Notifications decide who hears about what, and when. Configure them deliberately at onboarding — otherwise recipients decide for you, by filtering everything NetOp sends.
The rule that catches everyone
Two things people consistently get backwards:
- Being a platform user is not being subscribed. People enrolled on the platform do not automatically receive notifications. They must be added to the notification configuration explicitly.
- Being subscribed does not require a login. Someone with no NetOp account receives notifications perfectly well once added. This is how most external contacts are set up.
The five channels
All five live under Integrations › Notifications.
Email is the workhorse. It carries the full incident summary including Insights and Recommendations, and it is the only channel a non-user can receive.
Slack and Microsoft Teams suit an internal desk — fast, visible, and very easy to over-subscribe. Pick topics narrowly.
PagerDuty is for genuine out-of-hours escalation. Subscribe only to categories that justify waking someone.
Splunk is for downstream analytics and long-term retention alongside your other telemetry.
Recipients and topics
Each line pairs addresses with a topic selection, and you can add as many lines as you need. The five topics:
- Operations Incident — device, port, or VPN down. This is the one people feel, and it usually belongs with both the operations desk and the affected contact.
- Service Performance Incident — SLA QoE, latency, jitter, packet loss. External for anyone with an experience SLA; internal otherwise.
- Utilization and Network Utilization Incident — throughput, WAN saturation, and device resource pressure. Usually internal, since these become a capacity conversation rather than an outage call.
- Anomalies — the raw detection layer. Almost always internal only. Subscribing an external contact to anomalies is the fastest way to get your email filtered.
Review these lines whenever a contact changes. A notification going to someone who left is a notification going nowhere.
The four moments in an incident's life
- Creation — the incident opens. Always on, and the one that matters most: it is your chance to make contact before someone contacts you.
- Update — once a day while open. Useful internally on long-running incidents; consider keeping it off externally unless daily updates were requested.
- Severity change — fires in both directions. Valuable internally, because escalation timers move severity without anyone touching the ticket.
- Closing — the incident closes. Pairs with Creation to give a clean open-and-shut narrative without the noise in between.
A subscription design that works
For external contacts: Operations and Service Performance only, at Creation and Closing only.
For an internal desk: all four categories and all four moments, plus anomalies if you have the appetite for them.
What an email actually contains
Every notification carries an identifying header — severity, incident ID, creation time, site, category, main cause, and subcategories — which is enough to recognise the incident without opening the console. Below that sit Recommendations and Insights, both timestamped, then the constituent Observations with severity, time, element, and status, and finally a link into the console for anyone who has a login.
Bear in mind that the moment you subscribe an external contact, this content is external. Whatever the platform writes is read as coming from you.
Related articles
- Anomalies, observations, and incidents
- Reducing alert noise
Comments
0 comments
Please sign in to leave a comment.