After attending the OpenStack Summit in Boston, we saw first-hand both the growing prominence of OpenStack and also the growing challenges. One CIO told us “the combinations and permutations of problems in OpenStack are endless, so traditional monitoring doesn’t work.” Another told us, “OpenStack monitoring is so hard because it’s still a black box. I can’t see well into my environment to figure out the source of what’s going wrong.”
These kinds of comments underscored the fact that while we may have finally reached a critical mass of business digitalization, businesses are still lagging when it comes to adopting the latest technologies that could ease the cloud transition. OpenStack is a complex and evolving system that continuously generates vast amounts of metrics and log data. For IT teams to capably monitor OpenStack and get the visibility they crave, they need to look outward to another technology: artificial intelligence.
Here we will cover three of the top challenges with traditional OpenStack monitoring and how AI would overcome them.
1) Disparate Structures and Folders
Many OpenStack deployments are made up of several, disparate locations and monitoring across a broad range of facilities, clouds, hosts, and VMs is required. Inside OpenStack environments, many software components may be in use and each must be tracked to make sure that the delivery of business services is uninterrupted. These different structures make it challenging for IT Operations teams to get meaningful insights on deployments across the environment.
The beauty of AI is that it reduces comprehensive monitoring to not just one tool, but one dashboard. IT Operations teams would get centralized visibility across the Compute, Network, Storage, Identity, Telemetry, and other components of their OpenStack deployment. That would include all applications, microservices, and the full picture of dynamic infrastructures in real-time, including automatic updates on any changes to your environment like VM migrations.
“The combinations and permutations of problems in OpenStack are endless, so traditional monitoring just doesn’t work.” said one attendee at OpenStack Summit Boston 2017
2) Dynamic Environments
Most monitoring services are static rather than dynamic and don’t grow and shrink in step with elastic services, requiring extensive human intervention. In OpenStack environments, containers and VMs can be deployed almost instantaneously and may only run for minutes or hours. Compare that to legacy infrastructures where systems can run for months or years. When new containers spin up regularly throughout the day at OpenStack’s whim, IT Operations teams need to continuously analyze the new logs and metrics - something they may not be prepared for.
An AI-powered monitoring solution would actively monitor all of the components that could impact OpenStack performance in real time so that IT Operations teams get instant, deep-level visibility into each microservice, including proprietary application logs. It would grow and shrink in step with all OpenStack services, including the underlying compute, storage, and network infrastructures and OpenStack data-plane components. Even with dynamic deployments, IT Operations teams would get breadth and depth of visibility across the environment and would be able to see how they’re interconnected if issues pop up.
3) Legacy Mixes
It doesn’t help that organizations are usually running a mix of traditional infrastructures and cloud services in addition to OpenStack. According to OpenStack.org, 82% of OpenStack deployments are running in parallel with other cloud platforms. More specifically, 77% reported that their deployments are also interacting with Amazon Web Services (AWS). Unfortunately, open source monitoring tools can’t capably monitor hybrid IT environments. Any organization that adds OpenStack-specific monitoring tools has good intentions, but the result will be multiple, disparate monitoring tools - with all the additional costs and resources.
Artificial intelligence would give digital businesses unprecedented insights into their hybrid cloud setup, including the correlation of logs across the nodes of disparate deployments and the root cause tracing behind any alert. With instant visibility into the relationship between the host logs and network metrics, teams would be able to see how problems on a given host affect the whole network (and possibly other hosts). IT Operations teams would be able to determine exactly which part of the system has an error and get a recommendation on how to resolve it.
The Promise of AI
The CIOs who told us that traditional monitoring is no match for OpenStack and that OpenStack is a black box are exactly right, but OpenStack monitoring doesn’t have to be that way. If OpenStack is matched with an AI-powered monitoring tool that can get instant deep-level visibility into each microservice, grow and shrink in step with services, and connect with hybrid deployments, IT Operations teams will get the monitoring they need to catch issues before they affect the delivery of business services.
Even more, AI can give IT Operations teams visibility across their entire stack in one dashboard and highly intelligent alerts based on any metric or log structure. They’d be notified when it matters - no more noise - and they’d get full operational insights without any manual configuration, including the instant detection, isolation, and root cause analysis of any anomaly. Every cloud transition needs the right tools, and one tool IT departments shouldn’t go without is AI.
Loom Systems delivers an AIOps-powered log analytics solution, Sophie,
to predict and prevent problems in the digital business. Loom collects logs and metrics from the entire IT stack, continually monitors them, and gives a heads-up when something is likely to deviate from the norm. When it does, Loom sends out an alert and
recommended resolution so DevOps and IT managers can proactively attend to the issue before anything goes down.
Get Started with AIOps Today!