0 Replies Latest reply: Jun 19, 2013 3:10 PM by EdStevens RSS

    computing effective diskgroup space


      Oracle SE-One

      GI , stand-alone ASM, no RAC

      Oracle Linux 5.6


      I have two disk groups ..


      one based on ssd, has two failure groups, one on the ssd the other on SAN.  Normal redundancy.


      the second purely on the SAN, with external redundancy.


      We are getting to the point of being concerned about maxing out the ssd group and I am looking at identifying some low-use files for movement to the other DG.    I would like to be able to compute the effective percentage of the SSD group that will be freed up and the effective percentage of the san group that will be consumed, given the size of a given file under consideration.


      I'm doing the same exercise for a non-ASM database, but there the calculation is pretty straight forward, given the size of the two file systems in question.  But with the ASM and the effects of redundancy I'm not really sure which metrics I should be looking at.  I know that v$asm_diskgroup.usable_file_mb can actually go negative (though if it does you'd want to address that).  Would it be as simple as 1/2 of v$asm_diskgroup.total_mb for normal redundancy DG and all of v$asm_diskgroup.total_mb for the external redundancy?


      In other words, given:

      DG_SSD has total_mb of 1024  (1 gb)

      DG_SAN has total_mb of 4096  (4 gb)

      file my_ts_01.dbf is 128mb.


      So that 128 mb is effectively taking 25% of DG_SSD  (128 / (1024 / 2 )) .. allowing for normal redundancy on DG_SSD

      and it would be 3.1% of DG_SAN (128 / 4096)


      Or am I way out in left field in my understanding of how redundancy impacts capacity and how ASM computes and reports capacity?