AWS AmazonCloudWatch documentation change
Summary
Added documentation for composite SLOs that aggregate Availability metrics across multiple operations. Reorganized SLO creation steps, integrated dependency monitoring into main workflow, and moved naming step later in the process.
Security assessment
The changes focus on feature enhancements for monitoring reliability (composite SLOs) and UI workflow improvements. There is no evidence of security vulnerability fixes, security incident documentation, or new security capabilities. The changes relate to operational monitoring best practices without specific security implications.
Diff
diff --git a/AmazonCloudWatch/latest/monitoring/CloudWatch-ServiceLevelObjectives.md b/AmazonCloudWatch/latest/monitoring/CloudWatch-ServiceLevelObjectives.md index ab825591d..9897dcc9d 100644 --- a//AmazonCloudWatch/latest/monitoring/CloudWatch-ServiceLevelObjectives.md +++ b//AmazonCloudWatch/latest/monitoring/CloudWatch-ServiceLevelObjectives.md @@ -81,0 +82,2 @@ An SLI can be based on one of the new standard application metrics collected by +In addition to creating SLOs on a single operation or on all operations of a service, you can create _composite SLOs_ that monitor a subset of operations for a service. Composite SLOs aggregate the `Availability` metric across multiple operations, giving you a unified view of reliability for a group of related operations. You can select between 2 and 20 operations to include in a composite SLO. For more information, see Create a composite SLO on multiple operations. + @@ -307,0 +310,2 @@ For example, for a period-based SLO, meeting an attainment goal of 99.9% means t + * Create a composite SLO on multiple operations + @@ -323,3 +327 @@ Use the following procedure to create a period-based SLO. - 4. Enter a name for the SLO. Including the name of a service or operation, along with appropriate keywords such as latency or availability, will help you quickly identify what the SLO status indicates during triage. - - 5. For **Set Service Level Indicator (SLI)** , do one of the following: + 4. For **Set Service Level Indicator (SLI)** , do one of the following: @@ -327 +329 @@ Use the following procedure to create a period-based SLO. - * To set the SLO on either of the standard application metrics `Latency` or `Availability`: + * To set the SLO on a service operation, all operations, or the dependency of a service, using either of the standard application metrics `Latency` or `Availability`: @@ -329 +331 @@ Use the following procedure to create a period-based SLO. - 1. Choose **Service or Service Operation**. + 1. For **Type** , choose **Service**. @@ -335 +337,7 @@ Use the following procedure to create a period-based SLO. - 4. Select the operation that this SLO will monitor. To create a service-level SLO that monitors the overall health of your service across all operations, select **All Operations**. Otherwise, select a specific operation to monitor. + 4. For **Type** , choose one of the following: + + * **Service Operations** — to create an SLO on a service operation, all operations, or a subset of operations. + + * **Service Dependency** — to create an SLO on a dependency of the service. + + 5. If you chose **Service Operations** , select the operation that this SLO will monitor. To create a service-level SLO that monitors the overall health of your service across all operations, select **All Operations**. Otherwise, select a specific operation to monitor. @@ -337 +345,11 @@ Use the following procedure to create a period-based SLO. - 5. For **Select a calculation method** , choose **Periods**. +To create an SLO that monitors a subset of operations, see Create a composite SLO on multiple operations. + + 6. If you chose **Service Dependency** , do the following: + + 1. Under **Select an operation** , select one specific operation or select **All operations** to use the metrics from all operations of this service that calls a dependency. + + 2. Under **Select a dependency** , search and select the required dependency for which you want to measure the reliability. + +After you select the dependency, you can view the updated graph and historical data based on the dependency. + + 7. For **Select a calculation method** , choose **Periods**. @@ -341 +359 @@ The **Select service** and **Select operation** drop-downs are populated by serv - 6. Choose either **Availability** or **Latency** and then set the threshold. + 8. Choose either **Availability** or **Latency** and then set the threshold. @@ -345 +363 @@ The **Select service** and **Select operation** drop-downs are populated by serv - 1. Choose **CloudWatch Metric**. + 1. For **Type** , choose **CloudWatch Metric**. @@ -359,9 +377 @@ For more information about these screens, see [Graph a metric](./graph_a_metric. - * To set the SLO on the dependency of a service on either of the standard application metrics `Latency` or `Availability`: - - 1. Choose **Service Dependency**. - - 2. Under **Select a service** , select the service that this SLO will monitor. - - 3. Based on the selected service, under **Select an operation** , you can select one specific operation or select **All operations** to use the metrics from all operations of this service that calls a dependency. - - 4. Under **Select a dependency** , you can search and select the required dependency for which you want to measure the reliability. + 