[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-fsdevel
Subject:    LVM snapshots and journal file systems
From:       "Peter J. Braam" <braam () mountainviewdata ! com>
Date:       2001-05-21 15:53:26
[Download RAW message or body]

Hi,

I'm confused about the following:

What happens with an LVM snapshot during journal recovery of a file
system?  How does the LVM snapshot stay consistent?

Here is an example scenario I worry about.  Suppose that an ext3
transaction has been committed on disk to the journal inode and a block is
written to the file system volume.  The old data block is copied to the
snapshot volume and some redirection table is updated.  Suppose the system
crashes and this transaction is replayed from the journal: how can the LVM
determine if the copy has already been done (if it cannot, it runs the
risk of copying new data into the snapshot volume)?

There is probably a good solution to this and I'd like to know what it is.

Thanks!

- Peter -



-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org

[prev in list] [next in list] [prev in thread] [next in thread]