Red Hat introduced Red Hat Enterprise Linux 9, the Linux operating system designed to drive more consistent innovation across the open hybrid cloud, from bare metal servers to cloud providers and the farthest edge of enterprise networks.
DEVOPSdigest asked the top minds in the industry what they think AIOps can do for DevOps and developers. Part 2 covers AIOps capabilities such as observability and automation.
Start with What Can AIOps Do For DevOps? - Part 1
The security, identity, and privacy issues that come with the widespread use of the cloud as well as the scaling and resiliency needed for most web apps needs the toolset that AIOps promises in the form of automation. Incorporating automation is a must with the sprawling IT ecosystems of today. It will soon be impossible to have a bird's eye view of your entire ecosystem (if it isn't already). One practical use of automation comes in the form of observability which provides a way to proactively spot trends and issues enabling proactive management rather than reactive which is what we have with non-AI-enabled monitoring tools.
VP of AI and Data Content Strategy, O'Reilly Media
DevOps teams are drowning in observability and monitoring data. DevOps teams can leverage AIOps technologies to analyze workload and environment behaviors to alert when anomalies are detected. This frees up time for DevOps teams to focus on delivering services to the business.
Head Geek, SolarWinds
AIOps enables developers and DevOps engineers to be superhumans. It allows them to make sense of the massive amount of data that comes their way. It makes troubleshooting and pinpointing problems a lot more efficient.
VP of Engineering and Technology for Monitor & Platform, Sysdig
RESOLVE NEW FAILURE TYPES
AIOps is a tool that provides a force multiplier of the DevOps staff and reduces or removes mundane tasks, allowing the team to focus on complex or new failure types. New failure types should be processed by AIOps tools that can learn how the issue was resolved. This means the next time it can automate a response, or more quickly escalate to a person or team that can resolve it. This provides them with the best knowledge about what is happening now, allowing them to quickly focus on fixing the problem. The AIOps tool should be able to detect issues before humans can respond so that many events are never noticed by the end-user.
As developers and enterprise IT teams are put under continued strain to manage complex infrastructure, AIOps helps developers to focus their efforts on problems that cannot be solved with siloed tools. The onset of COVID-19 has triggered a rapid adoption of new services to accommodate the move to remote work. This acceleration has also seen an increase across DevOps team tasked with monitoring applications for warnings and malfunctions. By utilizing advanced technology like machine learning, AIOps can better evaluate and alert of application warnings, eventually being able to predict the impact of threats.
Product Marketing Manager, Micro Focus
As companies prioritize customer experience and focus on their digital transformations, DevOps practices have emerged to streamline processes, remove inefficiencies, and reduce risks in software deployments. By proactively integrating AIOps within the DevOps infrastructure and tools already in place, businesses can monitor the entire life cycle of their software and predict issues before they even occur. This automated approach can result in faster recovery times across the board, from software bugs to cloud performance issues.
Director, Developer Experience, Cisco
The single most important value add of AIOps for IT Operations teams is the self-learning capability. As networks become more complex and the applications they carry become more rich and varied, it is practically impossible for network administrators to set pre-defined boundaries for good or bad performance. This is where the fundamentals of AIOps capabilities such as baselining, dynamic thresholding, anomaly detection, paired with human-in-the-loop feedback enables self-learning behavior. Depending on the maturity of the ML/AI algorithms paired with accurate domain-specific interpretation of data, AIOps systems can truly become powerful self-learning systems that will only bubble up the most relevant events, thereby reducing noise and letting the IT Operations teams focus on real issues.
VP of Engineering, LiveAction
DevOps teams want to operate quickly and efficiently. AIOps helps with this by allowing teams to easily set up AI-assisted automations on top of their data flows and reduce manual toil.
Mohan Kompella, VP Product Marketing,
Adam Blau, Director of Product Marketing,
Anirban Chatterjee, Director of Product Marketing, BigPanda
AIOps enables teams to implement more automation, alleviating developers of repetitive and low-value tasks so they can focus on innovation.
CTO, Perforce Software
As AI enters the DevOps stack, it can perform the monotonous and repetitive tasks that keep employees from prioritizing knowledge work. Using patterns from deployment data, analytics, process insights, and more, enterprises can build scripts for hyperautomation. This allows the human team to focus on critical tasks, alarms, and potential failures, while empowering them to maintain the type of "high quality and high process compliance" culture that provides comfort to internal and external customers.
CEO, Newgen Software
AUTOMATION THROUGH CI/CD
AIOps enables the developers to automate the process of product delivery through CI/CD, release management, production availability and resiliency.
VP Product Development and Cloud Operations, OpsRamp
With DevOps, organizations can achieve business agility and expedite go-to-market timelines by providing valuable updates to critical applications. However, DevOps is primarily focused on automating application delivery with CI/CD. Much of the application performance monitoring, and root cause analysis continue to be a manually intensive process where DevOps still has to collect data from multiple sources such as logs, metrics, events etc. AIOps provide insights to DevOps by automating data collection and inferences that help decrease mean time to recovery (MTTR). Additionally, AIOps provide predictive analytics that could ease much of the capacity planning and cost saving driving faster and sound decision making.
DOING MORE WITH LESS RESOURCES
The insights AIOps gives to DevOps teams enables them to glean more clarity with less effort by automatically analyzing their environment to uncover noteworthy trends that may otherwise have been missed. This allows engineers to work across previously siloed lines and to conduct investigations more efficiently, even if they involve unfamiliar parts of the system, such as those managed by other teams. As a result, small teams can feel like they have the resources of a big team, and big teams can communicate with the efficiency of a small team. The advantage, then, is multiplying the impact of an organization's engineering resources, so they can do more with less.
SVP of Product, Datadog
The greatest benefit that AIOps delivers to DevOps teams is the scalability of the "You Build it You Own it" (YBYO) model. Even at the lowest maturity levels of AIOps through automated incident detection, alerting and response coupled with ChatOps and on-call management provides immediate relief from the burden of staffing a 24/7 watch/op-center. At higher maturity levels, automated self-healing, outage prediction/automated failover, and automated cost vs. performance optimization provide even greater reduction of the cognitive load of successful teams. This not only allows teams to focus more on the user experience, quality, and security of the solutions they develop and operate but creates the opportunity for better work life balance, increased collaboration and inter-sourcing between teams, and ultimately is the key to sustained success at scale.
VP of Software, SAIC
HELP SMALL TEAMS
DevOps teams or developers have a lot of jobs and one of these is keeping systems up and running. Vendor solutions that surface anomalies are helpful but often create too much noise. Vendors who can save engineering time when troubleshooting an issue are critical. If you can shortcut steps by either referencing useful information in context or suggest what a user might end up doing these approaches really help small teams who wear many hats. This lets engineers get back to work quickly, which should be all of our goals in the vendor world.