5. If you selected **Service** in step 4, set the period length for this SLO. @@ -369,3 +379 @@ For more information about these screens, see [Graph a metric](./graph_a_metric. -After you select the dependency, you can view the updated graph and historical data based on the dependency. - - 6. If you selected **Service Operation** or **Service Dependency** in step 5, set the period length for this SLO. + 6. Enter a name for the SLO. Including the name of a service or operation, along with appropriate keywords such as latency or availability, will help you quickly identify what the SLO status indicates during triage. @@ -434,3 +442 @@ Use the following procedure to create a request-based SLO. - 4. Enter a name for the SLO. Including the name of a service or operation, along with appropriate keywords such as latency or availability, will help you quickly identify what the SLO status indicates during triage. - - 5. For **Set Service Level Indicator (SLI)** , do one of the following: + 4. For **Set Service Level Indicator (SLI)** , do one of the following: @@ -438 +444 @@ Use the following procedure to create a request-based SLO. - * To set the SLO on either of the standard application metrics `Latency` or `Availability`: + * To set the SLO on a service operation, all operations, or the dependency of a service, using either of the standard application metrics `Latency` or `Availability`: @@ -440 +446 @@ Use the following procedure to create a request-based SLO. - 1. Choose **Service or Service Operation**. + 1. For **Type** , choose **Service**. @@ -444 +450,5 @@ Use the following procedure to create a request-based SLO. - 3. Select the operation that this SLO will monitor. To create a service-level SLO that monitors the overall health of your service across all operations, select **All Operations**. Otherwise, select a specific operation to monitor. + 3. For **Type** , choose one of the following: + + * **Service Operations** — to create an SLO on a service operation, all operations, or a subset of operations. + + * **Service Dependency** — to create an SLO on a dependency of the service. @@ -446 +456 @@ Use the following procedure to create a request-based SLO. - 4. For **Select a calculation method** , choose **Requests**. + 4. If you chose **Service Operations** , select the operation that this SLO will monitor. To create a service-level SLO that monitors the overall health of your service across all operations, select **All Operations**. Otherwise, select a specific operation to monitor. @@ -448 +458 @@ Use the following procedure to create a request-based SLO. - 5. The **Select service** and **Select operation** drop-downs are populated by services and operations that have been active within the past 24 hours. +To create an SLO that monitors a subset of operations, see Create a composite SLO on multiple operations. @@ -450 +460,13 @@ Use the following procedure to create a request-based SLO. - 6. Choose either **Availability** or **Latency**. If you choose **Latency** , set the threshold. + 5. If you chose **Service Dependency** , do the following: + + 1. Under **Select an operation** , select one specific operation or select **All operations** to use the metrics from all operations of this service that calls a dependency. + + 2. Under **Select a dependency** , search and select the required dependency for which you want to measure the reliability. + +After you select the dependency, you can view the updated graph and historical data based on the dependency. + + 6. For **Select a calculation method** , choose **Requests**. + + 7. The **Select service** and **Select operation** drop-downs are populated by services and operations that have been active within the past 24 hours. + + 8. Choose either **Availability** or **Latency**. If you choose **Latency** , set the threshold. @@ -454 +476 @@ Use the following procedure to create a request-based SLO. - 1. Choose **CloudWatch Metric**. + 1. For **Type** , choose **CloudWatch Metric**. @@ -474,11 +496 @@ For more information about these screens, see [Graph a metric](./graph_a_metric. - * To set the SLO on the dependency of a service on either of the standard application metrics `Latency` or `Availability`: - - 1. Choose **Service Dependency**. - - 2. Under **Select a service** , select the service that this SLO will monitor. - - 3. Based on the selected service, under **Select an operation** , you can select one specific operation or select **All operations** to use the metrics from all operations of this service that calls a dependency. - - 4. Under **Select a dependency** , you can search and select the required dependency for which you want to measure the reliability. - -After you select the dependency, you can view the updated graph and historical data based on the dependency. + 5. Enter a name for the SLO. Including the name of a service or operation, along with appropriate keywords such as latency or availability, will help you quickly identify what the SLO status indicates during triage. @@ -547 +559 @@ You can create SLOs to monitor the performance of your CloudWatch RUM app monito - 4. Enter a name for the SLO. Including the app monitor name and appropriate keywords will help you quickly identify what the SLO status indicates during triage. + 4. For **Set Service Level Indicator (SLI)** , choose **RUM AppMonitor**. @@ -549 +561 @@ You can create SLOs