I wonder how my monitoring could be fault tolerant. I have to choose
an agent to be the proxy between my MS and my SNMP device. But I can't
choose more than one. What can I do in order to set my monitoring
fault tolerant ? Can I discover my devices twice ? Thanks for your
help.
Regards,
--
P.A.
Hey there - I actually added a bit to this post with that in mind:
http://pavleck.net/2008/07/07/setting-agent-failover-servers-switching-snmp-device-proxies/
- It's the second half of it. Figured it was a good addition since I
was on the topic anyway.
Essentially, I don't see a method of setting a fail over like you can
for agents, but you can change the proxy agent. So if you had a
failure and needed a new watcher node, you would need to proactively
run something like the below to switch your devices from their current
proxy to a new one. If needed, you could tie this to a monitor on the
RMS/MS that would run this as a Recovery Task if the watcher node went
down - it's flexible there.
# Pick a new agent to use as a proxy for the remotely managed devices,
and grab it's object
$proxyAgent = Get-Agent |? {$_.Name -eq "MYSERVER.CORP.COM"}
# Gather the remote devices
$remDevices = Get-RemotelyManagedDevice
# Loop through devices, changing their proxy agent
ForEach($device in $remDevices) {
Set-ProxyAgent -ProxyAgent $proxyAgent -Device $device | Out-Null
}
# Done!
This is an extremely simple example, but you get the gist of it.
Hrmm, I swore I answered this earlier. In case I didn't, please see my
post at http://pavleck.net/2008/07/07/setting-agent-failover-servers-switching-snmp-device-proxies/
- the bottom half has to do with network devices and changing the
device proxy.
Thanks !
--
P.A.