Re: Training for Oracle Performance tuning

Hi,
see comments below. Please, don't treat my comments as an attempt of any kind of personal insult. Definitely, the Method-R is not as simple to implement in the real life as it might be initially thought.

fmhabash wrote:
> I have been involved in so many performance issues since I attended
> this class, all of them have been diagnosed and resolved. However,
> only in a few of them I had to get to the level where I really needed
> to use M-R.

Seems, one important point in the Method-R is missed here - talk to the _*right person*_ (and on the right topic).

In most cases, where new (for the tuner) company is involved you meet the "middle tier" manager, who is quite often more concerned in the safety of his personal asshole and not the real needs of the business.

The "litmus paper" in this case would be the question from the manager's side similar to the "how we can stabilize the situation and _*avoid the same problems in future*_". That was a positive test. The negative one would be "fix it asap, and I don't care how you would do that".

I'm still involved in the "on-call support" for the two major mobile phone providers here in Russia (multi-TB OLTP databases with a most active part of ~0.5-1 TB though), not mentioning a number of TOP-10 banks as well as some government organizations, which require 24x7 support for there business critical systems. ... In the beginning of this century their approach was ... humm... as with the negative test :-). The requests for the session tracing, etc usually were rejected with the explanation "we have no time, asap, ..."... Of course, those issues were resolved (otherwise they would not be our customers till now, right?).

Now they are taught to pass the "positive test" and think of the right things :-). So in the Method-R there is some place for the customer education from the tuner side.

Sorry for the semi-offtop.

• Andrey

--
http://www.freelists.org/webpage/oracle-l