Hi all,

I have a HP-UX Server with 4 gigabytes of physical RAM. When I use the 'Glance' utility to see what my memory utilization is, my memory usage shows up maxed out at 99%. I shut off all the known processes that I'm running on that box and the memory utilization is still at 78% (with Swap Space Util at 45%). Also, the buffer cache is at 1.5 GB usage.

I am wondering if there is a tool out there that will let me check exactly WHICH PROCESSES is using so much memory. I need the memory allocation breakdown by user processes. (does 'top' suffice?)

What is the buffer cache size? Is that a manual setting in the kernel setting?

thanks for any suggestions!

have you tried ipcs?

# ipcs -mop
IPC status from /dev/kmem as of Thu May 3 08:38:31 2007
T ID KEY MODE OWNER GROUP NATTCH CPID LPID
Shared Memory:
m 0 0x411829da --rw-rw-rw- root root 0 735 735
m 1 0x4e0c0002 --rw-rw-rw- root root 1 735 735
m 2 0x4124028d --rw-rw-rw- root root 1 735 747
m 19971 0x0c6629c9 --rw-r----- root root 2 1514 1512
m 4 0x06347849 --rw-rw-rw- root root 1 1514 1539
m 1541 0xffffffff --rw-r--rw- root root 0 1512 1512
m 45062 0x5e140024 --rw------- root root 1 1759 1759
m 519 0x30dfc4ec --rw-r----- oracle dba 61 5420 7453
m 8 0x9273c4c0 --rw-r----- oracle dba 58 5560 7462



go get kmeminfo from HP

# ./kmeminfo
tool: kmeminfo 7.02
unix: /stand/vmunix 11.11 64bit PA2.0 on "my.box.com"
core: /dev/kmem live
link: Wed Feb 1 07:45:48 PST 2006
boot: Sat Apr 28 17:36:56 2007
time: Thu May 3 08:37:33 2007
nbpg: 4096 bytes


----------------------------------------------------------------------
Physical memory usage summary (in page/byte/percent):

Physical memory = 5242880 20.0g 100%
Free memory = 537474 2.1g 10%
User processes = 4008776 15.3g 76% details with -user
System = 683688 2.6g 13%
Kernel = 421544 1.6g 8% kernel text and data
Dynamic Arenas = 96880 378.4m 2% details with -arena
M_TEMP = 43860 171.3m 1%
VFD_BT_NODE = 15210 59.4m 0%
M_SPINLOCK = 7245 28.3m 0%
KMEM_ALLOC = 5432 21.2m 0%
ALLOCB_MBLK_LM = 2813 11.0m 0%
Other arenas = 22320 87.2m 0% details with -arena
Super page pool = 7021 27.4m 0% details with -kas
Static Tables = 263990 1.0g 5% details with -static
pfdat = 119481 466.7m 2%
htbl2_0 = 65536 256.0m 1%
nbuf = 21024 82.1m 0% bufcache headers
bufhash = 20480 80.0m 0% bufcache hash headers
pfn_to_virt = 19913 77.8m 0%
Other tables = 17555 68.6m 0% details with -static
Buffer cache = 262144 1.0g 5% details with -bufcache
UFC file mrg = 0 0.0b 0%
UFC meta mrg = 0 0.0b 0%
#

current top info:

System: mybox Thu May 3 08:39:30 2007
Load averages: 0.09, 0.07, 0.07
250 processes: 210 sleeping, 39 running
Cpu states:
CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
0 0.12 5.9% 0.0% 0.0% 94.1% 0.0% 0.0% 0.0% 0.0%
1 0.11 13.9% 0.0% 5.9% 80.2% 0.0% 0.0% 0.0% 0.0%
2 0.02 18.8% 0.0% 1.0% 80.2% 0.0% 0.0% 0.0% 0.0%
3 0.11 4.0% 0.0% 2.0% 94.1% 0.0% 0.0% 0.0% 0.0%
4 0.06 3.0% 0.0% 1.0% 96.0% 0.0% 0.0% 0.0% 0.0%
5 0.16 8.9% 0.0% 5.9% 85.1% 0.0% 0.0% 0.0% 0.0%
6 0.06 3.0% 0.0% 2.0% 95.0% 0.0% 0.0% 0.0% 0.0%
7 0.10 7.9% 0.0% 2.0% 90.1% 0.0% 0.0% 0.0% 0.0%
--- ---- ----- ----- ----- ----- ----- ----- ----- -----
avg 0.09 7.8% 0.0% 2.9% 89.2% 0.0% 0.0% 0.0% 0.0%

Memory: 16015140K (13469144K) real, 24703812K (22087940K) virtual, 2180160K free Page# 1/6

I don't know if this works on HP but I use these on AIX to find "top hogs":
ps aux | tail +2 | sort -n +2 -r | head ; Top cpu hogs
ps aux | tail +2 | sort -n +3 -r | head ; Top memory hogs

Are u running oracle , there must be a memory leak in some applications ;
try this commeand in a cron "UNIX95= ps -e -o comm,vsz,etime,stime,time" and append the information to a log file and check which process is eating up your memory .

Check the kernel parameter settings swapmem_on ...


cheers