Overview

Prometheus is an open-source systems monitoring and alerting toolkit with a dimensional data model, flexible query language, efficient time series database, and modern alerting approach.
Prometheus collects metrics from monitored targets by scraping metrics HTTP endpoints on these targets. Since Prometheus also exposes data in the same manner about itself, it can also scrape and monitor its own health.

In Prometheus, a user configures service exporters to store application metrics.

With new design, monitor calls single URL to get response of all standard metrics with specified properties in monitor configuration file and for custom metrics, one URL each is dedicated. For example, if a user provides five custom metrics and five standard metrics in a configuration file, five URLs for custom metrics and one URL for standard metrics are invoked by the monitor.

And monitor will perform this action by single thread. By using this approach, the captured data is reduced. Therefore, monitor performance to capture / parse that response is increased.

Prometheus Configuration

To configure Prometheus, follow the below mentioned steps:
1. Login to a machine, go to Product UI Home page, click Monitors icon, and then click the Monitors menu item.

2. This displays the Monitor Group window. Select a ‘topology’ from the drop down list and click a ‘monitor group’ name.

3. This displays a list of Monitors. Expand the Prometheus group. A list of following monitors is displayed:

• ApplicationHealthStats
• KubernetesContainerNamespaceStats
• KubernetesContainerOverallStats
• KubernetesContainerStats
• KubernetesDeploymentStats
• KubernetesNodeStats

4. To configure any monitor, click the Configure Monitor icon.

5. This displays the Monitor Configuration window.

6. Provide the details for each field as follows:

• Exclude Tier: From the drop down list, the user can choose a tier that they want to exclude. The user can also type a pattern as a regular expression that allows them to exclude a tier. The user can type multiple regular expressions separated by comma.
• Type the Instance Name.
• Type the Prometheus Base URL.
• User Name: Type the user name for basic HTTP authentication.
• Password: Type the password for basic HTTP authentication.
• Filter Settings: Select the filter type as None, Positive Filter or Negative Filter from the drop-down, and then type the filter in the text box. The user can type multiple filters separated by comma in a text box.

• Proxy Settings: Type the Proxy URL, Proxy User Name, and Proxy Password.

• Recovery Settings: Type the Retry Count. By default, the retry count value is 60.

• Advanced Settings: Provide the following details under advanced settings.
  • Type the Absolute Metrics Configuration File Path. This is used to give externally configured metrics.
  • Vector Persist Count: Type the number of count to hold vector if it is freguently going up and down. The default count is 5.
  • URL Response Timeout (Secs): Type the maximum time used to get response for specified URL. The default value is 10 seconds.
  • Thread Pool Size: Type the number of threads used in parallel to get response from Prometheus URLs. If the number of thread is zero, it will calculate automatically based on the configured metric_url size.
  

7. Click Add to add the monitor configuration.

8. To delete one or more configured monitors, click the Delete icon in the lower-right corner of the window.

9. Select the check box to enable the monitor and click Save.

Prometheus Capabilities

ApplicationHealthStats

KubernetesContainerNamespaceStats

KubernetesContainerOverallStats

KubernetesContainerStats

KubernetesDeploymentStats

KubernetesNodeStats