Code360 powered by Coding Ninjas X Naukri.com. Code360 powered by Coding Ninjas X Naukri.com
Last Updated: Mar 27, 2024

Basics of Metrics, Time Series and Resources in Cloud Monitoring

Leveraging ChatGPT - GenAI as a Microsoft Data Expert
Speaker
Prerita Agarwal
Data Specialist @
23 Jul, 2024 @ 01:30 PM

Introduction

Cloud Monitoring gathers data from hosted uptime probes, application instrumentation, Google Cloud, Amazon Web Services (AWS), and events. You can also gather this information using the BindPlane service from more than 150 popular application components, on-premises systems, and hybrid cloud systems. In this article, the reader will learn about Metrics, Time Series, and Resources in Cloud Monitoring. We will also discuss the Cloud Monitoring metric model.

introduction

Components of the metric model 

Three key ideas make up the monitoring data paradigm used by Cloud Monitoring:
 

  • Monitored-resource types
     
  • Metric types
     
  • Time series

These ideas are broadly described in the Cloud Monitoring metric paradigm. Please read the page first if you are unfamiliar with these ideas.

Get the tech career you deserve, faster!
Connect with our expert counsellors to understand how to hack your way to success
User rating 4.7/5
1:1 doubt support
95% placement record
Akash Pal
Senior Software Engineer
326% Hike After Job Bootcamp
Himanshu Gusain
Programmer Analyst
32 LPA After Job Bootcamp
After Job
Bootcamp

The Cloud Monitoring metric model

A metric is a group of associated measurements you use to track a particular characteristic of a resource. The number of tables in your SQL database, the number of widgets sold, the latency of queries to a service, the quantity of disc space on a machine, and so on are some examples of measurements. Virtual machines (VMs), database instances, storage, and other resources are examples of resources.

A metric in the context of cloud monitoring often consists of three parts:

  • Information on the measurement's source.
  • Measurements of a particular attribute. Each measurement is saved as a value with a time stamp.
  • Specifics regarding the values of the variable being measured.

Consider the scenario where you wish to monitor the number of widgets that your stores are selling. The following ways the model's elements correspond to this example:

Source of the measurements

 

Each resource being tracked is documented by the metric model. Depending on the type of resource being tracked, several pieces of information may be gathered, such as disc IDs, method names, geographic locations, and other data that could be used as a measurement source.

A monitored resource is the source of monitoring data.

Example: The retailers that sell the widgets are the monitored resources in the widget-sales scenario.

Measurements

The metric model records measurements of a property as a collection of time-stamped data points called data points.

Usually, values are expressed in numerical form, but this depends on the measurement.

The measurements in the widget-sales example keep track of sales data over time. These measurements might be like the ones below:

[(150, 2019-05-23T17:37:00-04:00),

 (229, 2019-05-23T17:38:00-04:00),

 (138, 2018-05-23T17:39:00-04:00),

 ...]

Information about the values

Without knowledge of how to evaluate them, the measurement values have no meaning. You need to know the "type" of the values, such as the data type, unit, and nature of each measurement:

 

  • Is the value a string or an integer?
     
  • Is the figure in radians or miles per hour?
     
  • Does the value reflect the sum as of that moment or the difference from the prior value?

Every set of characteristics pertaining to an item you want to monitor is referred to as a metric type by Cloud Monitoring.

Example: Using the widget-sales example, this data may reveal the following to you:

It stores each value as a 64-bit integer.

Each amount corresponds to the number of sold widgets.

Each value corresponds to the number of widgets sold since the previous measurement was taken.

Time series: putting the components together

The time series data structure is what this model in Cloud Monitoring is based on (the singular and the plural forms are the same).

The three model elements are represented by each time series:

  • A description of the resource under observation that was the source of the measurements.
     
  • The collection of measures connected to a single resource under observation.
     
  • A metric-type description of the object you are measuring.
     

An illustration of a time series in the widget-sales case is as follows:

 

  • A summary of the retail location where the widgets included in this time series were sold.

 

  • The collection of measurements was taken at this shop.
     
