I am having this same issue. Twice a day at 2p ET and 2a ET I receive a series of alerts indicating both "SNMP Agent shut down message" and "SNMP Agent cold start message" from ALL of our iDRAC 7s. These alerts do not come from any other systems. I've tried everything I can to somehow filter out these alerts at the source (on the iDRAC by disabling various checkboxes under "SNMP Trap" column on the iDRAC "Alerts" page but I can't figure out how to get these to stop.
My big problem here is that I cannot filter these out on OME side. Our organization actually depends on the "SNMP Agent cold start message" traps as a way to [potentially] identify server reboots (yes I know that this message does not always indicate a server reboot but on the other hand a server reboot ALWAYS generates this trap).
Having the iDRACs sending out SNMP Agent cold start messages twice a day causes a lot of noise in our monitoring/alerting system that we don't need to see. And I can't filter these out based on device since an iDRAC and a server are consolidated into one single device in OME (so for example I can't filter out the alerts from only the iDRACs without also filtering them out from our ESXi hosts).
Please let me know if anyone has found a solution to this or if some additional control for this issue can be built into either the iDRAC or OME.