opsmode

Today when installing OpsMgr R2 agents on a couple of Solaris servers we ran in to a problem on the Solaris 9 servers.

The error we got in OpsMgr:
<![CDATA[Executing command: echo -e mail=\ninstance=overwrite\npartial=nocheck\nidepend=quit\nrdepend=quit\nconflict=nocheck\naction=nocheck\nbasedir=default\n > /tmp/scx-$USER/scx;uncompress -f /tmp/scx-om_unix/scx-1.0.4-248.solaris.9.sparc.pkg.Z

We found this on the Solaris box:
[ verifying class <config> ]
## Executing postinstall script.
ld.so.1: .scxsslconfig: fatal: libssl.so.0.9.7: open failed: No such file or directory
Killed
pkgadd: ERROR: postinstall script did not complete successfully
Installation of <MSFTscx> failed.
bash-2.05# ssh –version
Sun_SSH_1.1, SSH protocols 1.5/2.0, OpenSSL 0x0090700f
Bad escape character ‘rsion’.

And if you look in Appendix A in the Operations Manager 2007 R2 Operations Guide you se that OpenSSL is required. So after installing OpenSSL the agent install went fine!
Sun does not provide a version of OpenSSL for Solaris 9 SPARC. There is a version available from Sunfreeware.

Based on the data you get from the cmdlet get-agent you can list all agents with a specific setting.

This line for example, will list all agents with the Agent Proxy setting enabled:
get-agent | where {$_.ProxyingEnabled -like $true}

Change $_.ProxyingEnabled to $_.ManuallyInstalled and you’ll get a list with all manually installed agents.

In OM2007 SP1 there is a bug on updating the ManuallyInstalled property, so that wont work. This is corrected in R2.
If you dont have to use powershell there is a workaround in this thread.

 
Your options are:
PrimaryManagementServerName
Id
LastModified
Name
DisplayName
HostComputer
HostedHealthService
HealthState
PrincipalName
NetworkName
ComputerName
Domain
IPAddress
Version
RequestCompression
CommunicationPort
MaximumSizeOfAllTransferredFilesBytes
MaximumQueueSizeBytes
ManuallyInstalled
InstallTime
InstalledBy
CreateListener
AuthenticationName
ActionAccountIdentity
HeartbeatInterval
ProxyingEnabled
ManagementGroup
ManagementGroupId

I have been trying to fix an issue with a SQL 2008 cluster node restarting the Health Service every 10 minutes. Logging EventID: 6024 with the description:
RestartHealthService.js : Restarting Health Service. Health Service exceeded Process\Private Bytes threshhold.

It didnt mater if I did an override on the monitor Health Service Private Bytes Threshold. I got the event anyway

Then I realized that this is a multi-homed agent, so when I did an override on the same monitor but in the other management group as well, the problem was resolved.