1 2 3 4 Previous Next 54 Replies Latest reply: Feb 10, 2010 4:09 AM by g777 Go to original post RSS
      • 45. Re: Is ORACLE slower on Windows then on Linux ?
        jgarry
        Billy  Verreynne  wrote:
        jgarry wrote:
        It's because all the people without jobs can get their hands on is Linux.

        Linux is popular because it's available to the great unwashed, that's the only reason . . .
        Well, I may not agree with other things he says or his manner of saying them, but I don't agree that this is bigoted, and I believe to a large extent he is correct.
        Of course it is bigoted... what does [+great unwashed+| http://en.wiktionary.org/wiki/the_great_unwashed] implies? A contemptuous term for the populace, particularly the working class.

        An elitist and egotistical and discriminatory attitude against the so-called "+unwashed+" that do not have the same opportunities, same access to so-called "real" Unix, cannot afford to attend Harvard/Stanford/whatever, etc...

        And I take great exception to a bigotry and discriminatory view like that.
        Elitist is not the same as bigoted. I agree that it is a terrible attitude. We can perhaps agree to disagree as to whether his usage here is bigoted or not. I'm just saying he's right, it's popular because it's available. I personally was a big fan of linux for a long time partly because of its egalitarian nature. I'm still a fan of usenet for that same reason. I have mixed feelings about the Oakies for that same reason, even though I strongly agree with their technical approach (and congratulate all the authors on the Practices book about to come out). Yes, I'm saying even a merit-based access to an elite can be elitist, it depends how non-clique persons are excluded.

        (And I wish I had time to pursue your other post about the actual technical arguments. If you want to post or post links to explanations, feel free. I would like to know which kernel developers are Oracle experts.)
        • 46. Re: Is ORACLE slower on Windows then on Linux ?
          Jonathan Lewis
          Niall Litchfield wrote:

          The tests that I and others have done (*in the 2.4 kernel timeframe*) showed that workload based tests (as opposed to Roby's operation based tests) gave windows about a 5-15% advantage when properly configured. It's high time these tests were re done though.
          Coincidentally, there's [+*this thread*+|http://forums.oracle.com/forums/thread.jspa?threadID=980410&tstart=0] about a system slowing down after moving from Windows to Unix (HP-UX in this case, rather than Linux).

          Familiarity with "why" and "how" is probably far more important than the label on the packet. Anyone can make a mess of a good product and then blame the product. Fewer people are able to explain why the available products differ, and how best to compare the costs and benefits of two products.

          Regards
          Jonathan Lewis
          http://jonathanlewis.wordpress.com
          http://www.jlcomp.demon.co.uk

          To post code, statspack/AWR report, execution plans or trace files, start and end the section with the tag {noformat}
          {noformat} (lowercase, curly brackets, no spaces) so that the text appears in 
          fixed format
          .
          
          "Science is more than a body of knowledge; it is a way of thinking" 
          Carl Sagan                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            
          • 47. Re: Is ORACLE slower on Windows then on Linux ?
            jgarry
            Billy  Verreynne  wrote:
            jgarry wrote:

            I'm not even going to address the memory management mess, except to note, in linux it has changed rapidly and still hasn't been gotten right for Oracle databases. Figuring that out is an exercise for the student. Have fun going up and down on the elevator in the swampiness, I mean, swapiness.
            Very vague. Would like to see some hard technical fact and evidence to back that up that Linux memory management is inferior and not suited for databases. BTW, I run Oracle on Solaris, HP-UX and Linux in production.
            OK, so not really what we were talking about, but go here and look at the entry for Fri Jun 26 20:27:40 2009. [Translation.|http://www.theregister.co.uk/2009/11/03/linux_kernel_vulnerability/]
            • 48. Re: Is ORACLE slower on Windows then on Linux ?
              Billy~Verreynne
              jgarry wrote:

              OK, so not really what we were talking about, but go here and look at the entry for Fri Jun 26 20:27:40 2009. [Translation.|http://www.theregister.co.uk/2009/11/03/linux_kernel_vulnerability/]
              Vulnerabilities exist in all operating systems. So what is the real issue?

