1 2 3 4 5 6 Previous Next 163 Replies Latest reply: Oct 22, 2009 5:23 PM by Hoek Go to original post RSS
      • 45. Re: Index rebuild
        108476
        Hi Alan,
        all things in the database stay the same but through monitoring of indexes
        Do you have a script that you use to detect index rebuilding candidates?
        we do see an improvement is performance when indexes are rebuilt.
        Do you worry about accidentally rebuilding an index that might not need it?

        Do you rebuuild them all on a scheduled cron?
        I am no great expert on the intracacies of the internals of the database
        Oracle is a large computer program, written by people.

        It follows no laws of nature, and many people see no benefit to ripping into the internals of a man-made program. . . .

        It's easier for me to open an SR and have them look at the source code!
        • 46. Re: Index rebuild
          108476
          Hi Alan,
          The model for assessing whether an index needs to be rebuilt is some complicated mathematical equation which an ex colleague created.
          Great!
          From experience this has always proved to be correct over a wide range of databases.
          That's how good scripts are created, it's a process of successive refinement and empirical testing.

          Is it possible to share this script?
          • 47. Re: Index rebuild
            108476
            Hi Marcin,
            I think that this is a technical forum and I can't understand why to mix religion with technology.
            Me neither!

            That's why I found the Jonathan Lewis athiest quote to be so offensive.

            He tends to do thing like this when he gets backed into a corner, nobody understands why . . .

            But yeah, people don't like being called gullible fools just because they believe in things that cannot be proven
            So why test case are useless ?
            1 - Because they cannot be generalized effectively - exceptions exist

            2 - Because they can be used to decieve - You can change init.ora parms, schema stats, and get whatever results you desire.

            You cannot beat real-world empirical observations of a working database . . .

            It's the old saying "when all you have is a hammer, everything looks like a nail".

            The problem is that some people here don't have access to any production databases . . .
            • 48. Re: Index rebuild
              Jonathan Lewis
              Don Lewis wrote:
              +"Your manuscript is both good and original; but the part that is+
              +good is not original and the part that is original is not good."+

              Samuel Johnson

              Er, no. Not Johnson. No PROOF he ever said that.

              Thank you for that correction - there seems to be a website devoted to his sayings that agrees with you:
              http://www.samueljohnson.com/goodorig.html

              Regards
              Jonathan Lewis

              +"Your manuscript is both good and original; but the part that is+
              +good is not original and the part that is original is not good."+
              Samuel Johnson (attrib)
              • 49. Re: Index rebuild
                Jonathan Lewis
                Don Lewis wrote:
                "If you claim that something is true, I will examine the evidence
                which supports your claim; if you have no evidence, I will not
                +accept that *wat* you say is true and I will think you a foolish +
                and gullible person for believing it so."
                Richard Dawkins


                Knew it: Dawkins is illiterate.
                If you examine the evidence available in this thread alone, you would realise that the error is more likely to be a typing error on my part.


                Regards
                Jonathan Lewis

                +"If you claim that something is true, I will examine the evidence+
                +which supports your claim; if you have no evidence, I will not+
                +accept that what you say is true and I will think you a foolish+
                +and gullible person for believing it so"+

                Richard Dawkins
                • 50. Re: Index rebuild
                  706417
                  You don't say?


                  Love - Don Lewis

                  In this house, we OBEY the laws of thermodynamics!
                  Homer Simpson
                  • 51. Re: Index rebuild
                    635471
                    burleson wrote:
                    Oracle is a large computer program, written by people.

                    It follows no laws of nature, and many people see no benefit to ripping into the internals of a man-made program. . . .

                    It's easier for me to open an SR and have them look at the source code!
                    I think that this would make a great corporate philosophy for a consultancy business.

                    *"We don't try to understand the internals. We just open SR's!"*

                    A bold statement.
                    • 52. Re: Index rebuild
                      MarcinP
                      burleson wrote:
                      Hi Marcin,
                      I think that this is a technical forum and I can't understand why to mix religion with technology.
                      Me neither!

                      That's why I found the Jonathan Lewis athiest quote to be so offensive.
                      Don,

                      You started writing about religion not Jonathan. Quote is a quote and it can be anything.
                      If something is too offensive to you just skip it.

                      So why test case are useless ?
                      1 - Because they cannot be generalized effectively - exceptions exist
                      So testing of software is wasting of time ?

                      >
                      2 - Because they can be used to decieve - You can change init.ora parms, schema stats, and get whatever results you desire.
                      As I said - well defined input - well defined output - most of Jonathan test cases (especially in his book) are very well documented
                      You cannot beat real-world empirical observations of a working database . . .
                      But you can try figure out how it is working - using a test cases - for 24/7 systems there is no time for testing a new solution on production database.
                      It's the old saying "when all you have is a hammer, everything looks like a nail".
                      but what if you put hammer on your foot ?
                      It can be very painful and your patient can die. Ask air lines to stop booking a tickets and rebuild indexes in their databases ;)
                      Last time when one of that kind of system was out of service - it was huge delay on most of European airports.


