8 Replies Latest reply on Dec 3, 2015 6:34 AM by Beauty_and_dBest

    NODELAY vs REAL TIME

    Beauty_and_dBest

      Hi ALL,

       

      EBS R12.2.4

      11gR2

      Rhel6.5

       

       

      I setup physical standby for our ebs.

      I put it in "real time apply" mode using the command:

      ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT;

       

      My question is, is this the same as:

      ALTER DATABASE RECOVER MANAGED STANDBY DATABASE NODELAY;


      Or which of the two is better to use?


       

      Thanks a lot,

      jc

        • 1. Re: NODELAY vs REAL TIME
          Pravin Takpire

          hi,

          both are not same

           

          ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT;

          Here is redo changes are applied to Standby Redo Logs also and are applied to Standby Database. Here DR does not wait for archiving to be completed on primary and being shipped to Standby Site

           

          ALTER DATABASE RECOVER MANAGED STANDBY DATABASE NODELAY;

          This command is nothing but make sure that there is no delay in applying archived redo log to standby database. This is normal standby configuration command where archived logs are shipped to Standby Database and are applied. Above command just specifies that once shipped it should be applied immediately.

           

          I personally prefer second architecture rather than real time apply. But it all depends on Infrastructure you have, kind of budget you have and what is business requirement.

          I would suggest you read

           

          http://docs.oracle.com/cd/B28359_01/server.111/b28294/log_apply.htm. Check the Chapter 7 Figure 7.1

           

           

          regards

          Pravin

          1 person found this helpful
          • 2. Re: NODELAY vs REAL TIME
            Beauty_and_dBest

            Thanks Pravin,

             

             

            If I use the 2nd, can I still run reports at the standby server and the output is reliable?

            • 3. Re: NODELAY vs REAL TIME
              Pravin Takpire

              yes it is reliable

              • 4. Re: NODELAY vs REAL TIME
                Beauty_and_dBest

                I am still confused,

                 

                What if my redo log is 1Gb size and  supposing it is not filled up for 1 day, how can it switch the logfiles and do transport?

                How does the NODELAY get affected? Will it still apply logs which are not yet transported? Will our reports run at standby be not synced with the primary?

                 

                 

                Thanks,

                • 5. Re: NODELAY vs REAL TIME
                  Pravin Takpire

                  If your database is not generating 1GB or archive log for a day then I don't think so you need Stanby database.

                  Reduce the size of redo log to 100 MB.

                  In Normal standby architecture, archived are applied only when shipped to standby.

                  If you have good network infrastructure then you can use real apply also

                  regards

                  Pravin

                  1 person found this helpful
                  • 6. Re: NODELAY vs REAL TIME
                    Beauty_and_dBest

                    Thanks Pravin,

                     

                    But it is true that I can force the transport even if my redo log is not filled up by setting ARCHIVE_LAG_TARGET = 15 minute?

                    Then I can combine it with  REAL TIME?

                     

                     

                    Kind regards,

                    • 7. Re: NODELAY vs REAL TIME
                      Pravin Takpire

                      Hi

                       

                      ARCHIVE_LAG_TARGET limits the amount of data that can be lost and effectively increases the availability of the standby database by forcing a log switch after the specified amount of time elapses.

                      A 0 value disables the time-based thread advance feature; otherwise, the value represents the number of seconds. Values larger than 7200 seconds are not of much use in maintaining a reasonable lag in the standby database. The typical, or recommended value is 1800 (30 minutes). Extremely low values can result in frequent log switches, which could degrade performance; such values can also make the archiver process too busy to archive the continuously generated logs.

                       

                      You don't need to combine this with REAL TIME APPLY. I would suggest you set value of this parameter to 900 (Seconds). Log swtich will happen every 15 mins, archive will be generated and shipped and applied to Standby.

                       

                      regards

                      Pravin

                      1 person found this helpful
                      • 8. Re: NODELAY vs REAL TIME
                        Beauty_and_dBest

                        Thanks you very very very much Pravin.

                        I  understand your explanation better than the documentations