Monitoring Metrics from Anypoint Insights
Feature available in Usage-based and Titanium subscription tiers |
Monitor performance data and issues across your deployed Mule apps and API implementations (entities) with Anypoint Insights, in Anypoint Monitoring:
-
Check metrics across multiple Mule apps and API implementations.
Get up to the last 7 days (Advanced package) or 24 hours (Starter package) of high-level metrics on requests, errors, and responses. Use filters to narrow your results. You can download most chart data to CSV or XSL.
-
Check metrics on individual Mule apps and API implementations.
Get up to the last 30 days of metrics on latency, requests, responses, errors, Mule messages and the JVM for Mule apps, and for API implementations, data on policy violations, performance, and client apps. You can download most chart data to CSV or XSL.
Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
| 1 | Dashboard that provides an overview of metrics across of your deployed Mule apps and APIs |
| 2 | Environment selector for your deployments (such as Sandbox or Production) |
| 3 | Filters to narrow the data in charts |
| 4 | Performance metrics and trends across your deployed Mule apps and APIs |
| 5 | List of each Mule app and API that you can click to display detailed metrics on a specific deployment |
| 6 | Selector for the time range covered by the charts
This selector is available to all Insights dashboards. |
| 7 | Selector for auto-refresh interval, displayed when auto-refresh is enabled
This selector is available to all Insights dashboards. |
Before You Begin
Required prerequisites:
-
An Anypoint Platform account in a control plane that supports Anypoint Insights:
-
A Mule app or API (Mule Gateway or Flex Gateway API) running on a supported version of Mule runtime engine (see Mule Runtime Release Notes) and deployed to a supported target:
-
CloudHub 2.0
-
CloudHub (for US and EU clouds, in Advanced and Titanium tiers only)
-
Hybrid (for US and EU clouds, in Advanced and Titanium tiers only)
-
Runtime Fabric
-
-
For Canada Cloud and Japan Cloud:
-
Enable Mule runtime access to Anypoint Platform so that you can display entity metrics in Anypoint Monitoring.
For guidance, see Configuring Mule Properties to Access Hyperforce.
-
For information about Runtime Fabric feature support in these clouds, see Features Supported on Hyperforce.
-
-
The Monitoring Viewer permission to Anypoint Monitoring and the permissions required for viewing apps and APIs (see Managing Permissions for Anypoint Monitoring)
Monitor Your Deployments
Monitor Mule apps and APIs within a selected business group and environment.
-
Meet the prerequisites in Before You Begin.
-
Log in to Anypoint Platform, and navigate to Anypoint Monitoring.
-
In the Anypoint Monitoring sidebar, click Insights to open Anypoints Insights.
-
In Anypoint Insights, set your environment, such as Design, Production, or Sandbox.
For guidance, see Change the Environment.
-
To capture deployments for another business group you belong to in the Entities List, select the business group.
For guidance, see Change the Business Group.
-
Set the time period to capture in your charts.
For guidance, see Filter Data by Time Period.
-
Filter the deployment data to display in the charts, as needed, to focus on areas of interest.
For guidance, see Filter Deployment Data.
-
To get additional metrics for a specific app or API:
-
Find and click its name in Entities List.
Apps in the Started state and APIs in the Active state are listed.
-
Click View Entity Metrics, for example:
-
Review metrics in the charts for the entity, for example:
-
| 1 | Details page for a Mule app or API implementation |
| 2 | Charts with detailed metrics
Detailed metrics are available in dashboards specific to Mule apps and API implementations. |
| 3 | Selector for the time range covered by the charts
By default, the date period in dashboards with detailed metrics on specific entities matches the time period selected for cross-entity charts in the Anypoint Insights overview. |
| 4 | Selector for auto-refresh interval, displayed when auto-refresh is enabled
By default, the auto-refresh interval in dashboards with detailed metrics on specific entities matches the time period interval for cross-entity charts in the Anypoint Insights overview. |
Filter Monitoring Data Across Deployments
You can narrow the data in Anypoint Insights charts by deployment details, such as the deployment type, and by time range.
Filter Deployment Data
Use one or more filters to narrow the data in your charts. The filters support operators = and IN. Multiple filter settings are allowed. For example, this filter returns Mule apps deployed to CloudHub 2.0:
entity_type = 'APP' deployment_type = 'CLOUDHUB2'
This table lists supported filters:
| Filter | Operators | Description |
|---|---|---|
|
|
Filter by deployment target, such as CloudHub 2.0 (CLOUDHUB2) or Runtime Fabric (RTF), for example: deployment_type IN ['CLOUDHUB2', 'RTF'] |
|
|
Filter by deployment type (APP or API), for example: entity_type = 'APP'` |
|
|
Filter by specific app or API deployments, for example: entity IN ['99aea888-77de-6e06-5c55-d1234de56789', '09f99999-888d-7b77-6da5-bebd43210fc1'] |
IN captures exact matches to any comma-separated element in the array.
