Chapter 18. How subscription watch gets and refreshes data

The data collection tools gather and periodically send data, including data about subscription usage, to the Cloud Services platform tools that analyze and process this data. After the data is processed, the data that is needed for subscription watch, including the data related to subscription usage and capacity, is sent to subscription watch for display. This data is sent once per day. Therefore, the data that displays in subscription watch is a daily tally of the results in the form of a snapshot, and is not a real-time, continuous usage monitor.

The Red Hat Enterprise Linux data pipeline

The following image provides additional detail about the data pipeline that moves RHEL data from collection to display in subscription watch. The data collection tool, whether you are using Red Hat Insights, Satellite, or Red Hat Subscription Management with the Subscription Manager agent, sends data to the Cloud Services platform processing tools. After data is processed, it is available to Cloud Services platform services such as the inventory service. Subscription watch uses a subset of the data that is available to the inventory service to display data about subscription usage and capacity.

Figure 18.1. The RHEL data pipeline for subscription watch

The RHEL data pipeline for subscription watch

The Red Hat OpenShift data pipeline

For Red Hat OpenShift version 4.1 and later data collection, the tools available in the Red Hat OpenShift Container Platform monitoring stack monitor and periodically sum the CPU activity of all worker-based nodes, while ignoring the activity of infrastructure-based nodes. That data is sent to Red Hat OpenShift Cluster Manager at different intervals for new clusters, resized clusters, and clusters with deleted entities, to maintain currency.

Red Hat OpenShift Cluster Manager then updates the cluster size attribute for existing clusters and creates entries for any new clusters in the Cloud Services platform inventory tool. Lastly, subscription watch analyzes the inventory data and creates an account-wide Red Hat OpenShift Container Platform usage entry that accounts for both core and socket usage to display in the subscription watch interface, along with capacity data.

Figure 18.2. The Red Hat OpenShift data pipeline for subscription watch

The Red Hat OpenShift data pipeline for subscription watch

Heartbeats for data collection tools

The frequency at which the data collection tools send data for processing, also known as the heartbeat, varies by tool. This variance can affect the freshness of the data that subscription watch displays.

The following table shows default heartbeats for the data collection tools. In some cases, these values are configurable within that data collection tool.

Table 18.1. Heartbeats for data collection tools

ToolConfigurableHeartbeat interval

Insights

No

Daily, once every 24 hours

Red Hat Subscription Management

Yes

Multiple times per day, 4 hour default

Satellite

Yes

Monthly, configurable with the Satellite scheduler function

If used, the Satellite inventory upload plugin reports daily, with a manual send option.

Additionally, to maintain accurate information about the mapping of virtual guests to hosts, a best practice is to run the virt-who utility daily.

Red Hat OpenShift

No

Several tools are involved in the data pipeline, including tools in the Red Hat OpenShift Container Platform monitoring stack and in the Cloud Services platform, with differing intervals:

Red Hat OpenShift Container Platform monitoring stack:
New clusters identified every 15 minutes
Cluster size updated every 2 hours
Cluster cleanup for deleted entities updated every 5 hours

Red Hat OpenShift Cluster Manager:
New clusters identified to Red Hat Subscription Management every 15 minutes
Existing clusters synchronized every 6 hours

subscription watch:
daily, once every 24 hours