Problem with cont. get next requests to hrSWRun under high CPU load at Solaris

We are working with Net SNMP Version 5.1.2 on Solaris Version 5.8.

Under high CPU load of the machine we are observing the following problem:

- We are monitoring some processes on this Solaris machine using the Host MIB RFC1514.
- To differ the processes we sometimes need to request hrSWRunName, hrSWRunPath and hrSWRunParameters.
- Normally the process index runs to 29680 (instance ID).
- We are using the net-snmp.dll API (TNGSNMP_GETNEXT_REQ_MSG) on a Windows 2003 system with net-snmp V.5.4.0-1.win32   to get the MIB information.
- Normally to the following requests
  * GetNext for 1.3.6.1.2.1.25.4.2.1.2.0 cont.
  * GetNext for 1.3.6.1.2.1.25.4.2.1.4.0 cont.
  * GetNext for 1.3.6.1.2.1.25.4.2.1.5.0 cont.
  a) we get replies from 1.3.6.1.2.1.25.4.2.1.x.1  to  1.3.6.1.2.1.25.4.2.1.x.29680
  b) followed by 1.3.6.1.2.1.25.4.2.1.x+1.0
- In cases of high CPU load of the monitored system (and only in this cases)
  c) we get replies from 1.3.6.1.2.1.25.4.2.1.x.1  to  1.3.6.1.2.1.25.4.2.1.x.3680 (or any other very small instance ID)
     (appears for hrSWRunName and/ or hrSWRunPath and/ or hrSWRunParameters)
  d) followed by 1.3.6.1.2.1.25.4.2.1.x+1.0

Any idea why in case of high CPU load a continous get-next request to the hrSWRun OIDs finishes soon as existing processes?

Thank You for help.

Sven Koos.


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Net-snmp-users mailing list
[email protected]
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users