Dynatrace Extends Container Reach
There are a lot more types of containers being deployed in production environments than just Docker. In fact, one of the most widely employed types of containers employed in a production environment can be found in the platform-as-a-service (PaaS) environment overseen by the Cloud Foundry Foundation (CFF).
To make it easier to monitor those containers, Dynatrace has extended the reach of its application performance management (APM) software to take advantage of the open-source BOSH tools for deploying distributed services developed by CFF. Via that add-on BOSH module, Dynatrace now can monitor Garden-runC containers to enable the auto-discovery of applications, microservices and underlying infrastructure.
Mike Villiger, a product evangelist for Dynatrace, says larger enterprise IT organizations especially will need to be able to monitor multiple types of containers running on-premises as well as in various public clouds. Developers may be voting with their feet to deploy Docker containers in a container-as-a-service (CaaS) environment. But when it comes to deploying hundreds of cloud-native applications at scale, there’s no substitute for a PaaS that can scale, Villiger says. To make it simpler for IT organizations to move applications between those environments, CFF makes it possible to run Docker images on top of the Cloud Foundry PaaS.
Villiger says the rise of microservices is about to force APM issues on several fronts. There’s a tendency for developers to employ open-source APM tools when they build their applications. But as IT operations teams get confronted with hundreds—possibly even thousands—of microservices, they will need monitoring tools that not only can scale but also are commercially supported. That’s potentially critical for APM vendors, says Villiger, because less than 5 percent of all applications deployed today are being monitored.
In addition to that classic build-versus-buy debate, Villiger notes there will be a much bigger need to infuse machine-learning algorithms into monitoring platforms. The need for those capabilities will make it more difficult for an open-source project to keep pace with what can be achieved by a commercial vendor. Dynatrace, for example, can take advantage of the massive amounts of data it collects to drive artificial intelligence (AI) into DevOps processes. The result should be a massive reduction in both the amount of time it takes to discover and issue and the mean time to resolution (MTTR), says Villiger. In fact, Villiger notes that Dynatrace has already demonstrated a voice-enabled AI capability, dubbed Davis, that combines natural-language processing with machine- and deep-learning algorithms to both proactively identify anomalies and answer questions any IT operations team might have.
Naturally, the implications of that capability might give some IT operations teams cause for pause. At the same time, however, it should be apparent to all that managing microservices based on containers at scale will be beyond the capabilities of mere IT mortals. Without some additional AI capabilities to augment the IT staff, the management of microservices will eventually collapse of its own weight and complexity.