You notice delays between the time your AWS integration makes an API request and infrastructure returns the metric data.
Depending on the AWS integration, the infrastructure agent may experience explicit delays or implicit delays in the timing between the API request and the metric data returned. For more on this cause, see Cause.
Normally, the delays do not sum up. With an explicit delay, New Relic expects the metric data to be at the delay point in time. For example, if New Relic uses an explicit delay of five minutes, at 9:00 the freshest data point should be the one for 8:55.
However, delays may vary by customer and account. If you experience unusual delays in your metric data:
- Check whether your infrastructure integration has explicit delays or implicit delays with API requests for metric data.
- CloudWatch users: Take a screenshot of the metric data in your CloudWatch console, and attach it when you request support at support.newrelic.com.
Using your CloudWatch console's screenshot, New Relic Support will troubleshoot if the delay occurs in infrastructure or directly in AWS.
Depending on the AWS integration, the infrastructure agent may experience explicit delays or implicit delays in the timing between the API request and the metric data returned.
Metric data delays
Explicit delays are the ones that New Relic sets in the code to get more reliable data. In some cases, the API request to AWS may return one value, but a request for the same metric a minute later returns a different value. To reduce the possibility of this occurring, New Relic sets explicit delays. For example:
(expected but not specifically set)
Implicit delays are patterns that New Relic has experienced with integrations. They are not always present, and they are not exact.
In general, implicit delays tend to come from requests for AWS CloudWatch metrics, including these Infrastructure Amazon integrations: