I am looking for a better way to alarm on ports that are deemed critical without seeing "BAD LINK DETECTED" alerts for ports we do not care about. The previous admin (no longer here) stopped the BAD LINK DETECTED alerts by configuring the alert to None severity in Event Configuration.
Today, I was requested to send alerts on six specific ports (3 links) that were deemed absolutely critical. When I set the BAD LINK DETECTED event back to critical, the Spectrum event window was filled with hundreds of BAD LINK DETECTED alerts that we do not care about.
Is there a recommended solution to disable the BAD LINK DETECTED alarms for everything but the ones we deem most critical? Any pro tips out there?
Take a look at the following knowledge document.
Port/Interface alarming in CA Spectrum when a port/interface model goes down
Per this document, I am assuming I can take all the current BAD LINK DETECTED alarms and set the value to False and that will prevent these from occurring again?
Attribute ID: 0x12a54
Are you available to discuss this further over the phone?
Correct. Setting GeneratePortStatusAlarms to False on the interface model will prevent the BAD LINK DETECTED alarm from reoccurring for that interface model.
Is there a way to have this set as default and then we only alarm on the ports we care about by setting it as a live pipe?
IT Analyst, Level 2
Duke Health Technology Systems | Duke University Health System
Hock I Plaza, 10032
Visit the DHTS website<http://dhts.duke.edu/> to learn about Duke Health IT services, order products and services and to report issues.
Confidentiality Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.
If I made a global collection of all ports (excluding the six ports I was concerned with) and set the following attributes on all ports using attribute editor:
GeneratePortStatusAlarms – no
Alarm On LINK Down Trap – no
Poll Port Status – no
This would stop the “BAD LINK DETECTED” against all ports except for the six we want the notification. Would this cause any other issue outside of the port polling that you can think of?
No other issue that I am aware of.
This is a perfect use case for Policy Manager, as Joe's KB article suggests. We created dynamic global collections by device type to collect interfaces where we do care about bad link alarms. For instance, we have a collection for switch uplink ports, a collection for all other switch ports, a collection for admin down router ports, and a collection for admin up router ports. We then used Policy Manager to automatically set each of the attributes in that KB article differently for each collection. Switch ports have everything disabled, but switch uplinks have live links enabled. Admin down router ports have everything but traps disabled, and admin up router ports have everything enabled. It took a little while to set up, but it has been extremely useful.