Re: Are NFS handles stable across major updates?

Jason Thorpe <[email protected]>
Newsgroups gmane.os.netbsd.devel.kernel
Message-ID <[email protected]>
> On Dec 13, 2024, at 3:52 PM, Edgar Fuß <[email protected]> wrote:
> 
>> raidframe without dk, these can be made to be stable
> The RAID has a GPT, so the file systems on it are dks.
> 
>> will return stale handle now.
> What I'm really concerned about is not ESTALE, but the possibility of an open 
> NFS handle to suddenly refer to /a different file/ after the update.

Right, the problem is NetBSD’s file handles have this little guy at the front:

typedef struct { int32_t __fsid_val[2]; } fsid_t; /* file system id type */

In a perfect world, each file system would have a UUID and the file handle would pack the UUID in front of the file ID rather than this fsid_t thing (which is really just container for a dev_t).

-- thorpej
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.