Humanizing Observability and Controllability

Humanism is a philosophical stance that is at the heart of much of what OpenSignals aims to bring to the table for service level management operations. It runs counter to the misguided trend of wanton and wasteful big data collection so heavily touted by those focused more on selling a service rather than solving a problem, now and in the future. At the core of humanism are progress and the value of human agency, whereas current observability techniques and technologies are anything but this. While cloud infrastructure has scaled to meet ever-increasing levels of change and complexity, observability has stayed with cave wall painting – logging and tracing.

To bring a human back to the table where s/he can add value and act intelligently requires a scaling up of the service level management model with an emphasis on simplicity and significance while offering new higher-order forms of sensing, selection, and sapience. Over the last decade, with the arrival of the cloud, there has been an incredible scaling of computing resource capacities and communication networks but not so much in the ability of humans to understand and reason about the state of system or service, or the significance of a situation. When it comes to scaling human cognition, less is more. Scaling agency and attention to this new world of micronization and interconnectedness requires a shrinking of the surface area and a shift to sensibility and selectivity.

While there might never be a one size fits all kind of model for observability, and in turn controllability, a suitable model and representation of reality should be able to frame a system (space) or situation (temporal) at various levels of scaling (abstraction and compression) without too much in the way of communication overhead or cognitive effort. This is where OpenSignals shines in comparison to other observability technologies such as distributed tracing, metrics, or event logging. While the OpenSignals model is simple and small, it is astonishingly powerful in compressing complexity, synthesizing semantics, amplifying analysis, and detecting dynamics – the same intellectual qualities we look to each other in leading the progress of society and guiding decision making.

Much of the conceptualism underlying OpenSignals is based on the early foundations of human (and animal) communication and control – behavioral signaling and state inference. OpenSignals brings a much needed human approach to how systems of services communicate, converse, and collectively coordinate cooperation. Signals are the ultimate in the clear and concise communication of operation or outcome. Signals and the traces of such, signs left within an environment, inform others of what is of significance. They augment what is communicated over different channels, mediums, or within request or response payloads. Signals are used to influence the behavior of clients. There is no need for inspecting or interpreting a message – ambiguity is all but absent in signaling.

Signals steer behavior in others who sense such signals and signs – happening first with the mapping of a signal to a possible status value. In human society, the mapping is generally universal. If another swiftly raises a clenched fist, most others within an environment will make a similar assessment of what is being signified and over time with repeated signaling accurately and similarily judge the state – another is angry. Optimally there is little need for contextualization of the signal – no additional data capture is required. Contextualization is conversational. In OpenSignals the set of signals has been carefully chosen based on the most important constructs, controls, and codes found in various existing service-to-service communication channels, contracts, and clients.

In human society, the inferred state of another is always subjective. The subjectivity encompasses the sensitivity of the receiver of the signal, the sequencing of signals received, as well as the scoring algorithm employed in assessing the state of another; over time, and within the scope of a contextual space or changing situation. From the small and simple model of OpenSignals comes complexity, but of a kind that is extremely vigorous and versatile in capturing the diverse degrees of resilience and fault tolerance mechanisms that invariably exist within an enterprise-scale system of services. OpenSignals can be used with relatively little effort to extract actual state dynamics and, in turn, predict future state trajectories at both service and system levels. This is the future way!