Dear All -
we're facing a serious problem with CPI 3.0.3 (but the same situation happened with 3.0 before). Symptom is that Cisco Prime randomly stops to send out email alerts.
Normally, we're monitoring hundreds of switches health, but based on the actual monitoring policy, we're focusing now for device reacheability (ICMP ping w/ 5 mins polling) only. Meaning, if one of our switches is going down, Prime need to simply send out an alert to the specified email address.
It was worked fine but after some week, Prime simply stops to send out alerts. System shows the alert on the Alert Dashboard, but not sending out emails about it. To avoid database overflow, we tried to disable every other monitoring piece except up/down data collection, didn't helped. Rebooted the primary device, didn't helped. Deleted all of the active alerts and generate outages at some test switches to see the email, didn't helped.
As a check, we started to shutdown / get back test switches to see how Prime is reacting. Well, Prime noted the activity, shows the alert on the dashboard, but not send out emails. Strange, because it was worked before, and we didn't changed anything on the configuration. As a test, we set alert AutoClears at the end of each day but now, this is does not work. We didn't got alerts from unavailable switches, nor the AutoClear, nothing.
BUT! When I'm indicating a test mail sending from Prime, we get this email. When I restarted the Primary appliance, I got an email from the system. But if a switch is going down, we don't get anything. It seems for me that this is a serious issue somewhere around the indicator of email sending fuction.
Any ideas?
Thanks.
