COVER COMING SOON

eBook: How to Achieve AWS, Azure, or GCP Observability at Scale

Table of Contents
Section I: Section II: Section III:
Growth & Challenges of Guide to Achieving Multi-Cloud How to Excel at Multi-Cloud
Multi-Cloud Observability Observability Across AWS, Observability Across AWS,
Azure, and GCP Azure, and GCP

Why are Enterprises 4 What Should You Monitor 14 How to Enhance Amazon 33
Adopting Multi-Cloud? on AWS? Cloudwatch to get Full-Stack,
Multi-Cloud Observability
Challenges of Monitoring 6 Why Choose Wavefront for 15
Applications Across Multi-Cloud AWS Observability How to Monitor Serverless 39
Applications in AWS
Overview of Wavefront by 9 What Should You Monitor 17
VMware: Enterprise Observability on Azure? Guide to AWS App Mesh 43
and Analytics Platform Observability and Analytics
for Multi-Cloud Environments Why Choose Wavefront 21
for Azure Observability Optimize Both Service 46
Performance and Cloud Costs
What Should You Monitor 23 with CloudHealth and Wavefront
on GCP?
Top 5 Lessons Learned 50
Why Choose Wavefront 27 from Customers with
for GCP Observability Multi-Cloud Monitoring

How to Monitor VMware 29 Why Customers Choose 55

Cloud on AWS? Wavefront for Multi-Cloud
Observability and Analytics

2
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Section I:
Growth & Challenges of
Multi-Cloud Observability

3
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

1. Why are Enterprises

Adopting Multi-Cloud?
The adoption of multi-cloud is on the rise among enterprises.
According to a survey conducted by analyst firm Forrester on
behalf of Virtustream, 86% of respondents characterized their
organizations’ cloud strategy as multi-cloud. Before going into the
details of why enterprises are adopting multi-cloud, let’s first clarify
what’s multi-cloud and how it differs from hybrid cloud.

Multi-cloud is the approach of using more than one public or private

cloud providers like Amazon Web Services (AWS), Microsoft Azure
(Azure) or Google Cloud Platform (GCP) to run different services.
Often enterprises run a service on one or two clouds. However,
because they own and run many different services – operated by
different teams and businesses, they end up using multiple public
and private cloud providers. No matter if you are using only one
or multiple clouds, this ebook walks you through how to achieve
observability for each of the major cloud providers including AWS,
Azure, GCP and VMware cloud on AWS.

4
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Before we go into the details of each Here are some of the reasons why enterprises
cloud, lets clarify the difference between are adopting a multi-cloud approach:
multi-cloud and hybrid cloud. Hybrid cloud
is a kind of multi-cloud which uses both Utilize best of the breed services Data sovereignty
private and public cloud with orchestration AWS, Azure, and GCP provide a breath Certain laws and corporate policies require
between the two clouds. For instance, of amazing services (IaaS, PaaS, SaaS), enterprises to store their data in certain
VMware Cloud on AWS is a hybrid cloud including many CI/CD, Kubernetes, locations, often within country/state
solution that enables you to extend your AI/ML services. Engineering teams want boundaries. Multi-cloud helps enterprises
on-premises Software-Defined Data Center to benefit from those services and not to scale to different geographies while
(SDDC) to AWS public cloud with VMware re-invent the wheel. abiding by data sovereignty laws.
vCenter acting as the orchestrator between
the two. Now that we know the difference Optimize cost Avoid vendor lock-in
between multi-cloud and hybrid cloud, Different cloud providers have different Enterprises are very aware that being
let’s concentrate on multi-cloud and why pricing models. Organizations often make all-in on one public cloud provider gives
enterprises are increasingly adopting a the decision to adopt different cloud that cloud provider an edge on pricing
multi-cloud strategy. providers or run dev and production negotiations. So, enterprises often
workloads on a different cloud, if the adop a multi-cloud strategy to avoid
price is favorable. vendor lock-in.

Mergers and Acquisitions High availability and redundancy

Large enterprises acquire different teams/ Maintaining high availability by hosting
businesses that use different cloud providers applications on different clouds is a common
as a result of mergers and acquisitions, way to achieve high availability and data
leading to a multi-cloud environment. redundancy, especially for data backups.

5
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

2. Challenges of
Monitoring Applications
Across Multi-cloud
There are many benefits of adopting a multi-cloud approach, as Siloed cloud vendor tools lead
shown above. In fact, in many ways, the adoption of multi-cloud has to slower troubleshooting and
made life much easier for engineering teams. They can now select significant revenue impact
the cloud of their choice, deploy in minutes, and in some cases
altogether, avoid having any datacenters. Furthermore, they can Organizations often start with the tools provided
take advantage of the best of breed IaaS/PaaS and SaaS services by the cloud vendor to monitor their applications.
provided by different cloud providers. But they soon realize that cloud vendor tools are
built for only one vendor and don’t effectively
But at the same time, multi-cloud has introduced many new monitor multi-cloud environments. To overcome
challenges for the engineering team as well. Now instead of one this problem, Ops/DevOps teams find themselves
cloud, engineering teams need to learn, deploy, and monitor across using multiple siloed monitoring solutions that fail
multiple cloud providers. Because cloud providers are so different, to provide a unified view across different cloud
monitoring across different cloud environments is not easy. Here are leading to increased troubleshooting time and
some of the challenges faced by engineering teams while monitoring significant revenue impact.
multi-cloud environments:

