In message <[EMAIL PROTECTED]>,Rol
and Kuhn writes:
>Why is it necessary to identify a PAG by something else than the  
>equivalent keyring? Forgive my ignorance about keyring internals, but  
>I would imagine that a PAG is represented by processes holding a  
>reference to a certain keyring, which in turn contains an  
>authentication token for AFS.

i dont really understand what you are asking.   i guess you are asking
why the session keyring doesnt contain a key with the afs authentication
token.  it could.  however, a generalized pag interface would simply
provide a unique identifier for each process group.
_______________________________________________
OpenAFS-devel mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-devel

Reply via email to