Jeff Layton

Yeah -- we probably need to add some documentation comments to `struct ceph_client_callback_args`. If you wouldn't mind opening a bug at tracker.ceph.com, we can add some comments in the near...

So, to be clear: > @jtlayton if I may bother you for a moment, since you added the original set of callbacks: is it correct to say that using (only)...

The high level API was made to mirror the POSIX filesystem API. It has its own file descriptor table, etc. to closely mirror how the kernel syscall API works. e.g.:...

> acl_get_qualifier will call malloc to get memory... > > function find_entry and function posix_acl_2_fsal_acl use the acl_get_qualifier without freeing the memory... Huh? acl_get_qualifier(3) says _nothing_ about needing to free...

Oops, I stand corrected...you have to free the memory with acl_free(). So this sounds like it's a real bug.

> why nfs-ganesha does not have the real self mem management pool... it is also fine to directly use the talloc in samba .... You haven't articulated why we'd need...

I don't know of any magic bullets. This configuration is just not built for speed -- it concentrates all of the data and metadata access through the same servers, for...

On Tue, 2022-01-04 at 02:14 -0800, Kevin Chen wrote: > hi，everyone > in my project，i use nfs-ganesha as nas gateway in front of cephfs. > as we know, "For each...

TBQH, FSAL_CEPH is just not currently built for speed. As @dang points out, all the file I/O gets routed through ganesha before being distributed to the OSDs. That's a lot...

> A three node ceph cluster with nfs-ganesha enabled on each node. Does this mean you're running ganesha in parallel on all three nodes? And you're just moving the IP...