6
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Lack of multi-cloud support this problem, Ops/DevOps teams find themselves

in legacy monitoring solutions using multiple siloed monitoring solutions that fail
to provide a unified view across different cloud
Enterprises transitioning to cloud or multi-cloud leading to increased troubleshooting time and
often start with using their legacy monitoring significant revenue impact.
solutions (which worked for their on-premises
environment). But they soon realize that legacy
Increased cloud cost due to
monitoring solutions lack multi-cloud support.
underutilized resources and
Many teams spend precious resources on building
dynamic environments
monitoring capabilities across multi-cloud instead
of focusing on scaling their business. Even open- In public cloud environments, pricing is often
source monitoring solutions have their challenges, based on which resources you use and for how
including scale limitations, limited reliability, and long. So, it’s critical to rightsize the resources you
lack of high availability. use and not have underutilized resources.
Furthermore, in public cloud environments,
resource utilization can be highly dynamic.
Siloed cloud vendor tools lead to slower
Depending on the time of day, month, or season,
troubleshooting and significant revenue
your cloud services environment (test vs. prod)
impact
likely requires change, and you may have a
Organizations often start with the tools provided dramatic rise in resource utilization. For example,
by the cloud vendor to monitor their applications. you may need more compute and storage
But they soon realize that cloud vendor tools are instances during the peak shopping season,
built for only one vendor and don’t effectively then no longer need them after peak times.
monitor multi-cloud environments. To overcome Engineering teams move quickly, often forgetting

7
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

unused cloud resources that they are getting

charged for. Visibility into changes in resource
capacity and performance becomes paramount to
ensuring that you are not paying for unused cloud
infrastructure.

Complexity of different cloud

services leaves DevOps teams blind

Each public cloud provider – AWS, Azure, or GCP

- is very different in terms of their architecture,
capabilities, and services they offer. So, it is not
trivial to learn each cloud provider and determine
how to monitor applications in each cloud.
The complexity of learning how to monitor
applications on each of these clouds leaves
DevOps teams struggling to monitor, measure,
and maintain SLOs.

8
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

3. Wavefront by VMware: Enterprise Observability

Platform Delivering Full-Stack Observability and
Analytics Across Multi-Cloud
Wavefront is an observability platform specifically Key use-cases of Wavefront include:
designed for enterprises that need monitoring,
• Multi-Cloud Observability
observability, and analytics for their cloud-native
• Full-Stack Observability
applications and environments. DevOps, SRE, and
• Application Observability
developer teams use Wavefront to proactively
• Kubernetes Observability
alert on, rapidly troubleshoot, and optimize the
• Infrastructure Visibility
performance of their modern applications running
• CI/CD Visibility
on the enterprise multi-cloud.

Wavefront is purpose-built for modern apps “Wavefront is our enterprise-wide observability

running on the enterprise multi-cloud at scale. platform used by hundreds of engineers across
It delivers full-stack observability with advanced Workday’s Ops and Dev teams. With its unique
analytics on metrics, traces, histograms, and analytics built on top of a common repository of
span logs gathered on distributed applications, full-stack operational data, Wavefront enables
application services, container services, and a better team collaboration and data-driven system
multi-cloud based on public, private, and hybrid improvements.”
cloud infrastructures.
– Kevin Cantoni VP Engineering

9
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Wavefront delivers full-stack observability and analytics across multi-cloud environments

To learn more about Wavefront and how it provides full-stack observability

and analytics across enterprise multi-cloud, check out wavefront.com.

10
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

4. How to Achieve Observability

for AWS, Azure, or GCP
Which AWS Metrics
Should You Monitor?
AWS is the leading public cloud provider.
It provides a vast array of metrics that can
be analyzed in Wavefront across many of its
key components. Using the Wavefront AWS
Integration, you get insight into ~40 AWS services
and their most commonly used metrics including:

• AWS performance and utilization metrics for

major cloud services such as Amazon EC2,
Amazon EBS, Amazon ECS, Amazon ELB,
AWS Lambda, Amazon Redshift, Amazon
DynamoDB, Amazon SQS and more
• Pricing and billing metrics, including reserved Wavefront provides AWS observability with built-in monitoring for ~40 AWS services
instances pricing and spot pricing visibility

11
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

The Wavefront AWS integration ingests

data from the following products:

CloudWatch
retrieves AWS metric
and dimension data.

CloudTrail
retrieves EC2 event information and
creates Wavefront System events that
represent the AWS events.

AWS Metrics+
retrieves additional metrics using AWS APIs
other than CloudWatch. Data include EBS
volume data and EC2 instance metadata like tags.
You can investigate billing data and the number
of reserved instances. Be sure to enable AWS+
metrics because it allows Wavefront to optimize
its use of Cloudwatch and saves money
on Cloudwatch calls as a result.

For more information on Wavefront AWS

metrics collection, please see Wavefront
AWS documentation.

12
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Monitoring AWS Health

The Wavefront AWS integration provides packaged visibility
into key AWS performance, health, and cost metrics. The
pre-packaged visibility from these dashboards enables cloud
monitoring teams to isolate potential resource bottlenecks
quickly. The Wavefront AWS integration provides both at-
a-glance summarized views across the entire AWS footprint
and detailed component views of individual AWS component
consumption. Data can be grouped by different dimensions
(tags), offering easy navigation to particular resource
bottlenecks or anomalies. Using the power of the Wavefront
Query Language, all packaged dashboards can be cloned and
customized, and packaged data is correlated with top-level Granular EC2 Visibility
composite service performance, giving engineering teams the
ability to tailor dashboards to their specific use cases. Wavefront AWS dashboards provide visibility into
the performance of Amazon EC2 hosts including:

Below are some of the packaged dashboards with

corresponding AWS metrics provided by Wavefront AWS • Number of successful and failed instances

Monitoring Suite. • Total number of running instances by region and type

• Overall CPU utilization
• CPU usage by region

13
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Real Time EKS Health Insights Continuous Lambda Monitoring

Wavefront EKS dashboards provide real-time Kubernetes Wavefront AWS Lambda dashboards provide
monitoring of key EKS metrics, including: continuous visibility into Lambda metrics including:

• Detailed health of your infrastructure and workloads. • Lambda function invocations

• Metrics by namespace: Details of your pods/containers • Lambda function errors
broken down by namespace. • Lambda function cold starts
• State of Kubernetes objects such as pods, • Execution time of the Lambda function
services, deployments and daemon sets. handler (in milliseconds)

14
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Why Choose Wavefront for AWS Observability?

