The technology industry now recognizes OpenTelemetry as the reference standard for its adoption. As companies in the industry move toward stacks of modern applications built in multi-cloud and hybrid environments, OpenTelemetry is essential for managing and optimizing availability and performance in fragmented and ever-changing cloud-native environments.
Faced with the constant pressure of continuous innovation, IT managers want to be able to take advantage of microservices and cloud services in order to accelerate production in order to meet the growing challenge of reducing time to market. They are aware of the flexibility and scalability offered by cloud-native technologies.
But this transition to modern applications is accompanied by complexity and a large amount of data. Schematically, if you have a single application on site, say a Java application, and you want to move that application to a cloud native architecture, you will have to split the application into 10, 12, or even 20 separate microservices files that can also be shared with other applications.
It is mainly for this reason that the OpenTelemetry standard is essential to businesses today. It’s the only way for IT teams to streamline deployment and take advantage of a neutral standard to develop and maintain their applications, tied directly to their CI/CD cycles.
When used alone, OpenTelemetry is not enough to improve application performance
However, while no one can deny that the cloud is the future of computing, the abandonment of on-premises applications and infrastructures will not happen overnight. Migrating to the cloud is very complex and, if poorly managed, can be very expensive. The current economic climate is also prompting companies to invest more in cloud computing, rather than embarking on wholesale migration.
Therefore, a transition period should occur, during which companies will move a portion of their applications and infrastructure to cloud-native environments, while still managing the majority of their stack in-house. IT teams must therefore develop a hybrid strategy in which they link OpenTelemetry to the general application mix through traditional agent-based monitoring systems. They need a monitoring platform that provides flexibility to cover both cloud-native and on-premises environments – with OpenTelemetry and agent-based entities built into the platform.
IT managers need to relate IT data to business indicators to give meaning to their business
IT teams therefore need a monitoring platform capable of extracting business (or business) transactions from OpenTelemetry data as well as other MELT data elements (metrics, events, logging and tracking), to the infrastructure, application and data stored by the cloud provider.
Many IT teams deploy separate tools for monitoring cloud-native and traditional applications. This means that they must use “split screen” mode and cannot see the full path of the transaction or the end user in the application stack as a whole. As a result, it becomes very difficult to solve problems, which means that the mean time to resolution (MTTR) and the mean time to occurrence (MTTX) are mechanically increased due to the inevitable manual processes.
This is why it is essential that IT teams can integrate OpenTelemetry directly into the unified monitoring platform, to have a clear view of the entire application journey, even when components are running. On-premises and cloud-native environments.
IT managers need a strong OpenTelemetry adoption strategy
Despite the excitement surrounding OpenTelemetry over the past couple of years, it must be admitted that it is still a relatively new technology. As an open source project that actually took off in 2019 after the merger of OpenTracing and OpenCensus, it has made rapid progress. Already, OpenTelemetry allows visualization of traces and metrics and will soon support records.
In fact, adoption of OpenTelemetry is still in its infancy and few companies have implemented it entirely in their native cloud environments, although the situation is changing rapidly. Indeed, it is difficult to find a company that would not mention OpenTelemetry for its future strategy. Most IT managers plan to use OpenTelemetry within the next two years. Therefore, it is important that they take a deliberate approach to OpenTelemetry to ensure that it can be implemented smoothly and sustainably, especially as technology is changing rapidly.
One of the major hurdles in implementing OpenTelemetry is that teams expect it to provide the same level of visibility that is achieved by monitoring through a proprietary proxy in on-premise environments. This is not the case. With OpenTelemetry, today’s IT teams only get traces and metrics, and a projection effort seems necessary at this point
When this benchmark is used alone, it is difficult to extract intrinsic value from it because it produces large amounts of data and makes it difficult to get an overview of the application. This is where IT teams need a broader monitoring platform, to leverage the data they get through OpenTelemetry and combine it with other data obtained through agents or from public cloud environments, such as those derived from Kubernetes.
OpenTelemetry cannot be adopted without learning. In many cases, the go-to for OpenTelemetry is made by developer teams who see it as a way to debug their code during the app development process. Development teams then required operations teams to adopt OpenTelemetry, but many of them lack a deep understanding of the technology, what it offers, and how it fits into their monitoring system more broadly.
For all IT managers, there is now an urgent need to ensure they have the knowledge and skills to adopt OpenTelemetry as part of a broader monitoring strategy. As organizations seek to accelerate their transition to a microservices-based architecture, OpenTelemetry will become increasingly important for IT teams to manage and improve availability and performance across multi-cloud and hybrid environments.
IT teams need to ensure they can integrate OpenTelemetry data into their existing monitoring platform and learn from it in a business context. They can ensure they have the unified vision and insights they need to deliver seamless, seamless digital experiences for users
Article by: Eric Salviak, Senior Business Value Consultant, Cisco AppDynamics