Event alert Exclusion for NAS cluster in event of maintenance

[Cavin]

Does anyone know effective way to surpress all alerts from list of NAS cluster that scheduled maintenance, e.g. ONTAP upgrade?

UM 7.3, 9.4, even with nas name, lif name,,etc indicated, but alerts still creep in.

that causing tons of events per cluster

is it necessary to create same no. "Resources - e.g. cluster, nodes, lif, shelf..etc" in EXCLUSION section as well? same as no. resources we defined in INCLUSION.

or any SIMPLEST way to indicate just a cluster name that BLOCK every alerts fired from cluster during takeover/down..

[mbeattie]

Hi Cavin

The is a um CLI available but it does NOT support a "um alert list" or "um alert modify" command. The API XML is:

<alert-list>
    <alert-info>
        <alert-id></alert-id>
        <alert-name></alert-name>
        <email-addresses>
            <email-address></email-address>
        </email-addresses>
        <event-names>
            <event-name></event-name>
        </event-names>
        <is-disabled></is-disabled>
        <repeat-interval></repeat-interval>
        <resource-types>
            <resource-type></resource-type>
        </resource-types>
        <send-snmp-trap></send-snmp-trap>
    </alert-info>
</alert-list>

IE each alarm has a property to define if the alarm is in an enabled\disabled state. Therefore in theory it would be possible to disable all alarms programmatically for the purpose of invoking scheduled maintence.

The supported method of disable alerts would be to use the UI.

For example, click configuration, alerting, select the top left checkbox to select all alerts and then click the disable button.

/Matt

[Cavin]

Thank you Matt for the solution, we are already using UI and indeed, easiest is to simply disable it, but if we're not looking at large scale of NAS clusters involved, e.g. just 10 -20 nodes, we do not want to disable entire alerts as the other non affected clusters events will be required for auto ticketing solution..etc based on the event occuring.

that's reason, we're looking for simplified way to exclude just bunch of clusters, instead of creating too many individual alerts or extensive manual setting the cluster/inclusion as the event resources

😉

[GidonMarcus]

Hi

seems to be resolved in OCUM 9.5

from release notes:

"New Unified Manager maintenance window functionalityYou can define a Unified Manager maintenance window to suppress events and alerts for a specific timeframe when you have scheduled cluster maintenance and you do not want to receive a flood of unwanted notifications. During a maintenance window the events related to all objects on that cluster are still generated, but they do not appear in any of the UI pages, and no alerts or other types of notification are sent for these events. See the topic "What a Unified Manager maintenance window is" in the online help for more details."