Real Time, Graphical Visualization Analytics-Driven Insights into

of Your Global AWS Resources AWS - Performance and Utilization
Enhanced Wavefront AWS dashboards provide a rich Wavefront provides analytics-driven, pre-built dashboards
graphical visualization of your AWS resources across the to monitor key AWS services including EC2, EBS, EKS, Kinesis
globe. With real-time, color-coded node-maps, SRE, and Firehose, CloudFront, CloudTrail, Lambda, ELB, DynamoDB,
DevOps teams can easily view their AWS resources across Redshift and many more. These insights equip you to make
different regions and quickly find nodes that exceed any data-driven decisions for things like AWS capacity planning
metric threshold. and resource optimization.

15
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Built-In Visibility into AWS Billing Unified, Full-Stack AWS

Can Save You a Bundle Monitoring That Reduces MTTR
Wavefront applies analytics to your real-time AWS resource Wavefront makes it easy to pull together metrics, traces,
consumption with detailed AWS pricing, to show you how to and histograms from applications, infrastructure, and different
reduce your cloud costs – many customers have used this to AWS services (including CloudWatch, CloudTrail, and other
save over $60K monthly! AWS APIs), providing a unified, full-stack view
that significantly reduces MTTR.

16
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

What Should You Below are some of the key

Monitor on Azure? Azure metrics to monitor for Azure.

Azure Virtual Machines and Azure Azure

Metrics
Azure is the second most popular public cloud Virtual Machine Scale Sets Storage Accounts Web Apps

provider after AWS and is quickly gaining market

share. In this section, we will cover key Azure Capacity Percentage CPU, CPU Credits UsedCapacity UsedCapacity
Remaining, CPU Credits (StorageAccounts/Blobs/ (StorageAccounts/Blobs/
metrics to monitor, provide details on Wavefront Consumed, AutoscaleSettings Tables/Queues/Files) Tables/Queues/Files)
ObservedMetricValue,
Azure integrations and go over some cost benefits AutoscaleSettings ObservedCapacity

of monitoring Azure metrics in Wavefront.

N/A Availability Http 3xx errors, Http 4xx
Errors errors, Http 5xx errors
Microsoft Azure Monitor provides base-level
infrastructure metrics and granular performance
and utilization metrics for most services in Azure
(see the documentation. To get a full picture Latency
Disk Read/Write Bytes/Sec, Average AverageResponseTime
Disk Read/Write Operations/Sec SuccessE2ELatency,
of the health of all services and corresponding Average
SuccessServerLatency
resources, DevOps and SRE teams often
categorize metrics into four categories, i.e.,
Network In/Network Out, Disk Total Transactions, Total AverageResponseTime
capacity, errors, latency, and throughput. Throughput Read/Write Bytes/Sec, Disk Read/ Ingress, Total Egress
Write Operations/Sec

Key metrics to monitor for different Azure services

17
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Easily Monitor Azure Virtual Machines,

Scale Sets, Storage Accounts and
Container Instances with Wavefront
Whether you are using Azure virtual machines Overview dashboard Performance Metrics
or have virtual machine scale sets configured to of Azure Scale Sets such as CPU utilization, disk
deploy, manage and auto-scale identical VMs for across different subscriptions, read and write bytes, IOPS, and
large-scale applications or have a containerized regions and resource groups network in and out throughput
environment using Azure containers, in each case, for selected scale set
you can easily monitor all your Azure resources Top 10 Stats
using the Wavefront Azure Integrations. Auto Scaling metrics
such as top 10 VMs by CPU
utilization or network in/out for such as observed capacity,
All Wavefront integrations come with pre-built
selected resource groups scale actions initiated and
dashboards that show you a view of all your
metric threshold vs. value
resources. You can filter information, change time
windows, compare different points in time, and
even clone the dashboard to add and remove
more charts. The screenshot below shows the
Wavefront dashboard for Azure Virtual Machine
Scale Sets. In one glance, you can see:

18
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Pre-built Wavefront dashboard for Azure Virtual Machine Scale Sets

19
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Similar dashboards are included for Azure

Virtual Machine instances, Azure Storage
Accounts and Azure Container instances as
well. The following screenshot shows the
dashboard for Azure Storage Accounts,
which shows all active Azure storage
accounts, and corresponding metrics
such as account used capacity, total
transactions, ingress, server latency and
E2E latency.

Pre-built Wavefront dashboard for Azure Storage Accounts

20
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Why Choose Wavefront for Azure Observability

Out-of-the-Box Visibility Into Azure Behavior-Based Alerts Identify Anomalies

Wavefront provides out-of-the-box curated dashboards to in Your Azure Usage and Performance
monitor key Azure services and resources such as virtual With Wavefront, you can easily compare current Azure
machines, scale sets, storage accounts, container instances, resource consumption and application performance with
and many more. DevOps and SRE teams can concentrate previous weeks and turn queries into intelligent, dynamic
on services and resources they care about because they can alerts that are triggered when Wavefront notices something
easily filter based on parameters like subscription, region, and abnormal in your Azure environment. Additionally, you can
resource groups. get alert notifications on the channel of your choice (Slack,
PagerDuty, ServiceNow, HipChat, and so on).

21
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Unified, Full-Stack Azure Observability

Can Significantly Reduce MTTR
Wavefront provides over 200 integrations, so you can easily unify
metrics and alerts from different Azure services, applications, and
infrastructure. A unified, full-stack metrics view becomes available
to your SREs and DevOps teams, including developers–leading
to a significant reduction in MTTR.

22
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

What Should You Monitor on GCP ?

