For most businesses, the decision of “to cloud or not to cloud” has been answered with a resounding yes. The burning question now is which cloud provider to use, and that has no easy answer because every cloud provider has different pricing models and strengths and weaknesses.
The reality is that almost every business of any significant size will embrace the concept of multi-cloud where a combination of Microsoft Azure, Amazon Web Services (AWS), Google Cloud Platform (GCP), and others are used.
Multi-cloud rapidly becomes the norm
Several factors are driving companies to multi-cloud — some business related and others technical that will continue to force this architecture upon companies. Pricing considerations and avoiding being beholden to a single vendor are a couple of issues that business leaders might be concerned with, whereas tools, innovation and functionality are factors that developers and IT individuals would prioritize when making a choice.
Given no two cloud providers are the same, it’s almost manifest destiny that multi-cloud become the norm.
Traditional network management tools don’t cut it in multi-cloud environments
The challenge with multi-cloud is that as businesses leverage resources from multiple cloud providers in a wide range of locations, troubleshooting application and network problems becomes impossible to do with traditional tools. Current network troubleshooting toolkits contain things like packet capture, flow-based analysis and SNMP monitoring, but these all leave big blind spots as they have no correlate information cross-cloud and very few can actually monitor in a cloud.
Sometimes a high-level engineer might be able to look at the data and understand the implications, but this would require a very seasoned network professional whose time would likely be better spent doing something other than looking through packet dumps and log files.
All of the major cloud providers have their own cloud resident monitoring tools, but the problem with these is that they monitor only what’s happening in the cloud itself and have no ability to extend out or to other clouds. I’ve heard the analogy that troubleshooting multi-cloud is like finding a needle in a haystack, but It’s more like finding one specific needle in a stack of needles.
ThousandEyes introduces multi-cloud Network Intelligence
This week, internet monitoring vendor ThousandEyes extended its Network Intelligence product to multi-cloud environments. The company has pre-provisioned IaaS vantage points, including 15 AWS, 25 Azure, and 15 GCP regions, giving them outstanding visibility into how specific cloud providers are performing in various geographies.
ThousandEyes can also do agent-to-agent testing between clouds, giving IT the ability to measure inter-region, hybrid, inter and intra-cloud performance. Plus, the data can be used to map network paths, understand traffic flows, and monitor connectivity between on-premises and cloud data centers.
The cloud agents are the main source of the data, but ThousandEyes also collects a wide range of other data, including enterprise and endpoints agents deployed in their customer’s networks and data centers. Some customers have been using the cloud agents to do their own cloud monitoring previously, but the new service extends that to multiple clouds.
ThousandEyes Network Intelligence provides visibility into every network path
The image below shows how ThousandEyes can be used to visualize every service delivery path in a multi-cloud environment to pinpoint the root cause of service outages. This path visualization shows every service delivery path from AWS and Google Cloud regions to the Chicago area in a single map. An outage on the Zayo network has impacted connectivity to Amazon’s US-East-1 region from the Chicago area. Users in Chicago may not be able to reach services hosted in this AWS region.
ThousandEyes multi-cloud monitoring isn’t just for troubleshooting, as it can be used to help organizations better plan their deployments. For example, a developer in London could see if AWS or Google Cloud offered better performance over a variety of network service providers and then choose the best one. This might be different than the same combination in another part of the UK.
Don’t guess – take a data-driven approach to multi-cloud
In many of my posts on Network World, I’ve made the comment that “you can’t manage what you can’t see,” and that is certainly true for multi-cloud environments. Without visibility into every path, network managers are making a critical business decision based on nothing more than gut feel or faith. This can have a negative impact on worker performance and even company revenue because of poor performing web services. With ThousandEyes, the business can make a data-driven approach to planning multi-cloud deployments and make the best decision for the company.
Note: ThousandEyes is a client of ZK Research.