Re: [SPAM] oracle-l Digest V8 #84
From: Greg Rahn <greg_at_structureddata.org>
Date: Thu, 24 Mar 2011 12:33:35 -0700
Message-ID: <AANLkTimL4W01Qb8JJTiBrx_YoXa7B7ZH3H5wUnDhfO8Z_at_mail.gmail.com>
On Thu, Mar 24, 2011 at 11:08 AM, Michael Fontana <michael.fontana_at_enkitec.com> wrote:
>> Timo wrote:
>> Hibernate is not a reason for a problematic query. It is just another SQL generator.
>
> I respectfully disagree. While Hibernate IS a SQL generator, it is a particularly horrible one.
> The concept is sound, but the product fails in actual practice, especially when implemented against
> Oracle databases. My client project is way behind schedule because it costs them hundreds of thousands
> of dollars in billable hours before they realized this fact.
Date: Thu, 24 Mar 2011 12:33:35 -0700
Message-ID: <AANLkTimL4W01Qb8JJTiBrx_YoXa7B7ZH3H5wUnDhfO8Z_at_mail.gmail.com>
On Thu, Mar 24, 2011 at 11:08 AM, Michael Fontana <michael.fontana_at_enkitec.com> wrote:
>> Timo wrote:
>> Hibernate is not a reason for a problematic query. It is just another SQL generator.
>
> I respectfully disagree. While Hibernate IS a SQL generator, it is a particularly horrible one.
> The concept is sound, but the product fails in actual practice, especially when implemented against
> Oracle databases. My client project is way behind schedule because it costs them hundreds of thousands
> of dollars in billable hours before they realized this fact.
There are two sides to this coin:
A) Is the SQL code generation optimal (is there better SQL for the
same set of tables?)
B) Is the execution of the SQL optimal (is there a better execution
plan for this SQL?)
The solution may entail optimizing A or B or both A and B.
-- Regards, Greg Rahn http://structureddata.org -- http://www.freelists.org/webpage/oracle-lReceived on Thu Mar 24 2011 - 14:33:35 CDT