GCP is a top three public-cloud computing provider along be visualized via dashboards, charts, and alerts.Stackdriver
with AWS and Microsoft Azure. Like other cloud providers, alerting is integrated with tools like Slack and PagerDuty,
GCP lets you spin up compute, storage, network resources so your team can get alerted on the tool of their choice.
on the fly. It offers a variety of services such as Google Cloud
Functions, Google ML Engine, and Google Container Engine. Stackdriver Trace
is a distributed tracing system for Google Cloud Platform.
Stackdriver, the Google Cloud’s embedded observability It collects latency data from your applications and generates
suite, is designed to monitor, troubleshoot, and improve in-depth latency reports which surface any performance
cloud infrastructure, software, and application performance. degradations seen in your applications. Stackdriver Trace
can capture traces from all your VMs, containers, or App
Engine projects.
Google Stackdriver’s main features include:

Stackdriver Monitoring Stackdriver Logging

provides visibility into the health and performance of cloud is a log-management service offered by GCP. It enables you
services running on GCP and AWS. Stackdriver Monitoring to store, search, analyze, and alert on log data and events
is based on collectd, an open-source daemon that collects from GCP and AWS. It’s a fully managed services that is built
system and application performance metrics. Stackdriver for scale and can handle system log data from thousands of
Monitoring collects metrics, events, and metadata from workloads. Furthermore, it enables you to analyze all the log
various sources, including GCP, AWS, hosted uptime probes, data in real-time.
and application instrumentation. All the ingested data can

23
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Achieve OOTB
GCP Observability
with Wavefront
In response to increasing demand from Wavefront
customers, we have introduced Wavefront GCP
Integration. The Wavefront GCP integration
collects, analyzes, and visualizes key metrics from
Google Compute Engine, Google Kubernetes
Engine, and other Google Cloud services.
It delivers metric-driven visibility into critical
cloud performance and usage metrics. In addition,
you can monitor all your applications that run
on Google Cloud with Wavefront.

The GCP integration is full-featured native

integration offering agentless data ingestion
of GCP metric data, as well as pre-defined
dashboards and alert conditions for certain
GCP services

24
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Wavefront provides in-depth integrations and pre-built

dashboards for 15+ GCP services, including the following:

Google App Google Cloud Google Cloud Google Cloud Google Cloud
Engine Billing Datastore Functions Logging

Google App Google Cloud Google Cloud Google Cloud Google Cloud
Pub/Sub Router Spanner Storage VPN

Google Compute Google Container Google Google ML Google

Engine Engine Firebase Engine BigQuery

See our documentation for more details and the most up to date list of Wavefront GCP integrations.

25
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Pre-built Wavefront GKE Dashboard Showing Nodes, Namespaces, Pods, Containers etc.

26
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Why Choose Wavefront for GCP Observability?

Out-of-the-Box, Real-Time Behavior-Based Alerts Identify Anomalies in

Visibility Into Google Cloud Your Google Cloud Usage or Performance
Wavefront provides turn-key visibility into GCP and resource With a Wavefront, you can quickly compare your current GCP
metrics with pre-built dashboards. We monitor key GCP resource consumption and application performance behavior
services and resources including Google Compute Engine, with previous weeks and turn that query into an intelligent,
Google Container Engine, Pub/Sub, Spanner, Storage, dynamic alert, that gets triggered when something abnormal
Datastore and many more. happens in your GCP environment.

27
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Unified, Full-Stack GCP Monitoring That

Can Significantly Reduce MTTR
Wavefront provides over 200 integrations, so you can easily unify
metrics and alerts from different Google Cloud Platform services,
applications and infrastructure – providing unified, full-stack metrics
view to your DevOps teams including SREs and developers and
leading to significant reduction in MTTR.

28
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

5. How to Monitor VMware Cloud on AWS

VMware Cloud on AWS is an integrated cloud Customers leverage VMware Cloud
offering developed jointly by AWS and VMware.
on AWS for the following solutions,
VMware Cloud on AWS enables you to extend
your on-premises VMware vSphere-based among others:
environments to the AWS cloud while keeping • Seamless cloud migration
a consistent operational model - retaining your • Flexibility of on-demand capacity
familiar VMware tools, policies, and management • Disaster recovery as a service with
as well as investments in third-party tools. VMware Site Recovery

Powered by VMware Cloud Foundation,

VMware Cloud on AWS integrates vSphere,
vSAN and NSX, is supported by VMware
vCenter management, and is optimized to
run on dedicated, elastic, bare-metal AWS
infrastructure. The applications running on
VMware Cloud on AWS have access to a broad
range of AWS services. IT teams can manage
their cloud-based resources with familiar
VMware tools.

29
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

VMware Cloud on AWS enables extending on-premises

VMware vSphere-based environments to the AWS cloud

For details, see the VMware Cloud on AWS documentation.

30
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Easily Monitor VMware Cloud on AWS with Wavefront

The Wavefront VMware Cloud on AWS integration provides visibility
into the health of your vSphere-based environment. The integration
automatically pulls metrics from vSphere and populates the following
dashboards with below-mentioned metrics:

Summary ESXi Host Summary

shows a summarized view provides summarized
of all dashboards for metrics for each ESXi Host
vSphere environment for one or more clusters

Cluster ESXi Host Detail

shows an overview of cluster provides detailed metrics
metrics including number for any ESXi Host
hosts, VMs, memorystores
Wavefront ingests vSphere metrics using the vSphere input plugin
in the cluster. VM Summary
for Telegraf. For details on the metrics and on how to configure
shows summarized metrics
Datastore ingestion, see the vSphere plugin details on GitHub.
for each VM for any ESXi host
provides detailed metrics
about your datastore VM Detail
shows detailed
metrics for any VM

31
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

VMware on AWS Wavefront integration vSphere Summary Dashboard

showing clusters, hosts, VMs and datastores summary

32
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

6. How to Enhance Amazon CloudWatch to get

Unified, Full-Stack, Multi-Cloud Observability
Amazon CloudWatch is a component of AWS and Microsoft Azure. Using Wavefront’s 200+
that provides detailed monitoring data, in near built-in integrations, you can easily visualize,
real time, for AWS resources and workloads analyze, and alert not only on metrics from
running on AWS. CloudWatch is a good start multiple cloud services but from different
for monitoring AWS services, but you’re going workload environments such as Kubernetes,
to need a lot more to assure the quality of your VMware Enterprise PKS, VMware vSphere, web
cloud applications for your users. Gathered from apps, data stores and metrics from instrumented
our customers, here are the five top reasons for applications code.
augmenting CloudWatch with Wavefront:

