With the new agent group setup in PCM3, it looks like messages from the network end up in arbitrary lists. I can select the "Devices" group and get the events from my devices. When I've dealt with those, I delete them, but then other events from switches are listed in, say, the default agent group.
Am I missing the point here? What's the logic behind?
I'd like all events from my switches to be listed under "Devices".
Most of the time I have selected "Devices" in the navigation pane of PCM+ and the screen shows the dashboard. When events occur in the network, they are shown in the Events view of the dashboard (with the columns), and I expect to be able to click this view see the events list. I am taken to the list allright, but the events are not there. Instead I have to navigate to "Default Agent Group" to see the events in the list.
The "Default Agent Group" can include multiple agents, and the Dashboard and Events tabs at this level of the PCM tree show all events that result from SNMP traps received by those agents. At the "Devices" level of the tree, which is a subnode of each agent group including the Default Agent Group, the content reflected in the Dashboard and Events tabs is events that result from SNMP traps received from devices that the agents have *discovered*. So if you see a discrepancy between what's shown at these two levels it is most likely due to the fact that SNMP traps from undiscovered devices will be reflected at the agent group level but not at the Devices level.
However, the data at each level in the tree should be consistent between the corresponding Dashboard and Events tabs. Some quick testing to see if your situation could be replicated with post-3.0 PCM code didn't indicate any obvious navigational bugs that could explain it but did reveal one other possible explanation: the Events graph on the dashboard refreshes its counts every 15 seconds whereas the Events tab itself refreshes its content every 30 seconds. So you can see a discrepancy between the two, where events are reflected on the Dashboard but not the Events tab, during every other refresh of the Dashboard but they should then agree on the following 15-second boundary when both refresh. Is it possible that you're so quick to respond to the appearance of events that this is the reason for the discrepancy you're seeing?
Good, glad to hear there's an explanation even if the behavior is not exactly what you desire. I will pass the feedback that the two periods should be the same to the appropriate PCM developers, and hopefully you'll see that change in an upcoming PCM autoupdate or new release before too long.
