What news from AWS re:Invent last week will have the most impact on you?
Amazon Q, an AI chatbot for explaining how AWS works.
Super-fast S3 Express storage.
New Graviton 4 processor instances.
Emily Freeman leaving AWS.
I don't use AWS, so none of this will affect me.
Cloud Services / Containers / Kubernetes

How to Treat Servers and Software as Cattle, Not Pets

How to get started down the path of cloud native computing.
Jan 22nd, 2020 9:41am by
Featued image for: How to Treat Servers and Software as Cattle, Not Pets

Humio sponsored this post.

John Porcaro
John joined Humio in 2019, working as part of the marketing and developer relations team in Seattle. Recently, he worked at PwC as a consultant for enterprise companies making digital transformations in business strategy, technology and user experience. Prior to that, he led CRM and online community for Xbox at Microsoft. He enjoys writing, hanging out online and spending time in the Cascade mountains.

As software and services move to the cloud, the days of fawning over individual servers and software installations as though they were fussy pets is over. By treating cloud-based servers uniformly — as though they were cattle — engineers implement policies and protocols that boost security and efficiency. The resulting standardization of these processes makes it easier to manage distributed systems and scale efficiently, because multiple servers and instances of software can be managed and updated from one interface, rather than one at a time.

By centralizing system management, IT and DevOps can keep track of the whole herd by enabling complete observability of the IT environment. By understanding what’s happening across networks, servers and apps, administrators are better prepared to deal with unexpected incidents as they occur. They have the ability to explore terabytes of data and discover the root cause of any issue, significantly reducing the time to recovery.

In this article, we describe how organizations can make standardizing management of distributed systems easier in hybrid and multicloud environments. We’ll explore the benefits and challenges involved and offer some solutions to improve the health of the entire IT infrastructure.

Why a Move to the Cloud Can Be Complicated

Transitioning to the cloud requires wide-scale adjustments for engineers as they adjust to using new paradigms and a new set of tools. Even though the benefits can easily outweigh the problems, it will take some work to fully embrace the journey. However, while there are complications that come with the shift to the cloud, they are easily outweighed by the benefits gained.

Learning and Adopting a Microservices Architecture

Because the cloud is built on microservices rather than the familiar monolith, engineers need to develop a familiarity with a new set of tools. Gaining proficiency in newly-developed languages like Ruby, Python, or Go helps with the challenge. Those that understand cloud native architecture and open-source software have an advantage, since most SaaS apps are community-sourced and supported.

Paying ‘Rent’ to Cloud Providers

Moving to the cloud requires investing in a new set of services. Rather than maintaining a datacenter with a private set of servers, companies make the shift to subscribing to services online. There will be a transition time when they have both on-premises hardware and software and cloud-based services and applications. This period of maintaining a hybrid cloud requires investment upfront, but most companies soon see their budgets level out as they reach the end-of-life of physical hardware and the declining amount of maintenance and support it requires. And unlike maintaining physical hardware, companies only pay for what they need.

Purchasing New Software

Deploying new software usually brings the challenge of acquiring new licenses. Rather than purchasing a piece of software with a fixed price contract, cloud vendors rely on subscriptions and variable prices depending on the volume of data processed or the number of users connected. Most traditional on-premise software providers already have SaaS versions of the same software, so making the move should be relatively problem-free. And for those looking to try a new SaaS solution, most have free trial versions and many are completely open source with no licensing fees for the software.

Moving to the cloud can have a positive impact on budgets because they usually require less maintenance and many have low-cost license fees.

Overwhelming Amounts of Data

As companies increase the services they offer employees and customers, they find the amount of data skyrocketing. A simple task like monitoring and maintaining logs created by hardware, software and system events can tax traditional tools used by operations or security teams. These tools may not be able to handle the increase in collecting, indexing and storing terabytes (or more) of data a day in systems designed to handle gigabytes of data.

The good news is that cloud platforms are built for this kind of volume and new companies are architecting all-new services built from scratch to take advantage of advances in distributed microservices.

For example, Humio is built using Kubernetes containers. It manages massive volumes of incoming data with Docker, Kafka and Zookeeper. Rather than indexing and storing data in a database before it’s able to be accessed or searched, it is designed with an index-free architecture that allows streaming data to be instantly available in memory of a state machine, so dashboards, alerts and searches can be performed with zero latency. And because data doesn’t need to be indexed before it can be accessed, any data can be searched without pre-defining schemas or having to decide what will and what won’t be indexed.

Large amounts of streaming data can also be optimized for cloud environments. For example, Humio compresses unindexed data by five to 15 times or more, so data storage is optimized, bandwidth is preserved and searches are made in a fraction of the time.