1. Expand Visibility to All Services of

Your Full Stack, Beyond Just AWS
CloudWatch monitoring is restricted to metrics
from AWS resources. That means CloudWatch
doesn’t aggregate or correlate metrics across
different clouds. Wavefront correlates monitoring
and analytics across multi-cloud environments,
including all the top cloud providers, AWS, GCP

33
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Wavefront Across Major Public Cloud Real-Time Visibility

34
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

2. Combine the Global View Across Your

AWS Environment Health with a Deep
Dive Across Layers
CloudWatch visualizes key single-instance, system-oriented metrics,
but it alone can’t show overall AWS environment health at different
levels of aggregation. Overall, multi-level views enable quick
identification of trouble spots. Wavefront added to CloudWatch
reduces AWS incident resolution from hours to seconds using
interlinked out-of-the-box dashboards for a comprehensive multi-
level view of your complete cloud infrastructure. Wavefront’s globe
chart best illustrates your AWS infrastructure health across regions
worldwide, and Wavefront’s node map chart shows you region-
specific metrics of your EC2 instances. You can drill down to a
troubled instance quickly and identify the root cause at a glance.

Wavefront Interlinked
Multi-View of AWS
Infrastructure

35
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

3. Troubleshoot Microservices
Faster with Metrics, Histograms,
Traces and Span Logs
For troubleshooting applications that use AWS
ECS or AWS Lambda, CloudWatch/CloudTrail
offers only a few metrics and some complex
logs. Consider using Wavefront’s enterprise
observability instead. We offer the unique
monitoring combination of out-of-the-box metrics,
histograms, and traces at cloud scale, with
potentially over a million data points per second.
Metrics tell you system-wide when there is a
problem. Histograms can overcome the limitations
of averages. Histograms help you find problems
with high-speed metrics like sub-second response
times and latencies. Distributed tracing, with
visibility into the entire API call flow, lets you drill Wavefront Enterprise Observability – Metrics,
down and isolate issues across? In? microservices. Histograms, Traces – Go System Wide to Detailed Flow
In short, Wavefront is a single platform that unifies for Faster Troubleshooting
metrics, histograms, and traces and significantly
reduces MTTR.

36
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

4. Enrich Your Analytics

Functionality and Empower
Novice Users
CloudWatch offers less than 20 elementary math
functions that you can use to query CloudWatch
metrics and to create new times series based
on these metrics. Using Wavefront, you can
take advantage of 120+ analytic functions, from
aggregation and raw aggregation, filtering
and comparison, standard time and moving
window time, conditional, rounding, missing
data, metadata, exponential and trigonometric,
predictive and histogram, event and
miscellaneous functions.

Furthermore for users who want to experience

the full Wavefront power but don’t have the time
to build their own analytics models, Wavefront Wavefront AI Genie Automated Real-Time Anomaly
offers AI/ML assistance through AI Genie. AI Detection on Streaming Metrics
Genie uses patent-pending AI/ML technologies
to automate anomaly detection and forecasting
for any time series metrics, for AWS resources
and the application workloads that run on AWS
infrastructures.

37
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Metrics Cloudwatch Wavefront

5. Smarten Your Alerts and
AI/ML NA Holt-Winters, AI Genie (AI/ML-based
Get Notified on Real Issues, anomaly detection and forecasting)

not False Alarms

Alerting Basic Behavior-based smart alerts
No behavior-based smart alerts AI Genie anomaly-based alerts
With CloudWatch, you’re limited to predefined
basic alarms created using limited rules and Analytics 13 basic math functions 120+ functions

math functions. Alert notifications , as well as

Datacenter Type AWS, On-premises Multi-cloud (AWS, Azure, GCP),
the automated actions you’re able to take, are
on-premises
also limited. With Wavefront Alerting there are
no limits. You can fine-tune your alarms using Granularity Basic: 5 min All: 1 sec (real-time)
/ Latency Detailed: 1 min High resolution:
the industry’s most advanced query language
10 sec, 30 sec, x60 sec
and analytics engine, using a single function or Custom: up to 1 sec
a combination of any of 120+ analytics functions
on one or multiple time series metrics. Alert Integrations Only AWS Services (70+) 190+

notification targets include email, PagerDuty,

OS Monitoring Linux, Windows Linux, Windows, Mac
OpsGenie, Slack, and more. You can even use AI
Genie to find an anomaly or predict bottlenecks,
OS Monitoring Complex 1 pps/month
then create an intelligent alert for critical incidents
directly from AI Genie. Retention and 15 months retention 18 months retention
Aggregation Aggressively rolls up: No rolloups
<60 sec for 3 hours
See the table on the right for
1 min for 15 days
a summary of what Wavefront 5 min for 63 days
offers on top of Cloudwatch. 1 hour for 15 months

Ways that Wavefront Expands on CloudWatch

38
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

7. How to Monitor Serverless

Applications in AWS
AWS Lambda, introduced by Amazon in 2014, AWS Lambda and
was the first public cloud vendor offering of a Amazon CloudWatch
functions-based, serverless computing service.
AWS Lambda empowers developers by letting AWS Lambda performs some automatic
them run code in production without provisioning monitoring of Lambda functions on your behalf,
or managing servers. You pay only for the and reports those metrics metrics through Amazon
compute time you consume; there is no charge CloudWatch. To help you monitor your code as it
when your code is not running. executes, Lambda can automatically track:

As a developer, if you’re using AWS Lambda, in • Number of requests

addition to writing AWS Lambda functions, you • Latency per request
also need to think about your monitoring strategy • Number of requests resulting in an error
for your functions running on AWS Lambda. After
all, a function is nothing but the code written by
the developer and with AWS Lamgda, it runs on
the AWS infrastructure automatically. But just
like any other piece of code, it’s the responsibility
of that developer to monitor those functions for
performance and reliability.

