Yes, but ReadOnly volumes are hopefully replicated to more than one
server. So if the only volumes that I was using from serverA (the one
about to be shutdown) were RO volumes (and I therefore had no
callbacks from serverA), the shutdown shouldn't effect me. If I need
to access anything else from that RO volume, when the CacheManager
determined its original choice was unavailable, it would simply find
and use another RO replica.
Depending on how long a server is going to be down and the contents of
the server (heavily used project and user volumes or just a bunch of
ROs and little used RWs), contacting just those who have current
callbacks may not be adequate. If a server is going to be down for an
unscheduled and extended period of time, and my project directory is
on it, I'd want to about it even if I wasn't currently using any data
>From my project directory.
Pierette