Filter Data by Time Period
Use the time-period selector (item 6 in the example) to show data points collected for a given time period (such as the last 15 minutes or the last 24 hours). Your setting applies across all data in Anypoint Insights overview, including items in the Entities List.
If you don’t see data on the page, check that the selected time range is correct.
Change the Environment
Charts in Anypoint Insights provide metrics on a specific environment (such as Production or Sandbox).
To change current or default environment:
-
Open Anypoint Monitoring.
-
Set a new environment:
-
To select an environment without changing the default, use the environment selector in Anypoint Monitoring, for example:
-
To select a default environment:
-
Click your profile icon (see item 2 in the example).
-
Select the name of your profile to open Access Management Settings.
-
In Settings, select an environment as the default environment to view for the selected business group in Anypoint Monitoring, for example:
-
Return to Anypoint Monitoring to check for the new business group setting.
Refresh the page, if necessary.
-
-
Change the Business Group
Charts in Anypoint Insights display metrics for deployments in a specific business group. If you belong to more than one business group, you can change the selected group and the business group that Anypoint Monitoring selects by default.
| 1 | Business-group selector |
| 2 | User profile, for selecting defaults |
To change the business-group setting:
-
Open Anypoint Monitoring.
-
Set a new business group:
-
To select a business group without changing the default, use the business-group selector in Anypoint Insights.
-
To select a default business group:
-
Click your profile icon (see item 2 in the example).
-
Select the name of your profile to open Access Management Settings.
-
In Access Management Settings, select a business group to use as the default, for example:
-
Return to Anypoint Monitoring to check for the new default setting.
Refresh the page, if necessary.
-
-
Download Chart Data
After opening a set of charts in Anypoint Insights, you can download metrics from charts that provide a download menu in the chart.
-
Click the
(More menu) on a chart to open the download menu. -
Select the file format for your metrics:
-
Download as CSV
-
Download as XLS
-
Insights Chart Reference
Understand the monitoring data in Anypoint Insights charts:
Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
Performance Metrics and Trends
Performance Metrics and Performance Trends provide insights across your entities:
Performance Metrics
Monitor high-level performance metrics across your entities. Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Total Entities
-
Total number of entities
- Entities with Errors
-
Total number of entities that have received errors
-
For API implementations, this metric counts HTTP status codes that identify errors (for example,
404or502). -
For Mule apps, this metric counts response codes that indicate an error (for example,
FAILED).
-
- Total Requests
-
Total number of requests to the listed entities
- Error Rate
-
For entities with request errors, the percentage equal to the total number of request errors divided by the total number of requests
For example, assume that 1 out of 5 total requests to API A produces an error and that 2 requests out of 5 to App B produces an error. The error rate is 3/10, or 30%.
Performance Trends
Monitor high-level performance over time across your entities. Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Entities with Slow Response Times
-
Response times, in milliseconds, of the entities with the slowest response times
The metric uses percentile values for response times (calculated from highest to lowest times) to identify entities with the slowest responses. You can select p99 (default), p95, p90, or p75.
- Error Volume
-
Number of errors received by the entities over time
-
For API implementations, this metric counts HTTP error codes (for example,
404or502). -
For Mule apps, this metric counts error response codes (for example,
FAILED).
-
- Request Volume
-
Number requests over time for the entities
Entities List
Check key metrics on individual Mule apps and APIs (entity), and click an entity’s name to get detailed charts for the entity. Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
include::partial$include-insights-charts.adoc[tags="data-scope"] The entities belong to the selected business group.
Entities List provides these metrics:
- Entity Name
-
Name of the entity
- Entity Type
-
Type of entity, such as API or Application (Mule app)
- Request Volume
-
Total number of requests to this entity
- Response Time (p99)
-
Response time, in milliseconds, for the 99th percentile of requests to this entity
This response time is greater than 99% of all response times for this entity.
- Error Rate
-
Percentage of requests to this entity that resulted in errors
- Deployment Type
-
Deployment target of this entity, such as CloudHub 2.0 or Runtime Fabric
Application Details
Anypoint Monitoring provides details on individual Mule apps deployed to a supported target.
Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
Mule Application Metrics
High-level metrics are available in Application Details dashboards for individual Mule apps.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Type
-
Mule application type (Application)
- Status
-
Status code that indicates whether this Mule app is running or in some other state
This code comes from Runtime Manager.
