In message <[EMAIL PROTECTED]>, Glen Batchelor
<[EMAIL PROTECTED]> writes
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:owner-u2-
[EMAIL PROTECTED] On Behalf Of Anthony W. Youngman
Sent: Saturday, April 26, 2008 9:14 AM
To: u2-users@listserver.u2ug.org
Subject: Re: [U2] Guaranteed unique sequential keys
In message <[EMAIL PROTECTED]>, Glen B
<[EMAIL PROTECTED]> writes
> It's not overkill if you need a multi-app/multi-process service that
offers
>a single source for sequential keys. The other option is disk file
locking.
>Any way you look at it, you have to go to one place for the key. Don't
relay
>on O/S random numbers either. Even those are flawed at high resolution.
Ummm....
My first reaction to this was "linux has a guaranteed true random number
generator" - the only problem is that if it runs out of randomness it
simply stops generating and anything asking for a number has to wait.
Oh - and what do you mean by "high resolution"? I think that, if you
have a true random number generator, the chances of having a collision
approach 50% when you've used only 10% of the number space. (The classic
question about this is "how many people do you need in a room to have
even odds that two of them will share a birthday?". I think the answer
is "about 20".)
That is true and what I mean by high resolution is extremely frequent
requests. The scale of that is dependant on the hardware and the application
making the requests. When the entropy pool runs out, you get a function
block until more hardware data is available to generate a new value. That is
what I mean by "flawed". If you have a tiny box with a small entropy pool,
you could exhaust the entropy pool often which could lead to performance
issues. Of course, the new pseudo generators don't block at all. That is
what you are referring to below. The problem there is that entropy pool is
reused to generate pseudo bits to regenerate new keys from, which increases
the chance of repeated keys.
Except that there, the whole point is you DON'T WANT the entropy pool
used at all, because then you can't guarantee unique keys ... :-)
But then, if you want unique, pseudo-random keys, I think there are
generators that are guaranteed to return every possible number, only
once, each cycle through the number space.
None of this, however, solves the OP's original problem, I don't
think...
Cheers,
Wol
That's true. Random keys can not be sorted by order of creation.
Nor can they guarantee uniqueness.
Cheers,
Wol
--
Anthony W. Youngman <[EMAIL PROTECTED]>
'Yings, yow graley yin! Suz ae rikt dheu,' said the blue man, taking the
thimble. 'What *is* he?' said Magrat. 'They're gnomes,' said Nanny. The man
lowered the thimble. 'Pictsies!' Carpe Jugulum, Terry Pratchett 1998
Visit the MaVerick web-site - <http://www.maverick-dbms.org> Open Source Pick
-------
u2-users mailing list
u2-users@listserver.u2ug.org
To unsubscribe please visit http://listserver.u2ug.org/