              It comes down to managing risk. Is your kernel exposed to attack? All my database servers run in fairly secure environments. Some so deep in the network core, that even I do not have access to them - and have to tunnel via 2 shh hops to get to them. With numerous firewalls protecting them. And this it internal - not even exposed to the Internet.

              So is such a kernel vulnerability an issue? Not in my case. And I would argue not in the case of most Oracle servers as these are usually not exposed to the outside. The app or web server is. And if any sensible security is implemented, there will be a firewall in between. If security is a serious consideration then those web/app servers will in turn be protected by reverse proxies for example.

              A lot of these kernel vulnerabilities are just sensationalism. Sure, they are serious. But in the end it comes down to how you manage your risks and handle your security, You can have a invulnerable kernel, screw up with a weak encryption method and expose all your accounts to attacks via ssh.

              The media seems to confuse real vulnerability issues (like those affecting million of desktops and have resulted in huge botnets) with those that are less serious (if managed correctly).
              • 49. Re: Is ORACLE slower on Windows then on Linux ?
                jgarry
                Billy  Verreynne  wrote:
                jgarry wrote:

                OK, so not really what we were talking about, but go here and look at the entry for Fri Jun 26 20:27:40 2009. [Translation.|http://www.theregister.co.uk/2009/11/03/linux_kernel_vulnerability/]
                Vulnerabilities exist in all operating systems. So what is the real issue?
                The issue is the size of the surface vulnerable to attack. Giving any psycho perennial CS student access to everything for free may obviate any advantage of many eyeballs looking for cracks.

                >
                It comes down to managing risk. Is your kernel exposed to attack? All my database servers run in fairly secure environments. Some so deep in the network core, that even I do not have access to them - and have to tunnel via 2 shh hops to get to them. With numerous firewalls protecting them. And this it internal - not even exposed to the Internet.

                So is such a kernel vulnerability an issue? Not in my case. And I would argue not in the case of most Oracle servers as these are usually not exposed to the outside. The app or web server is. And if any sensible security is implemented, there will be a firewall in between. If security is a serious consideration then those web/app servers will in turn be protected by reverse proxies for example.
                Too bad we don't have any actual metrics on outside exposure. I would like to think so, but don't have any idea if XE is actually used inappropriately or people put too much faith in firewalls and obscurity. There are certainly security folks who say the worst.

                >
                A lot of these kernel vulnerabilities are just sensationalism. Sure, they are serious. But in the end it comes down to how you manage your risks and handle your security, You can have a invulnerable kernel, screw up with a weak encryption method and expose all your accounts to attacks via ssh.

                The media seems to confuse real vulnerability issues (like those affecting million of desktops and have resulted in huge botnets) with those that are less serious (if managed correctly).
                You are, of course, perfectly correct. Personally, I have a very dim view of how well most people evaluate risk.
                • 50. Re: Is ORACLE slower on Windows then on Linux ?
                  Billy~Verreynne
                  jgarry wrote:

                  Too bad we don't have any actual metrics on outside exposure. I would like to think so, but don't have any idea if XE is actually used inappropriately or people put too much faith in firewalls and obscurity. There are certainly security folks who say the worst
                  Well, it is their job to say that. And they always see the worse given the job description. So they tend to have a different perception than non-security folk have. Guess we'll be a lot more paranoid about security if we were in those shoes.
                  You are, of course, perfectly correct. Personally, I have a very dim view of how well most people evaluate risk.
                  Agree. I still like to firewall our servers anyway (despite the IT firewalls). Use ssh heavily (no passwords used - certificates and trusted keys only). But security is a full time job and one cannot simply harden a server or an Oracle database and leave it at that. So how much time do you spend on those aspects and how time on your "real" job? Not an easy balancing act.
                  • 51. Re: Is ORACLE slower on Windows then on Linux ?
                    MarcinP
                    Hi,

                    After 10 years with Oracle on various platforms I can say now it depends.
                    Of course there are differences and limitation on both side and rest of that depends on OS admin skills and willing to work.

                    Personally I prefer Linux because it have more tracing possibility (strace) and more build in tool to search or filer logs (grep, awk).
                    Of course most of them has Windows version too.

