[9278] in Athena Bugs
fs flushvol: continued investigation
daemon@ATHENA.MIT.EDU (akajerry@Athena.MIT.EDU)
Wed May 6 01:04:59 1992
From: akajerry@Athena.MIT.EDU
Date: Wed, 6 May 92 01:04:04 -0400
To: basch@MIT.EDU
Cc: bugs@MIT.EDU, bug-afs@MIT.EDU, trb@MIT.EDU, athena-ws@MIT.EDU
In-Reply-To: "Richard Basch"'s message of Tue, 5 May 92 17:22:27 -0400 <9205052122.AA20365@tardis.MIT.EDU>
I agree that introducing potential bugs of unknown extent into AFS at
this point would be bad (e.i., don't cross the streams), and Richard's
research seems to suggest that there is no way to implement the right
thing without doing just that.
So how about implementing the wrong thing for now. That is, a user
command that flushes the whole cache that a user can execute at
his/her discretion upon logout. This would basically put us back to
the same place we were with NFS security; the system is insecure by
default, but there is a user command that can make the system more
secure, albeit with a severe preformance hit for the next user.
Preserving the status quo seems to be the best we can to ask for at
the moment.
--Jerry
