Same problem on Windows 2003 R2 32-bit, ML150 G5.
About every one-two minutes the Foundation Agents reports some of the services that are running locally (review the Data section of the event in the event log), then reports that they are unreachable. This cycle continues indefinitely.
Here is some of the data I have pulled from the Event log entry present when I was receieving Email notices stating an snmp trap was received with a Details stating NOT AVAILABLE :
#Domain Name Server.domain 53/udp
#Domain Name Server.bootps 67/udp
dhcps
#Bootstrap Protocol Server.bootpc 68/udp
dhcpc
#Bootstrap Protocol Client.tftp 69/udp
#Trivial File Transfer.gopher 70/tcp
.finger 79/tcp
.http 80/tcp
www www-http
#World Wide Web.hosts2-ns 81/tcp
#HOSTS2 NameServer.hosts2-ns 81/udp
#HOSTS2 NameServer.kerberos 88/tcp
krb5 kerberos-sec
#Kerberos.kerberos 88/udp
krb5 kerberos-sec
#Kerberos.hostname 101/tcp
hostnames
#NIC Host Name Server.iso-tsap 102/tcp
#ISO-TSAP Class0.rtelnet 107/tcp
#Remote Telnet Service.pop2 109/tcp
postoffice
#PostOffice Protocol-Version2.pop3 110/tcp
#Post Office Protocol -Version3.sunrpc 111/tcp
rpcbind portmap
#SUN Remote Procedure Call.sunrpc 111/udp
rpcbind portmap
#SUN R...WV
CPQTHRSH.HOSTALRT
I do not have many of these services running on the local machine.
When I received an alert email with no Details listed, the Foundation Agent event log event data in the System Event Log listed four processes that were and are running on the machine.
Reviewing: %systemroot%\system32\CpqMgmt\cqmghost\cpqmgmt.chm it appears that the Foundation Agent provides things such as process and service monitoring. You may be able to disable these two features and the problem may or may not go away (but this is what I'm looking to do).
After some investigation with process monitor, I see that the log entries correlate with cqmghost.exe (the Foundation Agents service) accessing two keys:
HKLM\SYSTEM\CompaqPermanent\
and
HKLM\System\CompaqHost\
In my monitoring, both of these seem to be accessed only by cqmghost.exe.
These appear to be a cache of some sort.
I wonder if some sort of issue exists when upgrading directly, without uninstallation first.
I will uninstall the HP Insight Management Agents, delete these keys, then reinstall later.
[update 1]
With the agents uninstalled, and the those keys deleted, the web UI is still populating without issue (since it uses WBEM or SNMP, in my case WBEM) [tested by reviewing sectors read and written on a RAID disk].
cp015760.exe is the file to install the component "HP Insight Management Agents for Windows Server 2003/2008" (taken from %systemdrive%\cpqsystem\hp\log\localhost\hpsum_log.txt)
In the summary of the previous install, I do see that SUM detected the older version of the component:
Component File Name: cp015760.exe
Component Name: HP Insight Management Agents for Windows Server 2003/2008
Original Version: 8.70.0.0
New Version: 9.10.0.0
Installation Result: Success
I ran cp015760.exe solo.
After install a window with manage agents ("C:\WINDOWS\system32\rundll32.exe" Shell32.dll,Control_RunDLL cpqmgmt.cpl /setup) pops up. Registry keys that were deleted are added again.
I unchecked Enable SNMP sets, noting that polling was set to 2 minutes by default... hmm... since I'm using the WBEM (wmi), and click OK.
A 1182 event is logged in the System Event Log.
Four minutes pass. No more 1182 events, because there is no more polling... Enabling polling causes events to be logged.
[Conclusion]
So the problem exists with the SNMP sets that are being polled. You can access the Management Agents configuration control panel applet by ruhning cpqmgmt.cpl or by accessing "HP Management Agents" from the Control Panel itself.
After this select your minimum usage. Since I figured out, by reviewing the Data logged in the 1182 events in the System Event Log, that there was at least one of the "Active Agents" that were causing a problem. I would go through trial and error to figure out which one is causing the issue. You will note that there is also a "Process Monitor" tab present, so this might be how the processes above were listed.
I began tested removing the Active Agents one-by-one. Luckily, Host Information (near the top of the list), seemed to be the problem cause.
I removed the Host Information agent from SNMP polling, waited 8 minutes, and only the single "OK" status change was logged (at three minutes).
I then re-added the Host Information agent, waited 3 minutes, saw the "OK" status change again. Then waited another 3 minutes, received the first status change event (1182, citing CPQTHRSH.HOSTALRT). It appeared that some processes were listed in the Event Data. I then waited another 5 minutes and nothing popped up.
The info is sparse, but the issue is clearly related to HOSTALRT, which is found in both events' Data.
For the sake of testing, I unchecked
[tl;dr]
Through the "HP Management Agents" control panel applet (cpqmgmt.cpl):
1) Try to clear all thresholds on the SNMP Settings tab...
2) If that fails, Try to disable Application Exception Traps...
3) If that fails, remove the Host Information agent from the Active Agents list in the "HP Management Agents" control panel applet. This clearly stops some polling from occurring, but I can not figure out which OIDs the Host Information Agent polls, or which data will be missing.
