Re: Quality of Oracle MetaLink Notes

From: Gus Spier <gus.spier_at_gmail.com>
Date: Tue, 25 Mar 2008 13:07:06 -0400
Message-ID: <6d3967610803251007h320bf8cbk9a0db3b5620b2563@mail.gmail.com>


Rather than relying on my evaluation of my expertise, how much better if metalink had a set of "rules" to characterize my previous interactions with metalink? The forums on OTN don't seem to have a problem awarding skill levels to contributors. Surely, Oracle can whip up a similar utility.

FWIW, R,

Gus
On Tue, Mar 25, 2008 at 11:55 AM, Robert Freeman <robertgfreeman_at_yahoo.com> wrote:

> I think part of this issue is one of language/culture with respect to the
> analyst. With a global support organization you are going to get people
> who's first language is not English (or even American English). Thus, their
> vocabulary will not be quite as wide as a natural speaker. Additionally this
> probably makes for more halting and harsh sounding reporting, when in fact
> that may not have been the intent. Finally I find the first level support at
> Oracle to often be lacking. If you get a first level analyst making the
> report (or perhaps they are trying in vain to paraphrase a development
> response), they may not have the knowledge to accurately write what has been
> reported. Some of the initial responses I get to SR's are nothing more than
> the analyst going through the docs and finding something that I already knew
> a long time ago.
>
> I kind of wish Metalink had a set of radio buttons so you could describe
> your skill set (expert, advanced, intermediate, beginner) so they would not
> do stuff like that. It wastes my time and I sometimes think that they do it
> in an effort to just get the SR "answered"..... irrespective of the fact
> that the answer is about as helpful as fire ants in Florida.
>
> It think it falls on us, as openers of an SR, to ensure that the reporting
> is accurate and correct. Granted, we pay for support and we can complain
> until the cows come home about how we should not have to do this, or do
> that, but in the end if we follow-up and ensure accuracy and appropriate
> grammar, etc we all benefit. That being said, I'll confess that there are
> times that I'm just so frustrated with the Analyst after working an SR that
> I just don''t want to have anything else to do with them.
>
> Just my opinion, YMMV...
>
> RF
>
>
> Robert G. Freeman
> Author:
> Oracle Database 11g New Features (Oracle Press)
> Portable DBA: Oracle (Oracle Press)
> Oracle Database 10g New Features (Oracle Press)
> Oracle9i RMAN Backup and Recovery (Oracle Press)
> Oracle9i New Feature
> Blog: http://robertgfreeman.blogspot.com (Oracle Press)
>
> ----- Original Message ----
> From: Hemant K Chitale <hkchital_at_singnet.com.sg>
> To: oracle-l_at_freelists.org
> Sent: Tuesday, March 25, 2008 8:27:51 AM
> Subject: Quality of Oracle MetaLink Notes
>
>
> In the past 1 (or 2?) years, the quality of notes on MetaLink has
> significantly deteriorated.
> Some are outright misleading (and potentially dangerous to novice DBAs).
> However, in recent months, I have noticed notes that are also
> arrogant or disrespectful
> to the customer.
>
> One example, one which I did send feedback to Oracle Support is Note
> 558846.1.
>
> This is the feedback that I have sent :
>
> I find the language used in Note 558846.1 :
> 1. Unclear
> 2. Arrogant or dismissive
>
> The Symptoms section states
> "Running a SQL script that returns a great amount of data on Windows"
> while the Cause section, referring to Bug6867504 states
> "On Windows if you issue highly recursive or very large SQL
> statements you will blow the RDBMS stack"
>
> Is the Bug logged against "a great amount of data"
> OR is it logged against "highly recursive SQL"
> OR is it logged against "very large SQL statement"
>
> What is "a great amount of data" ? 5MB ? 500MB ? xx number of records
> ? Some figure with respect to a fixed Buffer Size ?
> What is "highly recursive SQL" ? One that makes 10 recursive calls ?
> One that makes 100 recursive calls ?
> What is "very large SQL statement" ? One that has a text length of
> 5000 characters ? A length of 50000 characters ? A length of 5Mbytes ?
>
> Is the langauge "blow the RDBMS stack" one that is used by a
> Technical Support person talking to a DBA/Developer ?
> What does it mean by "blow .. the stack" ? Should it be "exceed the
> hardcoded stack size of 1MB " ?
>
> What is related to the stack size ? "a great amount of data" OR
> "highly recursive SQL" OR "very large SQL statement" ?
> WHERE is the problem ?
>
> Is the solution section
> "Note that any SQL statement that has a lot of repeated values is a
> poor SQL and will probably cause such problems so it's best never to
> use such bad SQL and try to tune your queries.
> If you have a statement that will not work within the 5 MB stack that
> you have adjusted, you will never know what the correct results are
> anyway."
> a REAL WORLD Solution recommendation ? (and, by the way what is "a
> lot of repeated values" ? how many is "a lot" ?)
>
> How does your analyst define "poor SQL" and "bad SQL" in the context
> of this particular Note and Bug ?
> If I have an SQL statement that contains a very long INLIST such that
> it exceeds a certain size (what size ?) is it "poor SQL" or "bad SQL" ?
> And what does the analyst mean by "you will never know what the
> correct results are anyway" ? Is THAT the sort of response
> I expect from an RDBMS vendor ?
>
>
>
>
> Hemant K Chitale
> http://hemantoracledba.blogspot.com
>
>
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>
>
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

--
http://www.freelists.org/webpage/oracle-l
Received on Tue Mar 25 2008 - 12:07:06 CDT

Original text of this message