A Guide to OpenTelemetry - Part 5: The Challenges
October 24, 2022

Pete Goldin
APMdigest

Share this

While OpenTelemetry offers many advantages, the experts point out several challenges as well.

Start with: A Guide to OpenTelemetry - Part 1

Start with: A Guide to OpenTelemetry - Part 2: When Will OTel Be Ready?

Start with: A Guide to OpenTelemetry - Part 3: The Advantages

Start with: A Guide to OpenTelemetry - Part 4: The Results

The Project is Not Mature

Maybe the greatest challenge for OpenTelemetry is that the project is not mature. While the tracing component is fairly well advanced, the metrics and logging parts are still being formed.

"Currently, the project is not mature enough to support every stack, language, and signal," says Michael Haberman, CTO and Co-Founder of Aspecto. "While we believe it'll get there, the road to full stability is long."

"OpenTelemetry remains a young project in many ways, and many components are still in alpha or beta," explains Austin Parker, Head of Developer Relations at Lightstep by ServiceNow. "There is still work being done to bring in new signals such as profiling, logging, or real user monitoring (RUM), and breaking changes in those signals can be frequent."

"Logs and metrics are slowly catching up, but the API is still unstable which can be an issue," adds Vladimir Mihailenco, Co-Founder of Uptrace. "Logs and metrics are not stable yet so using them is more bumpy and requires some involvement with OpenTelemetry development and reading various changelogs."

Varying Quality

"While the existing OpenTelemetry libraries are already viable for manual and automatic instrumentation, which is a key part of any observability solution, it varies in quality," says Daniel Khan, Director of Product Management (Telemetry) at Sentry. "With new versions of libraries being released almost daily, it is up to the Open Source community to reverse engineer and adapt the instrumentation for every new version. This is not sustainable. Now is the time when library and framework maintainers have to start adding OpenTelemetry to their code. If this doesn't happen, the production use-case for OpenTelemetry will stay rather limited."

"The maturity of documentation, specification, libraries, and collector varies. One's experience might be very different depending on what they want to achieve," Marcin "Perk" Stożek, Software Engineering Manager of Open Source Collection, Sumo Logic, adds.

Bugs

"OpenTelemetry is developing at a fast rate, and the instrumentation is rapidly changing. This can lead to frustrating bugs at times. But the community is actively taking steps to address concerns around instrumentation stability," says Pranay Prateek, Co-Founder of SigNoz.

Does Not Provide Backend Storage, Analysis or Visualization

Another important challenge to be aware of: OpenTelemetry does not provide any backend storage, analysis or visualization, so to gain full value of the project you need to implement these components on your own or with the help of a service provider.

"Without the proper tools and integrations, it can be challenging to make sense of what the data OpenTelemetry uncovers, and what was meant to provide visibility into IT performance and availability could actually end up creating more data noise instead," says Joe Byrne, VP of Technology Strategy and Executive CTO at Cisco AppDynamics.

Requires Large-Scale Initiative

To truly adopt OpenTelemetry requires a large-scale effort by an organization.

"Users will have to replace some of their existing toolchains of telemetry collection (especially for logs and metrics)," Haberman of Aspecto points out. "It will require quite a lot of effort, and usually, companies are not excited to make this large-scale shift."

"Migrating from current proprietary collection technologies to OpenTelemtry is non trivial for any large customer," adds Nitin Navare, CTO of LogicMonitor.

Difficult to Manage at Scale

"As OpenTelemetry evolves, it will become more complex and challenging to configure and manage at scale," warns Jonah Kowall, CTO of Logz.io.

Alois Reitbauer, Chief Product Officer at Dynatrace, agrees: "The challenge will be managing large-scale OpenTelemetry rollouts and monitoring their health."

Hard to Learn

"OpenTelemetry can be challenging for new developers to learn, as documentation gaps still exist due to the rapid pace of development," says Parker of Lightstep.

"Implementing OpenTelemetry across every part of the system requires deep knowledge and has a high entry point effort," adds Haberman of Aspecto. "This forces users to fully understand how OTel works and get involved in the project's updates. Though, as the project matures and the amount and quality of resources grows, adoption will get easier."

