Re: Serializability of Transactions and Automatic (Number) Generators
Date: 5 Dec 2004 07:58:25 -0800
Message-ID: <18c7b3c2.0412050758.4dc17009_at_posting.google.com>
>> Yes. Sequence numbers and the like bear no relevance to the
serializability
of transactions. These constructions are merely a convenient way of
creating
unique numbers. <<
Let's go one step further. Having gaps is not the fundamental problem. These various exposed physical locators are all based on the state of the machine that stores the data and not the data itself. They are not a proper attribute, much less a key.
Mixing physical levels with logical levels does destroy a lot of properties, like uniqueness and serialization if you use them as a key.
>> If TxA were executed before TxB, it would obtain the values V(n)
and V(n+1) -- rather than V(n) and V(n+2) as shown above -- but for
all practical purposes, the results are equivalent: two new unique
values unused by any other transaction were generated and the specific
values are immaterial. <<
Let's give a milion redundant duplicates a unique exposed locator value to identify it and see if it is immaterial. Just leave your coffee cup on the "submit" button over night ...not that anyone has ever done that, of course :) Received on Sun Dec 05 2004 - 16:58:25 CET