Well if those two commands show that Solaris detected 1GB of RAM then there is nothing else what you can do. Maybe your server was partitioned using LDoms, and that is why amount of RAM is so low in this instance of the system, but I don't have much knowledge about that subject.
You might be onto something there; this box does have LDOMs, and the primary partition is only 1Gb, so that makes sense. I would still expect prtconf/prtdiag to tell me how much physical RAM is in the box, though.
# ldm list
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- SP 2 1G 1.3% 23h 31m
litsun01-d01 active -n---- 5000 4 2G 0.4% 1d 22m
litsun01-d02 active -n---- 5001 4 2G 0.2% 1d 22m
litsun01-d03 active -t---- 5002 4 2G 25% 1d 22m
litsun01-d04 active -n---- 5003 4 2G 0.2% 1d 22m
litsun01-d05 active -t---- 5004 4 2G 25% 19h 41m
litsun01-d06 inactive ------ 4 2G
litsun01-d07 inactive ------ 4 2G
litsun01-d08 inactive ------ 4 2G
Yup, but the script that I have for collecting this info uses the "memory size" line at the moment; LDOM broke it. I need to come up with a new way to do this, which will work on either an LDOM or non-LDOM system.
That's nice and clean, but I don't know if this will work on all systems of this type. I'll have to try this on similar boxes in our other data centers.
The 'lgrpinfo' utility might do it. I don't recall offhand when that was rolled into Solaris 10 so you may not have it, and it may not work under LDOMs, but it's designed to show physical RAM configuration.