                      I know that you have university background and I can't understand why you are fighting with a test cases and models - most of new science is base on models.
                      Can you imagine a real time observations of big ban ?


                      regards,
                      Marcin Przepiorowski
                      • 53. Re: Index rebuild
                        UweHesse
                        >
                        On the other hand, there are people of questionable provenance, who I cannot trust, even when they "prove" something . . . They cheat, that much I can prove . . .

                        It's the very core nature of credibility, and that's why personal integrity is so important . . .
                        >

                        If somebody "cheats" (resp. come to wrong conclusions) in a test case, you can prove that because there are facts and evidence to refer to. That is a good thing, isn't it? That is almost impossible to do for mere claims.

                        I agree to you that personal credibility is important, though. But that alone is not sufficient, in my view. In other words: I would be inclined to believe if say, Tom Kyte states something about Oracle Databases, but I would strongly prefer to check it myself with data & facts provided by him that are reproducable for me. And that is what he (and Jonathan Lewis also) do time and time again.

                        Nobody is perfect, everybody is bound to make mistakes. Even experts with much experience :-)

                        Kind regards
                        Uwe

                        http://uhesse.wordpress.com
                        • 54. Re: Index rebuild
                          706417
                          But! With the best will in the world, none outside of Oracle does fully understand the internals. It's not open source, is it? If I see a problem that looks like an internals issue, guess what? I raise an SR. We pay enough for it. If my washing machine goes funny, I don't have a quick Google and then dig out the spanner set! It's time to call the supplier or the footpad plumber!

                          The real problem with the drive towards trying to fully understand that which one cannot hope to fully understand is that one ends up doing the equivalent of sticking a knitting needle down the back of a TV, thinking it's science. Then one opens an SR!

                          Regards - Don Lewis
                          • 55. Re: Index rebuild
                            Billy~Verreynne
                            Don Lewis wrote:
                            But! With the best will in the world, none outside of Oracle does fully understand the internals. It's not open source, is it? If I see a problem that looks like an internals issue, guess what? I raise an SR. We pay enough for it. If my washing machine goes funny, I don't have a quick Google and then dig out the spanner set! It's time to call the supplier or the footpad plumber!
                            Don't confuse your self proclaimed incompetence to fix a washing machine, with the competence of others.
                            The real problem with the drive towards trying to fully understand that which one cannot hope to fully understand is that one ends up doing the equivalent of sticking a knitting needle down the back of a TV, thinking it's science. Then one opens an SR!
                            Silly analogy. Only an idiot that is totally ignorant of electronics will stick a knitting needle into the back of a TV. Are you calling most of us idiots here, for trying to fully understand an Oracle problem?

                            Let's go back to you analogy. There are a number of steps that can be performed to troubleshoot a TV problem. In fact, most TV's has a support/configuration/setup booklet that includes a list of symptoms, how to diagnose the problem, and how to correct that problem.

                            The very same approach holds true for Oracle. And the more one knows about Oracle, how it works, the better one is at diagnosing a problem. It does not take a SR to find out that the problem is trashing of the shared pool caused by a bad application that does not use bind variables.

                            Does it mean that all problems can be solved without a SR? Of course not. That is as stupid as saying that problems can only fully be understood and addressed using a SR.
                            • 56. Re: Index rebuild
                              706417
                              No Billy, I do not think you are all idiots. You seem to be a little agitated today. May I suggest that you take few deep breaths... underwater?


                              Love - Don Lewis
                              • 57. Re: Index rebuild
                                MarcinP
                                Billy  Verreynne  wrote:
                                Don Lewis wrote:
                                But! With the best will in the world, none outside of Oracle does fully understand the internals. It's not open source, is it? If I see a problem that looks like an internals issue, guess what? I raise an SR. We pay enough for it. If my washing machine goes funny, I don't have a quick Google and then dig out the spanner set! It's time to call the supplier or the footpad plumber!
                                Don't confuse your self proclaimed incompetence to fix a washing machine, with the competence of others.
                                Hi,

                                If you have a basic knowledge that washing machine need a electricity and water you can make a first check yourself.

                                In my opinion it is way of thinking - I have Oracle support as a product so I don't need to analyze anything.
                                But keep in mind that Oracle is a huge organization and you never know who will pick up your SR at the first line
                                and what will be a first and next steps.
                                And who will be blamed if you lost data or SLA for your customers ? DBA or Oracle ?

                                regards,
                                Marcin Przepiorowski
                                http://oracleprof.blogspot.com
                                • 58. Re: Index rebuild
                                  706417
                                  Electricity & water. Hmmm... no danger there, then!
                                  • 59. Re: Index rebuild
                                    Billy~Verreynne
                                    Don Lewis wrote:
                                    No Billy, I do not think you are all idiots.
                                    Many years ago, in our parliament, a member called the opposition members, idiots. The speaker took him to task and he had to withdraw that statement. Which he did, by saying that not all opposition members were in fact, idiots.
                                    You seem to be a little agitated today. May I suggest that you take few deep breaths... underwater?
                                    Nice...
                                    1 2 3 4 5 6 Previous Next