tcpip$hr_mib process paging file quota slow down

[FELIX WONG_1]

Hi,

I've had an issue with TCPIP process.
My system is currently running OVMS v7.3-2 with tcpip v5.4 ECO5 on an alpha server es47 7/1150 model.

The paging file quota of tcpip$hr_mib process
launched by user "tcpip$snmp" was found reducing slowly. It has an inital value of 56,000 after system reboot. However this value reduces continuously and eventually bombed after a period of time (within 2 weeks).

I've tried to increase the pgflquota of tcpip$snmp account from default 10240 to the current value 60000, it doesn't help.
Why the process never re-gain the paging file quota and what's the fix on this issue !?

[Hoff]

That the available pagefile quota for the MIB process is slowly being depleted is usually indicative of a memory leak somewhere within the associated application or its underlying libraries.

Increasing the value of or the limit of the particular resource associated with the leak typically only forestalls the occurrence of the failure; such increases cannot and do not eliminate the leak itself.

If ECO5 is not current for V5.4 (and I haven't looked at that), then apply the current ECO and see if you can reproduce the leak. If you are current on your ECO, your recourse here is to periodically restart the MIB and/or to contact HP support with a formal problem report.

[labadie_1]

A workaround is to call $purgws regularly for that process. Be prepared to a bigger pagefile utilization, by adding pagefile space.

You can find various tools doing that, or write your own code.

Check the freeware for similar tools

[Jur van der Burg]

Calling $purgws is nonsense as that will only move the pages out of the workingset. The real problem is a memory leak as Hoff said.

Jur.