[OpenAFS] Re: Vos functions and clones and shadows

Steve Simmons scs@umich.edu
Tue, 26 Jun 2007 18:04:54 -0400

On Jun 26, 2007, at 5:50 PM, Adam Megacz wrote:

> I've updated the FAQ with the contents of this thread.
>>> What if you create a shadow volume which has the same name as a
>>> volume which exists elsewhere?  I assume that this would normally
>>> work, since the shadow doesn't get into the VLDB.  If then you
>>> made the shadow volume live, would it delete the VLDB entry for
>>> the original volume?  Perhaps without deleting it from the other
>>> disk?
>> Your assumption is correct, and the answer to all the above questions
>> is yes.
> What if you then run "vos syncvldb" against the server holding the
> original rw volume?  Will the volserver with the rw volume obliterate
> the formerly-shadow volume's entry?

We believe so, yes. At least, that's what's supposed to happen. IMHO  
this is a feature, not a bug, or it is for our purposes. Consider  
this scenario:

AFS servers goes down hard, looks like a multi-day outage. The  
appropriate shadows are promoted to production, even though they're  
five hours out of date. A few minutes later, the failed server  
springs back to life. What's better: to let the out-of-date shadows  
run, or to immediately recover/restore the production data of the  
last six hours. IMHO it's the latter.