Martin Thwaites, Developer Advocate at Honeycomb, explains further: "At first glance, OpenTelemetry can be challenging to get started with, especially for certain languages that don't provide documentation for orchestration. However, even with this, early adopters were more than willing to dig deep and make it work. Therefore as focus on documentation becomes a priority, this barrier will quickly be eliminated. And, as adoption grows into the early majority and beyond, this will be more important to new users looking to come on board."

"Furthermore, looking deeper at the various language and framework SDKs are doing, it becomes harder to understand," he continues. "Providing more 'easy mode' integrations like the Agents and Kubernetes Operators will be essential to broader adoption. This will ease the issue of sampling or managing high data volume."

Developer Priorities

"It's important to consider that observability is not the fundamental goal when developing new software," explains Sajai Krishnan, General Manager, Observability, Elastic. "A software developer's primary goal may be to make the application or library they are building meet key business requirements or reduce the risk of impacting the performance of their code. Implementing observability may not be a primary goal, as developers carry on with familiar logging as has been done for decades, which could hurt the broad adoption of OpenTelemetry."

Download the 2022 Gartner Magic Quadrant for APM and Observability

Lack of Commercial Support

As with any open source solution, tech support and upgrades could be an issue for users of OpenTelemetry because it is not backed by a commercial vendor.

Vendor Enhancements

"There is also the risk that the standardization and vendor neutrality benefits of OpenTelemetry are lost by vendor enhancements beyond the standard features in their downstream distribution," says Krishnan at Elastic.

Go to: A Guide to OpenTelemetry — Part 6: OTel and APM

Pete Goldin is Editor and Publisher of APMdigest
Share this

The Latest

March 27, 2024

Nearly all (99%) globa IT decision makers, regardless of region or industry, recognize generative AI's (GenAI) transformative potential to influence change within their organizations, according to The Elastic Generative AI Report ...

March 27, 2024

Agent-based approaches to real user monitoring (RUM) simply do not work. If you are pitched to install an "agent" in your mobile or web environments, you should run for the hills ...

March 26, 2024

The world is now all about end-users. This paradigm of focusing on the end-user was simply not true a few years ago, as backend metrics generally revolved around uptime, SLAs, latency, and the like. DevOps teams always pitched and presented the metrics they thought were the most correlated to the end-user experience. But let's be blunt: Unless there was an egregious fire, the correlated metrics were super loose or entirely false ...

March 25, 2024

This year, New Relic published the State of Observability for Financial Services and Insurance Report to share insights derived from the 2023 Observability Forecast on the adoption and business value of observability across the financial services industry (FSI) and insurance sectors. Here are seven key takeaways from the report ...

March 22, 2024

In MEAN TIME TO INSIGHT Episode 4 - Part 2, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at Enterprise Management Associates (EMA) discusses artificial intelligence and AIOps ...

March 21, 2024

In the course of EMA research over the last twelve years, the message for IT organizations looking to pursue a forward path in AIOps adoption is overall a strongly positive one. The benefits achieved are growing in diversity and value ...

March 20, 2024

Today, as enterprises transcend into a new era of work, surpassing the revolution, they must shift their focus and strategies to thrive in this environment. Here are five key areas that organizations should prioritize to strengthen their foundation and steer themselves through the ever-changing digital world ...

March 19, 2024

If there's one thing we should tame in today's data-driven marketing landscape, this would be data debt, a silent menace threatening to undermine all the trust you've put in the data-driven decisions that guide your strategies. This blog aims to explore the true costs of data debt in marketing operations, offering four actionable strategies to mitigate them through enhanced marketing observability ...

March 18, 2024

Gartner has highlighted the top trends that will impact technology providers in 2024: Generative AI (GenAI) is dominating the technical and product agenda of nearly every tech provider ...

March 15, 2024

In MEAN TIME TO INSIGHT Episode 4 - Part 1, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at Enterprise Management Associates (EMA) discusses artificial intelligence and network management ...