“We want to take back control.” It's a phrase we hear a lot from our customers. Logical, because apps and functionalities are delivered from increasingly complex landscapes. And an increasing number of those landscapes are in the cloud, where you have less control as an administrator. This often makes finding issues a frustrating, time-consuming scavenger hunt.
What they soon find out is that observability on a cloud-based data platform like Elastic is not just a technical implementation. It is also an organizational change. And the challenge already lies in formulating the question: taking back control. Whose?
Education and resistance
In the classic situation, each team is responsible for its own part of the chain. Vendor A provides the payment API, vendor B provides virtual machines, team X does the mobile app and Team Y does the database, etc. These teams all have their own management methodology and collect their own data.
The only way to diagnose performance issues in such an organization is to call all suppliers and teams as soon as something is wrong. Practice shows that some people prefer not to let their mouth shut and prefer to point to others. Or simply don't have the insights, because their own observability is not up to par.
With central monitoring, this whole dynamic disappears, because you, as an administrator, can always see where the problem is and also show others what is actually happening. You also see bottlenecks coming earlier, so identify patterns and therefore know what (or who) is usually the cause. And to make it even more transparent, Blis developed the topology dashboard. This (management) dashboard shows the performance and health status of your entire environment in one helicopter view.”
That's great, but at first it can be quite confronting for suppliers and colleagues. Because if you take back the grip, they feel like they just lose their grip. They can't hide anymore and have to get to work when the numbers indicate that their part of the chain is the bottleneck. No discussion. The “educational effect” that comes from central insights is therefore not always immediately met with cheers. But in our experience, resistance quickly disappears when the benefits become clear.
The benefits of observability
In practice, observability with Elastic means that you have a central, objective view of the health and performance of each part of your application chain. The benefits are enormous. Here are four of the most important ones:
- Clarity about the quality of your service. You know, even without customer complaints, whether your service meets the requirements of your end users and the SLAs you have concluded with your customers. This clarity means the end of many discussions and gives you the opportunity to manage and optimize more quickly and effectively.
- Intervene more quickly if things go wrong. Users only call the helpdesk when things are really wrong. Good observability often means that you identify performance problems long before that. So you can intervene before your users notice the problem.
- Proactive management. As you gather more data and learn more about your IT landscape, you can better see problems coming. This allows you to plan management and maintenance in such a way that performance remains high, while also optimally balancing costs.
- More effective use of your people. “Extinguishing fires” is stressful, tiring and inefficient. Especially when you have to do it based on incomplete information. The more insight you have into your performance and health, the better your administrators know where to look for the cause of any issues. And the more time, rest and attention they have left to work on structural improvements.
- Control over spend and energy consumption. Controlling costs and meeting sustainability requirements and goals are often an important part of the business case for observability. Because only with a central overview can you optimally balance costs, emissions and performance.
Attention to the process
Points 2 and 3 on that list still deserve some explanation. Because having the right data is one thing; the process associated with the data is ultimately even more important. Observability ensures that you identify or even prevent performance problems earlier, provided that you can also respond to them effectively. In practice, this means that you need new work processes and that responsibilities and activities are changing. This is likely to change the job content and responsibilities of your administrators. For example, they should start responding to alerts from the system instead of following their own instincts. Or their work is shifting from incident management to preventive maintenance, which requires a significant change in mentality for some people.
Changing culture: from finger-pointing to improving
Ultimately, every IT administrator wants the same thing: stability. If you are responsible for the performance of a system, you don't want to be overwhelmed by unexpected changes or fluctuations in performance. You want to know what to expect every day, so you know that you can deliver on your promises to customers and users and that you can quickly resolve the issues that do arise. And when your shift is over, you want to sleep peacefully knowing that everything is running as it should. No angry users, no discussions, no unplanned changes, and no stress.
Observability helps with that. That is also the reason that, after a period of familiarization, everyone without exception is positive about observability as we organize it at Elastic. And where the major commercial and legal interests often gave reason to point to others, to present the facts a little better or “hide until the problems blow over”, everyone involved actually comes closer when the facts are clear to everyone.
The culture is changing, both within your organization and in your cooperation with suppliers and partners. The emphasis is shifting from “putting out fires” to thinking together about structural improvements and optimizations. And that is the real benefit of observability.
Want to know more?
Want to know more about Observability? Or do you have a specific Performance challenge? Please contact: Cees Schrijen, Account Manager Enterprise IT (c.schrijen@blisdigital.com | 06-17503107).
Learn more about our Performance Monitoring and Analytics services
.jpg)
