V0: Establish that witness believes the description
of error described in the error report.

V1: You advertise your system as carrier-grade?
(check but I think they do.
Otherwise get some juicy piece of boasting from their web page.)

V2: Describe what that means.

V3: Do you consider fault tolerance to be part of that definition?

V4: What reliability do you advertise?
(5 9s or 7 9s)

V5: Five 9s means that your expected down time is 4 minutes per year
is it not?

V6: How do companies keep their downtime down?

V7: Reliable components?

V8: No single point of failure?

V9: Fast recovery?

V10: Please describe each of these as they pertain to the DMs.

[If no to V10]
V11: So your five 9s advertisement is false.

[If yes to V10]
V12: Let's explore the no single point of failure
aspect of reliablity for a moment.
If one DM has a hardware failure then does the system fail?

[If yes to V12]
V12.a: This violates the single point of failure goal completely.

[If no to V12]
V13: Do you consider the fact that one failure should not
cause a system failure to be important?

[If yes to V13]
V14: In your implementation of the DMs, when one DM has a B tree
some of whose leaves are greater than 80, then other DMs will eventually
have such a situation too, correct?

[if no to V14]
V15: Then what does the error report mean when it says "all DMs maintain
an exact copy of the real time database"?

[If yes to V14]
V15: Also the error report says that "certain characteristics of the
1993 algorithms, allowed, in the context of certain unusual
operating practices, the balanced binary trees to become severely
unbalanced, thereby triggering the service interruption".
Do you see that?

V15.a: The 1993 algorithms you refer to here. Are they the company Y
implementations as referred to earlier in the error report when you say
"the balanced binary tree algorithm was implemented in the Tiger
systems in 1993"?

[if no to V15.a]
V15.b: Please show us the published account of the algorithm implemented.

[if yes to V15.a]
V15.c: So, the algorithm you are talking about is the one that
company Y implemented "based on the widely accepted
computer science algorithmic studies of Dr. RS
of Princeton Univerisity and Dr. W of Drew University"?
