Re: experience w/ UUID's for surrogate keys across entire schema?
From: Sayan Malakshinov <xt.and.r_at_gmail.com>
Date: Fri, 14 Sep 2018 15:46:49 +0300
Message-ID: <CAOVevU51fq9gJiZbGqcFLkBgqUnjVTX_43QpRyEoW671V-Z=hw_at_mail.gmail.com>
Date: Fri, 14 Sep 2018 15:46:49 +0300
Message-ID: <CAOVevU51fq9gJiZbGqcFLkBgqUnjVTX_43QpRyEoW671V-Z=hw_at_mail.gmail.com>
Hi Chris,
Of course, you can extend option 2 as multicolumn (SITE_ID,UUID) and use partitioning by SITE_ID to avoid some performance problems, but anyway because of the nature of UUID all inserts will require split operations on indexes, so sequences usually are better choice. It's much easier to solve problems with concurrency for right block of index (for example, using hash partitioning) than deal with constant splits
--
Best regards,
Sayan Malakshinov
http://orasql.org
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Sep 14 2018 - 14:46:49 CEST