An essential requirement for managing IT operations (ITOps) in today’s complex digital enterprises is the availability of high-quality monitoring data. Systematically deploying a monitoring framework to capture behavioral characteristics of all compute, communication, and storage components is vital to optimizing ITOps efficiencies. Data observability augments conventional system monitoring by providing end-to-end visibility across an enterprise’s entire technology stack. This enables the complete internal state of ITOps across an enterprise to be proactively monitored through the collection, analysis, and correlation of system outputs such as logs, metrics, and traces. Deeper value and context can be added to observability data by integrating Artificial Intelligence Operations (AIOps), leveraging Artificial Intelligence (AI) and machine learning (ML) to apply reasoning and problem-solving, remove noise, identify potential issues, and prescribe best actions for ITOps.
Combining observability and AIOps, Development Operations (DevOps) teams can proactively optimize systems and software so they are able to monitor, run queries, test products, see results, and debug systems ‘on the fly.’ Having more accurate, granular, and targeted data and the means to derive greater meaning and actionable intelligence from it improves process confidence levels, drives more informed decision-making, and empowers teams to deliver more reliable products faster. Additionally, when new products are introduced, performance can be monitored, and any anomalies that have the potential to impact operations detected, automatically flagged, and resolved, or escalated for human resolution where necessary, decreasing potential downtime and improving Mean Time to Resolution (MTTR).
Getting the Balance Right
While this all sounds great on paper, it’s important to remember that in any ITOps system, resources are limited. Given the amount and complexity of data flowing through today’s digitized enterprise system, it’s imperative that administrators have a solid understanding of monitoring requirements. A structured, clearly defined approach is a non-negotiable, as the adoption of an ad-hoc, manual, and intuition-based approach can lead to inconsistent and inadequate data collection and retention policies, which defeats the entire point of the program.
When conducting observability, determining the appropriate monitoring level for observing the IT estate and collecting data is key. An approach to monitoring that is either too aggressive or too conservative can lead to either very large or very small volumes of monitoring data. Both approaches suffer from drawbacks that make them impractical for truly effective use. Very large amounts of data generated by very fine monitoring parameters can be difficult to store, maintain, and analyze. For instance, logging ten metrics at a rate of one sample per second will consume about 720,000 KB per hour, one sample every five seconds about 144,000 KB per hour, and one sample every ten minutes 1,200 KB per hour. The collection of large volumes of data can become unmanageable and carries the risk of missing genuine anomalies and valuable insights in the noise. On the other hand, a very small amount of monitoring data generated by coarse operational parameters carries the risk of missing events of interest, incomplete diagnosis, and insufficient insights.
Adaptive Observability for DevOps
The emergence of ‘adaptive observability’ provides an optimal middle path, making it possible, based on intelligent deep data analytics, to increase or decrease monitoring levels in response to the system health of specific IT operations. Adaptive observability seeks to replace the ad-hoc, manual, intuition-based approach with a more systematic, automated, and analytics-based approach for system monitoring.
The core of adaptive observability is to intelligently assess system performance and infer the health and criticality of various system components. This inferred health and criticality is then used to generate dynamic monitoring guidelines for these components. So, for example, a poorly performing component will require monitoring at a finer level, with many metrics collected at a high sampling rate. More resources are automatically allocated to gather more data more frequently to diagnose and suggest resolutions to the issue quickly and efficiently. Once the issue is resolved, when things are healthy, it is no longer necessary to collect as much data, and at such a high frequency, resources can be deployed elsewhere in ITOps. On the flip side, for a healthy component, a very low sampling rate may be adequate.
Adaptive observability integrates two relatively new and active areas of research – adaptive monitoring and adaptive probing, which have, historically, been two important approaches for the measurement, monitoring, and management of complex systems. However, traditionally these two approaches have been used in isolation. Passive monitoring techniques compute at-a-point metrics and can provide fine-grained metrics but are agnostic to the end-to-end system performance, whole probing-based techniques compute end-to-end metrics but lack in-depth view of a component. Adaptive observability combines these two techniques to actively complement each other to produce a highly effective monitoring solution.
In closing, while there’s no doubt that observability delivers a number of tangible improvements for DevOps teams, it’s important to understand and set appropriate monitoring parameters, factoring in adaptive observability to optimize ITOps resources, streamline decision-making and problem-solving, and drive cost efficiencies across enterprises.
