Reworked patch set in response to Neil's comments. This patchset
requires his patch as a prerequisite:
http://www.spinics.net/lists/linux-nfs/msg66341.html
These patches implements behavior that allows a failed server to be "cut
off" (e.g. all RPC tasks fail with -EIO) with the following sequence of
operations:
mount -o remount,serverfailed
umount -f
In the event you want to try and recover the server, you can remount to
remove the serverfailed status:
mount -o remount,noserverfailed
Although, I'm not sure what implications that might have for data loss.
For the most part I think it is pretty straight forward, the only thing
I am unsure about is the last two patches. This behavior is really
useful in my use case (where the server disappears) as it prevents the
umount -f from blocking while trying to send some DESTROY_SESSION
messages to the client (as long as you mounted with nosharecache).
Version 1 can be found at:
http://www.spinics.net/lists/linux-nfs/msg66348.html
Thanks in advance for the reviews and comments
Joshua Watt (7):
SUNRPC: Add flag to kill new tasks
SUNRPC: Expose kill_new_tasks in debugfs
SUNRPC: Simplify client shutdown
NFS: Add mount flags mask
NFS: Add serverfailed mount option
NFS: Propagate NFS_MOUNT_UNSHARED to clients
NFS: Propagate operations to unshared clients
fs/nfs/client.c | 9 +++-
fs/nfs/inode.c | 6 +++
fs/nfs/internal.h | 1 +
fs/nfs/nfs3client.c | 2 +
fs/nfs/nfs4client.c | 4 ++
fs/nfs/super.c | 98 ++++++++++++++++++++++++++----------------
include/linux/nfs_fs_sb.h | 1 +
include/linux/sunrpc/clnt.h | 1 +
include/uapi/linux/nfs_mount.h | 1 +
net/sunrpc/clnt.c | 11 ++---
net/sunrpc/debugfs.c | 4 ++
net/sunrpc/sched.c | 3 ++
12 files changed, 95 insertions(+), 46 deletions(-)
--
2.13.6
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
