Connor,
Thanks for the examples, and the
It's useful to know that Windows and (at least some of the) Unix
implementations are completely different - with all considerations implied
by the sequential/random side effects.
Regards
Jonathan Lewis
On Mon, 4 Oct 2021 at 15:47, Connor McDonald <mcdonald.connor@xxxxxxxxx>
wrote:
Windows
SQL> select sys_guid() from all_tables where rownum <= 5;
SYS_GUID()
--------------------------------
CD48E63999534BEABA6BD8A58ACE134B
096CB75B5BDF472F92CD04DAEF98EFFC
E440AB8C9A3E4DE79E8F8043EFE0B0AB
F56E2B6AC420482A9B45C3470FA04F33
B3259B96759A454BABAF0C8C6C927CB1
and
SQL> select sys_guid() from all_tables where rownum <= 5;
SYS_GUID()
--------------------------------
F03D31EF992C4F3D92B496315A4BE319
D18A8AE49AB64E848938F4E328C91B97
B2A38E71E34F4436BAA13BAE13DAB5AA
6996772B3C2946869DDADF41BE1037CC
C249BA6CD0B14147A013B3A0E02E86B0
For as long as I can remember, Windows has been the only one to look
suitably "random". I've seen various projects (inside and outside of
Oracle) that steer clear of GUID for keys etc for this reason, often using
a blend of sequences and dbms_crypto.randombytes
On Mon, Oct 4, 2021 at 10:34 PM Jonathan Lewis <jlewisoracle@xxxxxxxxx>
wrote:
Quick follow-up.
Does anyone have an instance running on Windows that they can do this
test on
Thanks
Jonathan Lewis
On Fri, 1 Oct 2021 at 10:28, Jonathan Lewis <jlewisoracle@xxxxxxxxx>
wrote:
I put out a tweet yesterday (
https://twitter.com/JLOracle/status/1443302780266029061 ;) about
sys_guid() and the different pattern of results you could get depending on
the platform you were running on. The most recent check I had done was
12.1.0.2, and I don't have access to anything other than OEL 7.9 at the
moment to check more recent versions.
It would be nice if a few people could run the following query on their
systems and report the results, along with O/S, for 19c to see if there are
stll differences.
Here's my result from 19.11
SQL> select sys_guid() from all_tables where rownum <= 5;
SYS_GUID()
--------------------------------
CD47CA7FBE987B90E0530100007FE7D5
CD47CA7FBE997B90E0530100007FE7D5
CD47CA7FBE9A7B90E0530100007FE7D5
CD47CA7FBE9B7B90E0530100007FE7D5
CD47CA7FBE9C7B90E0530100007FE7D5
5 rows selected.
Almost simultaneously from another session
SYS_GUID()
--------------------------------
CD47CB493173114BE0530100007FA8A8
CD47CB493174114BE0530100007FA8A8
CD47CB493175114BE0530100007FA8A8
CD47CB493176114BE0530100007FA8A8
CD47CB493177114BE0530100007FA8A8
This seems to have a sequential component tied to the session (possibly
to the moment the session connected** rather than the session id). If you
follow the links in the tweet you'll find that Oracle on Windows used to
produce very random-looking leading values
you reconnect once per session and run t
** If I reconnect once per session then the 7th digit of the guid
increments by 1.
Regards
Jonathan Lewis
--
Connor McDonald
===========================
blog: connormcdonald.wordpress.com
twitter: @connor_mc_d
"If you are not living on the edge, you are taking up too much room."
- Jayne Howard
*Fine print: Views expressed here are my own and not necessarily that of
my employer*
