Adding quorum disk causing wasted space?

Question

Hi,
Any idea whether this is a bug or an expected behavior?
Seeing this with ASM 11.2.0.4 and 12.1.0.4
Have a Normal redundancy disk group (FLASHDG22G below) with two disks of equal size. With no data on the disk group the Usable_file_MB is equal to the size of one disk, as expected.
But if I add a small quorum disk to the disk group, the Usable_file_MB decreases to 1/2 of the disk size. So, half of the capacity is lost.
Thoughts?
[grid@symcrac3 ~]$ asmcmd lsdsk -k
Total_MB  Free_MB  OS_MB  Name        Failgroup  Failgroup_Type  Library  Label  UDID  Product  Redund   Path
   20980    20878  20980  SYMCRAC3_A  FG1        REGULAR         System                         UNKNOWN  /dev/symcrac3-a-22G
     953      951    953  QUORUMDISK  FGQ        QUORUM          System                         UNKNOWN  /dev/symcrac3-a-quorum
   20980    20878  20980  SYMCRAC4_A  FG2        REGULAR         System                         UNKNOWN  /dev/symcrac4-a-22G
[grid@symcrac3 ~]$ asmcmd lsdg
State    Type    Rebal  Sector  Block       AU  Total_MB  Free_MB  Req_mir_free_MB  Usable_file_MB  Offline_disks  Voting_files  Name
MOUNTED  NORMAL  N         512   4096  1048576     42913    42707            20980           10388              0             N  FLASHDG22G/

Answer 1

There are two separate issues:
1) ASMCMD silently fails to add quorum failure groups. Adds them, but as regular failure groups.
2) Even if a quorum failure group is added with SQLPlus, the space is still lost – I have just confirmed it. And it doesn’t matter whether I add a quorum disk to an existing group or create a new group with a quorum disk.
For #2 here is the likely source of the problem. Usable_File_MB = (FREE_MB – REQUIRED_MIRROR_FREE_MB ) / 2.
REQUIRED_MIRROR_FREE_MB is computed as follows (per ASM 12.1 user guide):
–Normal redundancy disk group with more than two failure groups
   The value is the total raw space for all of the disks in the largest failure group. The largest failure group is the one with the largest total raw capacity. For example, if each disk is in its own failure group, then the value would be the size of the largest capacity disk.
Instead, it should be "with more than two regular failure groups".
With just two failure groups it is not possible to restore full redundancy after one of them fails. So, REQUIRED_MIRROR_FREE_MB = 0 in this case.
Also REQUIRED_MIRROR_FREE_MB should remain 0 even when there are three failure groups if one of them is a quorum failure group. But the logic seems to be broken here.