allocate a new block. Eventually old blocks would be reused of > course. Which means, if you're using that sort of filesystem and you really need to be sure that the blocks are overwritten, you may need something like pv -cptrb -N blanking < /dev/zero > zero.000 rm -f zero.000 but be warned that blanking all the free space on the device may take a while (so don't do it too often) :) /* PLUG: http://plug.org, #utah on irc.freenode.net Unsubscribe: http://plug.org/mailman/options/plug Don't fear the penguin. */
