Cloud computing is one of the biggest and most disruptive IT technologies of all time, which has significantly changed the way IT is consumed by enterprises, both small and large. The widespread adoption and popularity of cloud computing services can be attributed to the numerous benefits the technology delivers, including low cost and accessibility of data. In fact, many organizations have become so comfortable with the cloud that they are considering to make the move from a single cloud to a multi-cloud environment.
Adopting a multi-cloud environment basically means using more than one public cloud. The goal is to distribute a business’ assets, applications, software, redundancies, and other computing resources across multiple service providers to enjoy improved performance, cost savings, and resiliency, while mitigating risks associated with single cloud environments at the same time. But while multi-cloud offers new possibilities to businesses, the question of how these complex systems perform still remains. Given the complexity and inherent unpredictability of Internet connectivity, robust visibility solutions are essential in order to create safe and secure cloud environments.
SourceForge had the chance to speak with Alex Henthorn-Iwane, the Vice President of Product Marketing at ThousandEyes, the Network Intelligence company, to discuss the shift towards multi-cloud and how ThousandEyes can help organizations gain the visibility needed to ensure performance in their multi-cloud deployments.
Cloud Deployment Models for Businesses
Alex Henthorn-Iwane, the Vice President of Product Marketing at ThousandEyes
As cloud computing gained widespread acceptance among various organizations and industries, so did the number of cloud computing models and deployment strategies. Various cloud services and deployment methods have emerged to address the specific needs of different users. There are three main models for cloud computing: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS). Cloud computing deployment models meanwhile include public, private, community, and hybrid clouds.
“In regards to cloud strategies, when you talk about single cloud, hybrid cloud, or multi-cloud, most of the time you’re talking about the infrastructure as a service (IaaS) as opposed to SaaS,” said Henthorn-Iwane . “And so one of the things that happens is that, within a company, is that they start moving outside of a data center, and this is a single cloud strategy.” According to Henthorn-Iwane, a single cloud strategy, in most cases, is a starter step. It is ideal for companies with a very limited number of applications to host. “The caution here is to not get locked in,” advised Henthorn-Iwane. “It’s similar to companies outsourcing a business process, and then later, companies realizing they can’t get what they want. his can happen when you use mass-produced cloud services.”
A hybrid cloud environment meanwhile involves an interaction or a share of a load between a business’ own data center and one more public cloud provider. “Hybrid clouds give you flexibility. What happens is that these hybrid cloud environments can often go in cycles,” stated Henthorn-Iwane. “Sometimes you see a company pushes everything (their production workload, their engineering workload, etc.) to a public cloud provider, and then they find out that they want more control over select things…so then some of these things come back in,” he said.
Henthorn-Iwane explained that with large public cloud providers, SLAs are highly standardized so most businesses simply have to accept the SLA given to them, without the ability to tune it. For public cloud service providers, having such a highly standardized SLA works since they can build and execute their services well at massive scale, which delivers economies of scale. But Henthorn-Iwane argues that for businesses, a one-size fits all SLA can introduce challenges because if it does not meet the company’s requirements for fine tuning performance and control, then they may have to re in-source certain pieces.
Multi-cloud, on the other hand, is best for companies that are nurturing different types of apps or have different geographic placements. Henthorn-Iwane said, “In multi-cloud, there is competition to take advantage of. And what’s interesting about multi-cloud is that there is a reason why major enterprises do it: to avoid getting locked in to a single source.” To do this effectively teams will adopt only a common denominator set of services across multiple cloud that protects them from vendor lock in, added Henthorn-Iwane.
But beyond the economic implications and concerns about vendor lock-in, Henthorn-Iwane said that the choice to adopt a multi-cloud could also relate to the fact that some clouds perform certain things better. “There are cloud offerings that may offer security capabilities and/or pre-built integrations that can meet the need of the development or production teams better than others, and so teams will start diversifying based on these needs,” he added. “And even though there is always a concern about lock-in, those capabilities may simply be required, so this ‘specialization’ keeps the tech stack robust.”
Gaining Visibility Across a Multi-Cloud Environment
Multi-cloud offers so many benefits, including easy deployment of new IT services. Unfortunately, a multi-cloud environment also has its downsides, including more silos and complexity. When combined with the fact that cloud involves the Internet, that complexity results in a pronounced lack of visibility.