Ever-Increasing Data Security, Privacy and Compliance Requirements

As customers grow familiar with sharing personal data, governments and industries have taken notice. New global, federal and local regulations are multiplying. Policies like GRC, GDPR, CCPA, to SOC, HIPPA and dozens more help protect the security of data and the privacy of consumers. Complying with ever-increasing regulations is required, whether organizations are running software on-premise or in the cloud.

Fortunately, every part of the cloud is designed to protect data throughout its lifecycle. PaaS vendors ensure that data is kept secure while at rest, when in motion and when in use. Cloud solutions are built with privacy and compliance in mind. Modern software is built to enforce appropriate authentication and permissions. Platforms and software are built to help users remain compliant, so relying on them can help save headaches and legal trouble.

With standardized servers on the cloud, software can be regularly updated to meet new regulations and can go into effect with just one install. Users can institute security policies as code in Git and then can enforce them across multiple clusters, creating uniform security compliance across all technologies, both on-prem and in the cloud.

Using a modern log management tool like Humio helps keep systems secure by making all activity across the network observable. This ensures that information is available immediately to act on security threats and potential vulnerabilities, whether the data is coming from hardware, the network, from apps in development, or from customers or the public.

When an incident occurs — whether it’s unauthorized users trying to access a system, employees sharing sensitive data, or misconfigured HTTP headers exposing secure data — alerts can be configured to automatically run webhooks to instantly shut it down and prevent an incident. And even when a problem is discovered later, having access to all logs and event data makes finding exactly what happened and why as easy as running a few searches, even if the data is months old.

Learning New (Better) Ways to Monitor System Health

Looking at dashboards and reports that have been used for years provides some peace of mind, but all too often, problems occur outside of what’s being reported. Traditional monitoring fails at providing complete observability. They were designed to share predefined metrics from a single application and the data isn’t correlated with other events happening throughout a distributed system.

Because nearly everything running in the cloud uses accessible APIs, it’s easy to collect everything that’s occurring across the environment and manage it centrally. SaaS tools like Humio capture every streaming log, metric, trace, or event across the system, so users see what’s happening as it happens. Because data is massively compressed and efficiently stored, it’s available to explore to answer any question and discover the root cause of any issue.

Centralization Makes Complete Observability Possible

As someone responsible for managing the technology environment — whether it’s a startup, a mid-sized company, an agency, an NGO, a division of a company, or for a global enterprise — maintaining observability is the most important thing ITOps and DevOps professionals can do.

Centralizing the collection and monitoring makes observability possible and using PaaS and SaaS solutions facilitates observing, learning and reacting to incidents or bugs.

Three Steps to Get Started

Ready to get started?

Step 1: Pick the Cloud Provider that’s Right for Your Business

At Humio, we work with IBM as our cloud provider of choice and make use of its tools in IBM Cloud Paks. Humio’s log management SaaS is also available as part of the IBM Cloud Multicloud Management Pak.

By the way, Humio also works on-premise, across hybrid clouds and with other popular cloud providers.

Step 2: Design Your Environment for Observability

As you get started trying SaaS apps and moving data, look for ways to migrate existing tools to the cloud. Most apps running on-premise have SaaS solutions that outperform older software.

As you install new services and design applications, make sure they are publishing good log data. As each service comes online, collect all the log data so it’s there when you run into problems. That way, you’re prepared for the “unknown unknowns.”

Step 3: Learn and Explore

Learn everything you can. There are countless podcasts and videos that’ll help explain the cloud native architecture and the benefits of moving to the cloud. We invite you to start by listening to our ongoing podcast series, The Hoot. You’ll also find informative talks and demos online, like on the Humio YouTube channel. And what self-respecting software company doesn’t have a blog, like the Humio blog? Subscribe, visit and share your thoughts as comments.

Join online communities like Slack (our engineers are active in Humio’s Slack channel). You’ll find inspiration, tips for getting started and answers to questions you might have. And share your experience as you go.

Get started with Humio by downloading a free 30-day trial (on-premise or Cloud). See how easy it is to collect data from every part of your infrastructure. Connect to data using popular APIs and prebuilt integrations for Rsyslog, Fluentd, Nginx, Zeek, Javascript, NodeJS, Logstash, Filebeat and others. Our engineers love solving problems and they’ll help you get up and running in no time.

See how Humio’s modern architecture redefines what is possible with log management. After a few searches, you’ll understand why we’re excited about how easy and how fast it is. Find out more by visiting our website:

Once you make the move to the cloud and begin to understand the benefits of moving to a flexible, powerful, cloud native environment, you’ll love the freedom that comes from achieving complete observability and treating your servers and software like cattle rather than pets.

Feature image via Pixabay.

Group Created with Sketch.
TNS owner Insight Partners is an investor in: Docker.
THE NEW STACK UPDATE A newsletter digest of the week’s most important stories & analyses.