39
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

AWS Lambda publishes these metrics to Amazon

CloudWatch. Using the Wavefront Amazon
CloudWatch integration, you can retrieve all
those metrics from? in? Wavefront. Once the
integration is set up and metrics are flowing to
Wavefront, you can run analytics to create charts,
dashboards, and alerts. But the question you
might ask is: “Why pay Amazon to store and
export metrics from CloudWatch?”, and as well,
“Why add a visualization lag by extending your
metric pipeline using CloudWatch?” Instead
of setting up the CloudWatch integration and
exporting AWS Lambda metrics from CloudWatch
to Wavefront, what if your functions can send the
metrics directly to Wavefront with minimal network
overhead, very little additional code, and lower
overall cost.

40
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Metric Name Description

How to Monitor
1 aws.lambda.wf.invocations.count Count of number of Lambda function invocations aggregated at the server.

2 aws.lambda.wf.invocation_event.count Count of number of Lamba function invocations.

Serverless Better 3

4
aws.lambda.wf.errors.count

aws.lambda.wf.error_event.count
Count of number of errors aggregated at the server.

Count of number of errors.

with Wavefront 5 aws.lambda.wf.coldstarts.count Count of number of cold starts aggregated at the server.

6 aws.lamba.wf.coldstart_event.count Count of number of cold starts.

Using the Wavefront Lambda SDK (Python, 7 aws.lambda.wf.duration.value Execution time of the Lambda handler function in milliseconds.

Go, Node.js), you get all the standard Lambda

metrics out-of-the-box sent directly from your To make analysis even easier, the Wavefront Lambda SDK also adds following
Lambda functions to the Wavefront service. As point tags to all the metrics.
an application developer, you only need to focus
Point Tag Description
on writing your Lambda function and to copy/
pasete a little little additional code to monitor 1 LambdaArn ARN (Amazon Resource Name) of the Lambda function.

those functions in Wavefront. There is no need 2 Region AWS Region of the Lambda function.

to invest time and money to set up Wavefront’s

3 accountld AWS Account ID from which the Lambda function was invoked.
AWS CloudWatch integration in order to monitor
4 ExecutedVersion The version of Lambda function.
your Lambda function. There is no need to pay
AWS CloudWatch for storing and exporting those 5 FunctionName The name of Lambda function.

metrics to Wavefront. Also, since the metrics are 6 Resource The name and version/alias of Lambda function. (Ex: DemoLambdaFunc:aliasProd).

coming directly from your Lambda function, they

7 EventSourceMappings AWS Event source mapping Id (Set in case of Lambda invocation by AWS Poll-Based Services).
become available on Wavefront in real time – in
contrast, when polling CloudWatch, you might get Along with the standard metrics, it is also possible to send custom business
them every 5 minutes or so. metrics directly from your Lambda function to Wavefront. Refer to this blog, to
learn more about sending custom metrics from Lambda functions to Wavefront.

41
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Lambda Integration
Sample Code
As an application developer if you’re developing
your Lambda function in say Python, all you have
to do is to decorate your AWS Lambda handler
function with @wavefront_lambda.wrapper, and
voila, all the standard Lambda metrics with the
point tags are sent directly to Wavefront.

import wavefront_lambda

@wavefront_lambda.wrapper
def handler(event, context):
# your code

We’ve included an integration tile in Wavefront to

Wavefront AWS Lambda Dashboards
give you easy access to out-of-the-box Wavefront
AWS Lambda dashboards. Here are a couple of
Note: You can also monitor custom metrics that are sent
example dashboards:
directly to Wavefront using these dashboards.

42
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

8. AWS App Mesh Observability and Analytics

AWS App Mesh is a service mesh based on the Envoy proxy activity metrics for different service clusters. Here are the metrics for
that makes it easy to monitor and control containerized the main components to monitor in a mesh.
microservices.

Category Metrics to Monitor

Critical KPIs to Monitor in AWS App Mesh

Service Mesh
In any service mesh, including AWS App Mesh, critical KPIs to • Uptime • Concurrency
• Allocated memory • Total connections
monitor are as follows:

Health of
• Heap size • Internal flush buffer size
• Live/Currently draining • File reopens
• Health of the service mesh
• Health of your services
Envoy provides the following service and cluster health metrics:
• Performance and activity of your services

Health of
services
• Active listeners
• Active clusters
Because App Mesh uses Envoy proxy as the data plane, • Total healthy/unhealthy clusters

you can determine the health of the mesh from Envoy proxy
health stats. Envoy proxy also provides many health and Envoy proxy provides many performance metrics
Some of the main service performance metrics are as follows:
activity metrics for different service clusters. Here are the
Performance

• Total sent/received bytes to/from a service

of services

metrics for the main components to monitor in a mesh.

• Total buffered sent/received bytes to/from a service
• Http request rate per service
Because App Mesh uses Envoy proxy as the data plane, • 3xx, 4xx and 5xx error responses

you can determine the health of the mesh from Envoy proxy
health stats. Envoy proxy also provides many health and

43
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

AWS App Mesh

Observability with
Wavefront
Wavefront collects health and performance
metrics for App Mesh and its services via the The overall health of App Mesh clusters shown in Wavefront
Envoy proxy. Over 600 App Mesh/Envoy health
and performance metrics are collected OOTB by
the Wavefront integration. Additionally, Wavefront
provides pre-built dashboards for App Mesh,
enabling SREs and developers to quickly visualize,
monitor, and troubleshoot any issues.