- URL
-
URL of this deployed Mule app
- Runtime
-
Mule runtime engine version of this Mule app
- Replicas
-
Virtual core fraction and count for this Mule app, for example, 0.1 vCores x 1 (= 10% of 1 vCore)
- Last Updated
-
Date-time stamp for the last update or redeployment of this Mule app
Overview Metrics for a Mule Application
Charts with overview metrics are available in the Overview tab for a Mule app in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Average Latency
-
Average latency, in milliseconds (ms), for requests to this entity
- P95 Latency
-
Latency value, in milliseconds (ms), for 95th percentile (p95) request for this entity
- Total Requests Volume
-
Total number of requests to this entity
- Mule Messages Volume
-
Total Mule message count for this app
For information about Mule messages, see Mule Events.
- Request Latency
-
Latency, in milliseconds (ms), at specific time points for the average request and the p90, p95, and p99 percentile requests
You can download data in this chart as a CSV or XSL file.
- Inbound Request Volume
-
Total number of inbound requests to this app at specific time points
You can download data in this chart as a CSV or XSL file.
- Failed Inbound Request Volume
-
Total number of unsuccessful inbound requests at specific time points
You can download data in this chart as a CSV or XSL file.
- Mule Messages
-
Mule message count for this Mule app at specific time points
You can download data in this chart as a CSV or XSL file.
For information about Mule messages, see Mule Events.
Request Metrics for a Mule Application
Charts with request metrics are available in the Requests tab for a Mule app in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Inbound Request Volume
-
Inbound request count total, which you can group by response time or protocol
You can download data in this chart as a CSV or XSL file.
- Outbound Request Volume
-
Outbound request count total, which you can group by response time, protocol, or by the 20 hosts (such as an Object Store or a
cloudhub.ioaddress) with the most outbound requestsYou can download data in this chart as a CSV or XSL file.
- Top Slowest Requests by Path
-
The 20 slowest endpoints in this app
The requests must take more than 25ms. If no data is available, then no requests take more than 25ms to process.
You can download data in this chart as a CSV or XSL file.
- Request Latency by Endpoint
-
Inbound or outbound latency, in milliseconds, by endpoint
The table captures the average latency and the latency value of requests at the p50, p75, p90, p95, and p99 percentiles.
Error Metrics for a Mule Application
Charts with error metrics are available in the Errors tab for a Mule app in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Failed Outbound Request Volume
-
Total number of failed outbound requests at specific time points
You can download data in this chart as a CSV or XSL file.
- Top 20 Failed Requests by Total Count
-
Total number of inbound or outbound requests to the 20 endpoints with the most failed requests
You can download data in this chart as a CSV or XSL file.
- Top 20 Failed Outbound Requests by Total Count
-
Total number of unsuccessful outbound requests to the 20 endpoints with the most failed requests
You can download data in this chart as a CSV or XSL file.
JVM Metrics for a Mule Application
Charts with JVM metrics are available in the JVM tab for a Mule app in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Old Gen GC Count
-
Total number of times old generation Java garbage collection occurred
You can download data in this chart as a CSV or XSL file.
- Old Gen GC Duration
-
Amount of time taken for old generation Java garbage collection
You can download data in this chart as a CSV or XSL file.
- Young Gen GC Count
-
Total number of times young generation Java garbage collection occurred
You can download data in this chart as a CSV or XSL file.
- Young Gen GC Duration
-
Amount of time taken for young generation Java garbage collection
You can download data in this chart as a CSV or XSL file.
- Heap Committed
-
Amount of memory the JVM guarantees to be available for the heap
You can download data in this chart as a CSV or XSL file.
- Heap Usage
-
Amount of physical memory used
You can download data in this chart as a CSV or XSL file.
- Thread Count
-
Number of active threads currently running in the JVM
You can download data in this chart as a CSV or XSL file.
- Committed Virtual Memory
-
Amount of virtual memory the operating systems guarantees to be available for the JVM
You can download data in this chart as a CSV or XSL file.
- CPU Utilization
-
CPU work as a percentage of total CPU time
You can download data in this chart as a CSV or XSL file.
API Details
Anypoint Monitoring provides details on individual API implementations deployed to a supported target.
Entities are deployable units (such as Mule apps and API implementations) that generate the data that’s displayed in Anypoint Insights.
API Implementation Metrics
High-level metrics are available in API Details dashboards for individual API implementations.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Type
-
API implementation type (API)
- Status
-
Status code that indicates whether the governance policies and SLA tiers are functional (Active) or not
For all codes, see information about tracking registration status in the API Manager documentation.
- Implementation URL
-
URL of deployed API integration
- Runtime Version
-
Mule Gateway or Flex Gateway runtime version used by the API implementation
- API Instance ID
-
Instance identifier for an API implementation incorporated into a Flex Gateway or API Gateway runtime
- API Version
-
Version of the API implementation
For more information, see Managing API Instances.