to monitor the performance of your CloudWatch RUM app monito - 5. For **Set Service Level Indicator (SLI)** , choose **RUM AppMonitor**. + 5. Select the app monitor that this SLO will monitor from the dropdown list. The list shows the app monitor name along with the supported platform (Web, iOS, or Android). @@ -551 +563 @@ You can create SLOs to monitor the performance of your CloudWatch RUM app monito - 6. Select the app monitor that this SLO will monitor from the dropdown list. The list shows the app monitor name along with the supported platform (Web, iOS, or Android). + 6. (Optional) Select a specific page or screen to monitor. If you don't select a page, the SLO will monitor all pages for the app monitor. @@ -553,3 +565 @@ You can create SLOs to monitor the performance of your CloudWatch RUM app monito - 7. (Optional) Select a specific page or screen to monitor. If you don't select a page, the SLO will monitor all pages for the app monitor. - - 8. For **Select metric** , choose the metric to use for the SLI. The available metrics depend on the platform: + 7. For **Select metric** , choose the metric to use for the SLI. The available metrics depend on the platform: @@ -561 +571,3 @@ You can create SLOs to monitor the performance of your CloudWatch RUM app monito - 9. For **Set condition** , select a comparison operator and threshold for the SLO to use as the indicator of success. + 8. For **Set condition** , select a comparison operator and threshold for the SLO to use as the indicator of success. + + 9. Enter a name for the SLO. Including the app monitor name and appropriate keywords will help you quickly identify what the SLO status indicates during triage. @@ -588,3 +600 @@ You can create SLOs to monitor the performance of your CloudWatch Synthetics can - 4. Enter a name for the SLO. Including the canary name and appropriate keywords will help you quickly identify what the SLO status indicates during triage. - - 5. For **Set Service Level Indicator (SLI)** , choose **Synthetics Canary**. + 4. For **Set Service Level Indicator (SLI)** , choose **Synthetics Canary**. @@ -592 +602 @@ You can create SLOs to monitor the performance of your CloudWatch Synthetics can - 6. Select the canary that this SLO will monitor from the dropdown list. + 5. Select the canary that this SLO will monitor from the dropdown list. @@ -594 +604 @@ You can create SLOs to monitor the performance of your CloudWatch Synthetics can - 7. For **Select metric** , choose either `SuccessPercent` or `Duration`: + 6. For **Select metric** , choose either `SuccessPercent` or `Duration`: @@ -600 +610,3 @@ You can create SLOs to monitor the performance of your CloudWatch Synthetics can - 8. For **Set condition** , select a comparison operator and threshold for the SLO to use as the indicator of success. + 7. For **Set condition** , select a comparison operator and threshold for the SLO to use as the indicator of success. + + 8. Enter a name for the SLO. Including the canary name and appropriate keywords will help you quickly identify what the SLO status indicates during triage. @@ -614,0 +627,70 @@ You can create SLOs to monitor the performance of your CloudWatch Synthetics can +### Create a composite SLO on multiple operations + +You can create a composite SLO that monitors the `Availability` metric across a subset of operations for a service. This is useful when you want to track the reliability of a group of related operations together, rather than monitoring a single operation or all operations. + +Composite SLOs support both period-based and request-based evaluation. You can select between 2 and 20 operations to include. There are two ways to select operations: + + * _Explicit selection_ — Manually pick individual operations from the dropdown. + + * _Pattern matching_ — Use a prefix or regular expression to automatically match operations by name. + + + + +###### Note + +Composite SLOs support only the `Availability` metric. The `Latency` metric is not available for composite SLOs. + +###### To create a composite SLO + + 1. Open the CloudWatch console at [https://console.aws.amazon.com/cloudwatch/](https://console.aws.amazon.com/cloudwatch/). + + 2. In the navigation pane, choose **Service Level Objectives (SLO)**. + + 3. Choose **Create SLO**. + + 4. For **Set Service Level Indicator (SLI)** , for **Type** , choose **Service**. + + 5. Select the service that this SLO will monitor. + + 6. For **Type** , choose **Service Operations**. + + 7. Select the operations to include in this composite SLO. Do one of the following: + + * To manually select operations, choose multiple operations from the **Operation** dropdown. You can select between 2 and 20 operations. + +The selected operations appear as tokens below the dropdown. You can remove an operation by choosing the dismiss icon on its token. + + * To select operations by pattern, select the **Use pattern matching** checkbox. Then do the following: + + 1. For **Pattern type** , choose either **Prefix** or **Regular expression**. + + * **Prefix** matches all operations whose names start with the text you enter. For example, entering `Invoke` matches operations named `InvokeFunction`, `InvokeAsync`, and so on. + + * **Regular expression** matches all operations whose names match the regex pattern you enter. For example, entering `^Invoke.*` matches the same operations as the prefix example.