App Mesh service performance metrics shown in Wavefront

44
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Identify Sources of Latency with App Mesh

Distributed Tracing in Wavefront
Distributed tracing allows developers and SREs to obtain the trace context. Wavefront, with its Zipkin integration,
visualizations of call flows in large microservices-based distributed ingests distributed traces information from Envoy proxies and
systems. It can be invaluable in understanding serialization, provides distributed traces visualization for App Mesh.
parallelism, and sources of latency enabling developers and SREs
to identify services that need optimization. Envoy reports tracing AWS App Mesh standardizes how your microservices
information for communications between services in the mesh. communicate, helping you ensure high availability for your
However, to be able to correlate the different pieces of the traces applications running in AWS. Wavefront provides built-in
generated by different Envoy proxies, and to get a complete call integrations for multiple AWS services including AWS App
flow across different proxies, the application needs to propagate Mesh. With Wavefront’s App Mesh integration, you get out-
of-the-box visibility into App Mesh metrics and distributed
traces, making it easy to debug and identify the cause of any
communication issues between your App Mesh services.
Furthermore, with Wavefront, you can easily monitor,
correlate and analyze the performance of different AWS
services, applications, and infrastructure components.

App Mesh distributed tracing view in Wavefront

45
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

9. Optimize Both Service

Performance and Cloud Costs
with CloudHealth and Wavefront

As enterprises are adopting more public cloud services, By using CloudHealth and
they are often surprised by growing, instead of shrinking Wavefront together, you gain:
expenses. As a result, insight into cloud cost ROI is essential
to make proper business decisions. CloudHealth by VMware 1) Rightsizing Analysis
helps you manage costs, usage, security, and governance
In many large cloud environments, it’s common to find
across your entire cloud environment. CloudHealth analyzes
severely underutilized infrastructure that drives up costs.
native metrics from cloud providers out-of-the-box, but in
Using CloudHealth In many large cloud environments, it’s
many cases, that is not enough for a 360 view. If you use
common to find severely underutilized infrastructure that
CloudHealth and Wavefront together, CloudHealth can take
drives up costs. Using CloudHealth and Wavefront together
advantage of Wavefront’s vast real-time, highly granular,
can help solve this problem. The Wavefront platform
multi-cloud (AWS, Azure, GCP) infrastructure usage metrics,
collects granular performance data from the application,
and provide effective recommendations for cost savings of
CPU, memory, network, and disk usage from instances
your service.
and virtual machines on AWS, Azure, and GCP. This tag
enriched data is then ingested and used by CloudHealth

46
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

to do a rightsizing analysis. CloudHealth provides specific you do rightsizing analysis by function, so you can quickly find
recommendations around downsizing, or even terminating which departments or teams need to adjust their behavior
instances, including the suggested instance type for your when provisioning new infrastructure.
services’ optimal performance. By following these rightsizing
recommendations, for example, downsizing an instance
that is underutilized, you can save time and money without 3 - Balance Cost with Performance
interruption to the service.
Not only can CloudHealth help identify assets that are
underutilized and that can be downsized for cost savings, but
“I can easily identify which Elastic it can also help with tracking and managing expenses in the
Load Balancers in a business unit cloud. You can allocate and amortize costs and can forecast
do not have an AWS EC2 instance spend across departments and teams. Data and analysis from
attached.This is just money on the CloudHealth can be extended further into the Wavefront
table that we can easily save using platform and correlated with application and infrastructure
Wavefront.” performance. Thus, DevOps teams can optimize cloud
service performance and set real-time, proactive alerts for
- Dieter Matzion, Staff Business
configuration changes and cost overruns.
Systems Analyst, Intuit

Together, CloudHealth and Wavefront, enable you to easily

2 - Visualization of Key Assets improve efficiency and reduce cost without sacrificing
productivity or performance.
Through a Business Lens
Together, CloudHealth and Wavefront give you real-time visibility
into the performance of your service and enable you to zero in on
the cloud data you care about by department, application, line of
business, or any logical business grouping. This visibility also helps

47
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

How Intuit Saved Millions Using

CloudHealth and Wavefront
After evaluating multiple vendors for their cloud “CloudHealth feeds data into
adoption transformation, Intuit chose CloudHealth Wavefront in an aggregated fashion
for true cloud cost management and Wavefront that is extremely powerful.”
for real-time performance monitoring, based on
– Dieter Matzion, Intuit Cloud Architect
advanced streaming analytics.

Intuit integrates Wavefront and CloudHealth Intuit also forwards aggregated CloudHealth cost
through the CloudHealth API, feeding metrics to Wavefront for correlated monitoring
Wavefront’s tag-enriched resource usage and analytics-driven alerting. Specifically, Intuit
data into CloudHealth’s cloud cost-savings uses the Wavefront powerful alerting engine to
recommendations, saving millions of cloud cost alert on cloud budget overruns.
dollars annually. Using rightsizing analysis, Intuit
was able to move over half of its workloads to run
smaller EC2 instances, saving them over $100,000
a year.

48
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

49
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

10. Top 5 Lessons Learned from

Customers with Multi-Cloud Monitoring
1 - Choose an observability solution that
is builtfor cloud and modern technologies

Over a period of time, we have seen our Common feedback from such customers was
customers modernize their services and adopt that it helped them to use an observability
emerging technologies. For instance, our solution (like Wavefront), which has support for
customer 8×8’s cloud service has evolved emerging technologies, including multi-cloud and
dramatically from being a monolithic piece of Kubernetes. With Wavefront, customers don’t
software. In line with industry trends and best have to worry about building observability support
practices, 8x8 moved to a services model, for emerging technologies. Traditional monitoring
and, ultimately, to microservices. We have and APM solutions, on the other hand, often lack
also seen many customers embark on the multi-cloud and emerging technologies support.
journey of containerization and adopt different
flavors of Kubernetes, including Amazon EKS
andOpenShift.

