Solaris 10 memory confusion

I have a Solaris 10 zone where I am having issues with Oracle, and I'm
somewhat confused about memory. The domain which contains the zone had 128GB
of RAM allocated, and the zone doesn't have a memory cap. However, it was
unable to start a further database instance.

After giving the domain a further 64GB of memory, the additional instance
could be started. Here's some information from the domain:

#  echo "::memstat" | mdb -k
Page Summary                Pages                MB  %Tot
------------     ----------------  ----------------  ----
Kernel                     646622              5051    3%
ZFS File Data              664314              5189    3%
Anon                      8335731             65122   34%
Exec and libs              199163              1555    1%
Page cache                1048136              8188    4%
Free (cachelist)          6543435             51120   26%
Free (freelist)           7440098             58125   30%

Total                    24877499            194355
Physical                 24831553            193996

# swap -l
swapfile             dev  swaplo blocks   free
/dev/zvol/dsk/rpool/swap 256,4      16 50331632 50331632

# swap -s
total: 66660144k bytes allocated + 3828760k reserved = 70488904k used,
68318968k available

# prtconf |grep Memory
Memory size: 196608 Megabytes

# vmstat 2 3
 kthr      memory            page            disk          faults      cpu
 r b w   swap  free  re  mf pi po fr de sr vc vc vc vc   in   sy   cs us sy
id
 0 0 0 13531088 53922272 114 497 179 0 0 0 0 3 3 35 35 11805 15398 13703 0 0
99
 0 0 0 68255128 111848016 3 320 0 0 0 0  0  0  0 43 43 19251 38258 22457 1 0
99
 0 0 0 68254296 111847240 0 226 0 0 0 0  0 72 72 31 31 17895 56356 20629 1 0
99

# prstat -t 1 1

 NPROC USERNAME  SWAP   RSS MEMORY      TIME  CPU
   932 oracle     65G   63G    33% 189:27:34 0.5%
   139 root     1174M 1185M   0.6%  25:15:12 0.0%
    25 daemon     40M   39M   0.0%  17:51:50 0.0%
     4 smmsp      12M   18M   0.0%   0:01:36 0.0%
     4 noaccess  568M  455M   0.2%   2:49:26 0.0%

So, now the questions:

1) This box has 216GB of virtual memory - 192GB of memory + 24GB of swap.
Where can I see that reflected?

2) User oracle is consuming about 65G of virtual memory now, after the extra
instance has been started. As this is much less than the memory available
before the extra 64GB was allocated, why wouldn't the extra instance start
before?




Kindest regards,



Niall  O Broin