1 Reply Latest reply on Aug 17, 2011 10:04 AM by philmodjunk

    which type of find would be faster?

    ijontichy

      Title

      which type of find would be faster?

      Post

      I am scripting a find for records inside a database that has over 25million records.  Each record represents  the metadata from a camera rig recording at 24 frames per second.

       

      I've currently got a mutlple find which says "look for images shot on THIS day and with THIS camera between the time stamps of This time and this time."

       

      But given the volume of records, would it make more sense to do a successive series of contraints?   "Find records from this date.   Now find records from this camera.   Now find the records between these two timestamps."

       

      Any thoughts on which would be quicker, if either?

       

      Thanks in advance.

       

        • 1. Re: which type of find would be faster?
          philmodjunk

          Assuming all fields receiving search criteria are stored, indexed fields, I believe that a single find would be the quickest.

          Successive finds can be much quicker if one or more of the fields are unstored/unindexed. In those cases perform a find with criteria entered only into the indexed fields. Then return to find mode to specify the values in the unindexed fields and constrain the found set.

          Ultimately, you may want to try out both approaches and see which works best...