FAQ

• Can I set both static and anomaly detection on an SLI? If I can, which one takes precedence?
• For static alert - I know the lower/upper bound and the x times out of y scenario. For anomaly detection, because it is seasonality-based - I have the advantage of not setting a threshold. But what are the criteria for an alert? When the dynamic threshold is breached, does it immediately alert? Is there a sustained period for which peak/drop is maintained before alerting? What factors come into play?
• How are upper and lower bounds used in static alerts?
• I have a component that is not managed by the Cloud, and I want it to be observed. What are my options?
• Why should I enable container insights on EKS; shouldn't AWS show EKS metrics by default like it does for other managed services like AWS RDS, AWS Elasticache, etc?
• When I set an SLO for a service, how do I know when will it get triggered? e.g. 1-day SLO at 90% - does this mean that if I get 100 requests in a day and 11 fail - my SLO is breached? Is the 1-day / 7-day rolling window?
• How do I set a global slack channel for all my alerts?
• How do I set per service slack channel for my alerts?
• What if I don't set per service slack channel settings, will I still get alerts?
• I am using EKS for observing my K8s setup. I cannot see some of my services on the dashboard. Where are they?
• Why has Last9 recommended additional costs to stream metrics to Last9?

Did we miss something? Should we add a new section? Tell us at [email protected]