memory leaks ?

[nwhwsup]

Hi, summarization and graphing of SZ column in "ps -elf" shows the table below.
(script attached if you want to measure your own memory leaks).
Memory is measure in pages of size 4K.

This is only for a month period.
This is temip 5.0 and hpux 11.11.

Why does ia64 stand out (isn't that some hw-monitor, that shouldn't use memory)?
How reliable is SZ for process private memory consumption on hpux ?

PID => MEMA MEMB Process
1581 => 1782 1790 /opt/Navisphere/bin/naviagent -f /etc/Navisphere/agent.config
1883 => 119 123 nvnmgr 3
2105 => 102 106 /sbin/sh /usr/dt/bin/dtrc
2145 => 36 40 /etc/opt/resmon/lbin/p_client
2726 => 23721 23741 temip_acs_fm
2803 => 3953 5273 /usr/sbin/stm/uut/bin/tools/monitor/ia64_corehw
2885 => 143 167 /usr/sbin/stm/uut/bin/tools/monitor/sysstat_em
12500 => 70922 88686 mcc_alarms_fm
21089 => 102 106 /sbin/sh
24021 => 2211 2951 /usr/opt/tns/bin/tns_server
25756 => 1098 1102 /usr/opt/temip/orbix/610/asp/6.1/bin/itlocator -ORBname iona_se
26199 => 2991 3019 tt_conn_am
26230 => 8775 8943 temip_alarm_handling_fm
26287 => 5371 5375 oracle_extended_archiving_fm
26428 => 5503 5551 TeMIP AHFM Child NNM_OC
26718 => 59717 59721 tal_orbix_server

[Bryan Strait]

Yes, ia64_corehw is a monitor that monitors power supplies, and fans, among other things. Some older versions of ia64_corehw have contain a memory leak which may be what you're seeing. What version of EMS are you using? Patching to the latest Support Tool Manager should take care of it.

http://www.docs.hp.com/en/diag/stm/stm_upd.htm#whyupdate

Hope this helps.

Bryan