Jeff Layton has uploaded this change for review.
FSAL_CEPH: add FS cluster ID into the FSAL_CEPH filehandle
Currently, the FSAL-specific portion of the filehandle is composed of a
vinodeno_t, which is just an inode number and snapid tuple.
In a follow-on patch we're going to allow exporting alternate, named
filesystems within the same cluster. At that point, the vinodeno_t is no
longer unique enough for this purpose.
This patch expands the ceph_handle with a new 64-bit fscid field, that
denotes the cluster ID within the filesystem. From there, teach the
cluster to encode the fscid in the handle when it's available.
Ceph reserves fscid 0 to denote "legacy" filesystems. If no filesystem
name string is specified, or if it is a legacy filesystem, then we don't
append the fscid to the handle, in order to maintain backward
compatibility.
Change-Id: Id8818f181f9abeb49b2845c551e6e2b83e80db9b
Signed-off-by: Jeff Layton <jlayton@redhat.com>
---
M src/FSAL/FSAL_CEPH/export.c
M src/FSAL/FSAL_CEPH/handle.c
M src/FSAL/FSAL_CEPH/internal.c
M src/FSAL/FSAL_CEPH/internal.h
4 files changed, 31 insertions(+), 14 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/60/430960/1
To view, visit change 430960. To unsubscribe, or for help writing mail filters, visit settings.