Overview Metrics for an API Implementation
Charts with overview metrics are available in the Overview tab for each API implementation that you can monitor in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Average Latency
-
Average latency, in milliseconds (ms), for requests to this entity
- P95 Latency
-
Latency value, in milliseconds (ms), for 95th percentile (p95) request for this entity
- Total Requests Volume
-
Total number of requests to this entity
- Policy Violations
-
Total number of requests rejected because of API policy violations
For information about policies, see Gateway Policies in API Manager.
- Request Latency
-
Latency, in milliseconds, for requests to this API at specific time points
The chart provides values for the average latency and the p90, p95, p99 percentiles.
You can download data in this chart as a CSV or XSL file.
- Request Volume
-
Total number of requests to this API at specific time points
You can download data in this chart as a CSV or XSL file.
- Failed Requests
-
Total number of unsuccessful requests to this API at specific time points
You can download data in this chart as a CSV or XSL file.
- Policy Violations
-
Total number of requests rejected at specific time points because of API policy violations
For information about policies, see Gateway Policies in API Manager.
You can download data in this chart as a CSV or XSL file.
Request Metrics for an API Implementation
Charts with request metrics are available in the Requests tab for each API implementation that you can monitor in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Request Volume by HTTP Method
-
Total number of requests at specific time points, grouped by the HTTP method used for the request
You can download data in this chart as a CSV or XSL file.
- Request Volume by Response Code
-
Total number of requests at specific time points, grouped by the response code (such as 2xx)
You can download data in this chart as a CSV or XSL file.
- Request Volume by Top N paths
-
Total number of requests at specific time points from the 20 endpoints with the most requests
You can download data in this chart as a CSV or XSL file.
Error Metrics for an API Implementation
Charts with error metrics are available in the Errors tab for each API implementation that you can monitor in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Failed Requests Volume by HTTP Method
-
Total number of unsuccessful requests at specific time points, grouped by the HTTP method used
You can download data in this chart as a CSV or XSL file.
- Failed Requests Volume by Response Code
-
Total number of unsuccessful requests at specific time points, grouped by the response code (such as 2xx)
You can download data in this chart as a CSV or XSL file.
- Failed Requests by Top N Paths
-
Total number of unsuccessful requests at specific time points from the 20 endpoints with the most requests
You can download data in this chart as a CSV or XSL file.
Performance Metrics for an API Implementation
Charts with performance metrics are available in the Performance tab for each API implementation that you can monitor in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Top 20 Paths by Average Response Time
-
Average response time of the 20 most requested endpoints
You can download data in this chart as a CSV or XSL file.
- Top 20 Paths by 95th Percentile Response Time
-
95th percentile value for each of the 20 most requested endpoints
You can download data in this chart as a CSV or XSL file.
Client Application Metrics for an API Implementation
Charts with client application metrics are available in the Client Applications tab for each API implementation that you can monitor in Anypoint Insights.
The scope of metrics depends on the selected environment, time period, and any query filters.
- Request Volume by Location
-
Maps with the number of requests to geographic areas across the globe
You can click the request volume number on the map to get more detailed information.
Data Collection Intervals in Anypoint Monitoring Charts
Anypoint Monitoring charts display data in configurable date-time ranges, which include the last 15 minutes, hour, week, a custom date-time range, and other periods.
Data Points in Anypoint Monitoring Charts
Anypoint Monitoring charts aggregate data into data points along a horizontal Time axis.
Maximum granularity of the data captured in charts:
-
1 minute for metrics captured within the last 7 days
-
1 hour for metrics captured between before the last 7 days
Up to the last 30 days of metrics are collected.
For example, if a chart displays data collected in the last 15 minutes, the chart contains 15 data points, one for each minute of data collection. Each data point aggregates data collected during that minute. If a chart displays the past seven days of data, each data point aggregates data collected within each four-hour period of the last seven days.
How Anypoint Monitoring Collects Metrics
The Anypoint Monitoring agent collects and aggregates metrics on a per-minute basis, so one minute is the shortest possible time increment you can examine in dashboards for your deployments. Metrics collection begins at a random second of that minute, and collection repeats at the same second in subsequent minutes. For example, if the agent collects metrics at 12:00:15, it collects them again at 12:01:15, 12:02:15, and so on.
Within a Mule app, metrics for calls made to and from a connector start on different seconds. For example, metrics for inbound calls to port 8080 might be collected at 12:00:45, 12:01:45, and 12:02:45, while metrics for outbound calls made to an endpoint might be collected at 12:00:23, 12:01:23, and 12:02:23. Because inbound call metrics collection occurs at different seconds during a minute than call metrics for outbound calls, the total number of inbound and outbound calls reported during any minute might vary. For example, if app A reports ten outbound calls made at 12:05, app B might report that the total number of corresponding inbound calls is split between 12:04 and 12:05, or 12:05 and 12:06 time windows (namely, 12:04 reports four calls and 12:05 reports six calls, totaling the ten calls).