  • The values are 64-bit integers that represent the number of widgets sold since the preceding value was recorded.
     

A single monitored resource type or cloud monitoring metric type can be linked to a large number of related time series. If there are 15 retailers selling widgets, there can be 15-time series documenting widget sales because each store selling widgets saves its data in a time series.

Metrics

Each time series that Cloud Monitoring generates stores a set of data points together with details on how they are organized and what they imply.

Metric objects

A reference to the data being recorded in a metric object is included in each time series.

The type of measurements and metric-specific details about those measurements are specified by the metric object contained in a time series. Metric provides a definition of the metric object data structure. Here is an illustration of the metric object that was taken from a sample time series of data:

{
      "metric": {
        "labels": {
          "log": "kubelet",
          "severity": "DEFAULT"
        },
        "type": "logging.googleapis.com/log_entry_count"
      }

This object informs you that measurements from the logging.googleapis.com/log entry count are present in the time series. The label values let you know that this particular time series only counts kubelet log items with severity DEFAULT.

INFO entries to the same log file show up in a distinct time series because there is one time series for every combination of label values.

The descriptor for that metric type specifies the set of labels gathered in that metric object.

Metric types

Metric types explain the measurements that can be taken from a resource that is being watched. An explanation of what is being measured and how the measurements are interpreted is part of a metric type. A rough estimate of 1,500 different kinds of metrics are supported by Cloud Monitoring, and you have the option to define new ones.
 

Metric kinds include disc utilization statistics, storage consumption, and counts of API calls, among many more.

A data structure known as a metric descriptor has a formal description of each metric type. See Metric descriptors for further details.

The Metrics list contains entries for each of the built-in metric kinds. The metric descriptors are used to produce the entries in these tables.

Monitored-resource

Both metric types and monitored-resource types support labels, enabling the categorization of data during analysis. For instance:

  • The location of the virtual machine and the project ID connected to the device may both have labels on a monitored-resource type for virtual machines. The values for the labels are recorded along with the information about the resource being tracked.
     
  • In addition to the labels predefined for the type of monitored resource, a resource under monitoring may also include system- or user-provided metadata labels.
     
  • The name of the method that was called and the request's status may be recorded in labels for a metric type that counts API requests.

Frequently Asked Questions

faqs

On what model does the Google Cloud Platform rely?

Similar to other public cloud services, the majority of Google Cloud services operate on a pay-as-you-go basis, where customers only pay for the cloud resources they actually utilize.

What distinguishes Google Cloud from the Google Cloud Platform?

A variety of online services that are part of Google Cloud can assist businesses in going digital. A component of Google Cloud is the Google Cloud Platform, which offers public cloud infrastructure for hosting web-based applications and is the subject of this blog article.

What are the four types of cloud storage?

The four types of cloud storage are private cloud storage, private cloud storage, hybrid cloud storage, and community cloud storage.

How many products does Google Cloud have?

There are over 100 products under the Google Cloud brand.

Conclusion

We covered Metrics, Time Series, and Resources in Cloud Monitoring in this article. We hope this article helps you to learn something new. And if you're interested in learning more. See our posts on AWS vs. Azure and Google CloudGoogle BigQueryAWS Vs Azure Vs Google Cloud: The Platform of Your Choice?Java knowledge for your first coding job.

Visit our practice platform.  Coding Ninjas Studio to practice top problems.  Attempt mock tests. Read interview experiences and much more.! 

Feel free to upvote and share this article if it has been helpful for you.
 

 

Topics covered
1.
Introduction
2.
Components of the metric model 
3.
The Cloud Monitoring metric model
3.1.
Source of the measurements
3.2.
Measurements
3.3.
Information about the values
4.
Time series: putting the components together
5.
Metrics
6.
Metric objects
7.
Metric types
8.
Monitored-resource
9.
Frequently Asked Questions
9.1.
On what model does the Google Cloud Platform rely?
9.2.
What distinguishes Google Cloud from the Google Cloud Platform?
9.3.
What are the four types of cloud storage?
9.4.
How many products does Google Cloud have?
10.
Conclusion