50
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

2 - Siloed tools increase MTTR and slow down productivity

Before using Wavefront, many of our customers had so many tools, not many people knew how
had one common problem. They were using to use every tool. (“You just cannot expect them
many siloed monitoring solutions and didn’t have to learn every tool when you have over half a
a unified/first pane-of-glass view across different dozen monitoring tools,” they said.) Furthermore,
clouds, full-stack, and tools. This lack of the first none of the teams had a unified view across all
pane-of-glass view led to long MTTR and missed the tools. So, whenever an incident happened,
SLAs. they would have to look through several tools
to troubleshoot a problem. This led to increased
Let me walk you through an example of one MTTR and missed SLAs. Additionally, when an
of our customers, one of the biggest Indian incident happened, it resulted in several alerts
e-ticketing platforms. They mentioned that their from different tools leading to alert storms. Very
SRE and DevOps team was using over half a often, an important alert was missed in this array
dozen siloed monitoring solutions. Because they of false positives leading to increased chances of
missing SLAs.

51
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

3 - Maintain sanity in cloud and save millions by controlling

cloud cost and by proactively alerting on budget overruns

Many enterprises adopt a multi-cloud strategy to integration. Intuit, one well-known Wavefront
reduce costs. But they are often surprised to see customer, uses CloudHealth as a caching engine
their cloud cost go up due to different cloud’s that retrieves and stores telemetry information
pricing models, varied types of instances, and from AWS. With the Wavefront CloudHealth
underutilized workloads. If you are using multi- integration, they feed this telemetry information
cloud, it is essential to select rightsized instances into Wavefront and use Wavefront to monitor and
and avoid budget overruns with proactive alert on budget overruns.
alerting.

Wavefront customers have saved millions of

dollars by using rightsized reserved instances
recommended by the CloudHealth and Wavefront

52
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

4 - Open-source monitoring solutions

don’t support cloud-scale and reliability

Many enterprises move to the cloud to take with more horsepower, or you need to spin
advantage of the the ability to scale up at a up another Prometheus instance. Both of the
moment’s notice. Reliability is one of the reasons tasks need engineering team effort and cycles,
for adopting multi-cloud. But while adopting leading to teams getting slowed down with scale.
multi-cloud for all these benefits, enterprises Soon engineering teams find they’re spending
often forget that their open source monitoring more and more time on scaling, patching,
solutions lack support for scale and reliability. and maintain their open-source monitoring
For instance, consider this scale-limiting use- solutions and not having that time to focus on
case of the open-source monitoring solution innovating and scaling their business. Reliability
Prometheus. As a customer’s footprint grows is another big concern with open-source solutions.
in the cloud, so do their metrics volumes as In fact, one of the biggest reasons for Boxever
engineers add more and more necessary to switch from Prometheus to Wavefront was
telemetry. But when one Prometheus node gets that Prometheus lacked support for a real high
too busy with the increased number of metrics, availability configuration.
it either needs to be hosted on a bigger instance

53
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

5 - To achieve exceptional SLAs, empower development

teams with the ability to monitor their SLOs

As enterprises adopt a distributed microservices- metrics and analytics information with different
based architecture, the overall customer teams. Wavefront offers teams a centralized
experience depends on performance and observability solution that different teams can
availability of many different microservices. These easily share. With Wavefront, engineering
microservices are often owned by different teams can easily monitor their SLOs, taking
development teams. As mentioned by Edgar complete responsibility for the performance and
Nidome, Principal Architect at 8x8, it is critical reliability of their services. In short, make sure
to empower all development teams to take to give development teams the ability to have
ownership of their metrics and performance. lightweight instrumentation, unify objectives and
Every development team should be able to empower them to get to addressing their SLAs.
measure and share the quality of service that they
provide with other teams. But often, engineering
teams find it hard to standardize how to share

54
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

11. Why Customers Choose Wavefront

for Multi-Cloud Observability and Analytics
Wavefront delivers turn-key observability across Here are the top three reasons why
AWS, Azure, and GCP. Well-known enterprises customers adopt Wavefront for Multi-
such as Intuit and Workday and successful Cloud Observability and Analytics
cloud-native companies like 8x8 and Reddit use
Wavefront for monitoring their applications and 1 - Real-Time Observability and
infrastructure across different public clouds. Analytics Across All Major Public
Cloud Platforms
“Wavefront provides unparalleled
visibility across all our cloud With Wavefront, you get out-of-the-box, real-
microservices helping hundreds of time observability across all major public cloud
developers become more productive platforms – AWS, Azure, and GCP – with many
focusing on innovation, while public cloud integrations and pre-packaged
enabling 8x8 to deliver exceptional dashboards showing key metrics for all major
SLAs and eliminate issues with our cloud services. You don’t have to use different
cloud services.” siloed products for different cloud providers.
Monitoring multiple cloud services and platforms
- Dejan Deklich, Chief Product Officer, 8x8 in a single solution has never been so easy!

55
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Real-Time Observability Across AWS, Azure, and GCP provided by Wavefront

56
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

2 - Unified, Full-Stack
Observability that can
Significantly Reduce MTTR

Wavefront provides over 200 integrations.

You can easily unify metrics, traces, and
histograms from different AWS, Azure,
and GCP services, applications, and
infrastructure. The result gives a unified,
full-stack view to your DevOps teams
including SREs and developers and leads
to a significant reduction in MTTR.

57
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

3 - Optimized Service
Performance and Cloud
Cost with Wavefront and
CloudHealth
With Wavefront, you can easily monitor
cloud consumption and cost metrics and
convert them into data-driven insights
and alerts for reducing your cloud cost.
Furthermore, by using CloudHealth and
Wavefront together, you get effective
recommendations for cost savings of
your service that use CloudHealth’s cost
governance platform and Wavefront’s
vast real-time, highly granular, multi-cloud
(AWS, Azure, GCP) infrastructure usage
metrics.

58
Section I: Growth & Section II: Guide to Achieving Section II: How to Excel
Challenges of Multi-Cloud Multi-Cloud Observability Across at Multi-Cloud Observability
Observability AWS, Azure, and GCP Across AWS,Azure, and GCP

Author
Chhavi Nijhawan
Chhavi is a Product Line Marketing Manager at Wavefront by
VMware. Before Wavefront, she worked at New Relic, SnapLogic
and Cisco, where she led product marketing and technical
marketing. She has over 10 years of IT industry experience.
She is also an AWS certified solutions architect.

59