“Let’s consider first and foremost what customers do in the cloud and their multi-cloud deployments,” said Henthorn-Iwane. “Typically they’re building service-based architectures and microservices, which means there’s a ton of API calls. Let’s say as an example that we have an application in a data center with APIs that are calling services built in AWS, and also using web APIs to call cloud-based services offered by SaaS vendors. There’s a ton of API calls happening from all sorts of different directions. The network paths taken across the Internet by various inter-service communications can be different. Any one of those can traverse multiple, different networks. If the application is critical to the business, you But you need to know how each one of these communication paths is performing in order to see if there are any performance impacts on the application.”
“Now think about how this looks end-to-end. A remote worker at home is using Salesforce, but to get there, has to go through a secure web gateway (SWG) provider, so that’s one Internet path. Salesforce may be calling additional cloud-based services, which represent other Internet paths. The problem is that in general the Internet is a black box for network engineers. When you have multi cloud-based applications, multiple API calls, …now there are all these blind spots,” explained Henthorn-Iwane.
Henthorn-Iwane said this is the picture they are trying to communicate to enterprise IT customers, showing them that the applications that their business processes depend on are reliant on networks and services they don’t own and can’t see with traditional tools. “And ThousandEyes exists to help solve this problem,” he emphasized.
“With our agents in over 150 cloud nodes across the world, combined with enterprise agents you can deploy in your data center and IaaS, plus endpoint agents you can deploy on employee devices, you can monitor and see end-to-end from the remote employee to the AWS instance to the API gateway provider,” said Henthorn-Iwane. “You can also use the cloud agents to see how customers are connecting to your website, how DNS is performing for them,and the Internet paths they are supporting all that communication… That way businesses can cover all the legs of communication and shine light on them. That way, if something breaks, you know where it’s happening, who to contact, and what data to share to fix the problem. Even though you may not own all the stuff, you still own the outcome. A new approach to Internet and cloud visibility is very important in this regard.”
Why ThousandEyes
A trusted network intelligence company, ThousandEyes is committed to giving businesses immediate visibility into their own network, the internet, and the cloud, to improve application delivery and reduce service interruptions. “We make it easy to activate monitoring agents everywhere you want visibility from,” said Henthorn-Iwane. ThousandEyes uniquely combines and correlates various types of data from all the agent, and turns that data into easy to understand visualizations about end-to-end performance.
“For example, you can see how an API call is performing from AWS to a cloud-based API gateway, and we give you the precise Internet path information. The monitoring agents are very lightweight and low overhead. With us, you are not dealing with enterprise software installs. And we use algorithmic approaches so the way we present the data addresses what matters to you. So if you are seeing a performance issue, you can view the http layer performance, availability and response time; then from the same page switch to look at end-to-end network metrics like loss, latency and jitter; then switch to a full path visualization including every Internet network (Autonomous System) the path crosses..”
For Henthorn-Iwane, there are three things that make ThousandEyes a strong competitor in the network monitoring market:
- Uniqueness. “We have the unique ability to see across the internet,” stated Henthorn-Iwane. “If you want to see how performance is being affected and where the problem lies, ThousandEyes does the job. Our ability to show you detailed paths across the internet and how they’re impacting your users and cloud-based apps, is very special.”
- Powerful visualization. “Second is that that we have this kind of super powerful visualization to harvest many sets of data, and then make it visual and easy to understand,” added Henthorn-Iwane. “Most engineers, when they see this visualization they say it’s both intuitive and powerful.”
- Easy data sharing. “Third is the fact that we make it easy to share that data with your vendors, ISPs, and across teams,” said Henthorn-Iwane. He explained that a lot of times, traditional network monitoring tools create siloed sets of data that end up being black boxes themselves., It means that in the end, only the network team can get anything out of these tools. This makes it very difficult to share data of a whole event. ThousandEyes, on the other hand, make it easy to share snapshots of detailed data and visualizations that provide context across teams and between partners and vendors. “We allow you to useshared data to drive more productive interactions with your providers. It’s an ecosystem of sharing data,” stated Henthorn-Iwane.
About ThousandEyes
Established in 2010, ThousandEyes is a leading network and performance monitoring company. They offer a powerful Network Intelligence platform that gives organizations visibility into every network they rely on. This enables faster application delivery, enhanced end-user experience, and improved infrastructure investments. With headquarters in San Francisco, California, ThousandEyes is trusted by top global companies and members of the Fortune 500 to boost the performance and availability of their business-critical applications.