                    One of big Windows disadvantage is that some process which are not related to Oracle opens a Oracle DLL and cause patching/upgrade problems.
                    Most odd one was a WINLOGON.EXE service which had open handler to ociw32.dll ;) Try to kill that process to perform a upgrade ;)

                    For other side Linux has some problems too - especially when we are working with not simple solutions.
                    I had a lot of problem with poor working FC cards and changing of Ethernet interface order in 4 port Intel card (after every reboot eth0 become ex. eth1 and then eth4).

                    regards,
                    Marcin Przepiorowski
                    http://oracleprof.blogspot.com/
                    • 52. Re: Is ORACLE slower on Windows then on Linux ?
                      Billy~Verreynne
                      Marcin Przepiorowski wrote:

                      After 10 years with Oracle on various platforms I can say now it depends.
                      Of course there are differences and limitation on both side and rest of that depends on OS admin skills and willing to work.
                      Exactly.
                      One of big Windows disadvantage is that some process which are not related to Oracle opens a Oracle DLL and cause patching/upgrade problems.
                      Most odd one was a WINLOGON.EXE service which had open handler to ociw32.dll ;) Try to kill that process to perform a upgrade ;)
                      This can happen on Linux too - another non-Oracle process using a shared object library file that needs to be patched during an upgrade.

                      The winlogon one is indeed odd however. How sure are you that it was diagnosed correctly as winlogon has no reason to load an OCI driver? Perhaps it was a dependency issue where another DLL loaded by winlogon did a load lib of that DLL.
                      For other side Linux has some problems too - especially when we are working with not simple solutions.
                      I had a lot of problem with poor working FC cards and changing of Ethernet interface order in 4 port Intel card (after every reboot eth0 become ex. eth1 and then eth4).
                      This happens due to using inferior firmware and hardware in my experience. Some h/w vendors attempt to chase the bottom line too hard and use inferior parts in certain server models. We actually had an experience like this earlier this year - and bluntly told the vendor to remove their cr@p from our data centre. And judging from a google we did, we were not the only ones. Several blogposts (sysadmins) also slammed this specific model harshly.

                      Unfortunately, some will not look beyond the o/s and then blame it for being at fault. Which is why both Linux and Windows become the receiving end in threads like these. A pity as ignorance of operating systems turns into a bias against certain operating systems.
                      • 53. Re: Is ORACLE slower on Windows then on Linux ?
                        MarcinP
                        Billy  Verreynne  wrote:
                        One of big Windows disadvantage is that some process which are not related to Oracle opens a Oracle DLL and cause patching/upgrade problems.
                        Most odd one was a WINLOGON.EXE service which had open handler to ociw32.dll ;) Try to kill that process to perform a upgrade ;)
                        This can happen on Linux too - another non-Oracle process using a shared object library file that needs to be patched during an upgrade.

                        The winlogon one is indeed odd however. How sure are you that it was diagnosed correctly as winlogon has no reason to load an OCI driver? Perhaps it was a dependency issue where another DLL loaded by winlogon did a load lib of that DLL.
                        Hi,

                        I have used a process explorer (http://technet.microsoft.com/en-us/sysinternals/bb896653.aspx) as a diagnostic tool and I have used Find Handler functionality with Oracle Home as a searching string.
                        Yes you right it could be dependency issue between DLL's but anyway winlogon.exe has that file open and only one way to solve it was to hide a partition with Oracle Home and reboot a server. It started complaining about some missing files but I was able to set a drive letter to Oracle partition and patch it.
                        It had happen on Windows 2000 server running in windows domain (not a domain controller) and Oracle 9i.

                        regards,
                        Marcin Przepiorowski
                        http://oracleprof.blogspot.com/
                        • 54. Re: Is ORACLE slower on Windows then on Linux ?
                          g777
                          hi,

                          I'm for Linux mostly due to flexibility and available tools to tune. But let's look at this test:
                          [http://www.scribd.com/doc/3713222/Comparison-Oracle-Windows-Linux]
                          1 2 3 4 Previous Next