Monitoring is a capability that every organization running cloud applications needs. This is why there is a constant stream of new startups in the monitoring space, and at the same time, a steady number of acquisitions of these startups by larger enterprises. In this article, we look at three such instances where enterprise cloud giants IBM and Splunk made big moves by acquiring startups to bolster their DevOps monitoring offerings.
IBM acquires Instana
Instana is the biggest and most well-known of the three startups featured in this post. It operates in the application performance monitoring (APM) space and competes with the likes of New Relic and Splunk. IBM has acquired Instana to add APM capabilities natively to its IBM Cloud platform.
With its monitoring features spanning end-user monitoring, container monitoring, APM, CI/CD pipelines, serverless computing, and service dependencies, Instana is a very capable monitoring tool.
Instana touts that its product’s uniqueness is automation. What this means is that it uses AI to automatically discover and map components across the entire system. In a large system, this manual task can consume a lot of time and be prone to errors. By automating this crucial step, Instana takes much of the pain out of APM. Since this is the all-important first step, it ensures that the incoming data is clean and ready to monitor and that it will not mislead.
Distributed tracing is a new field of interest in the DevOps monitoring space. It goes deeper than metrics and logs to reveal the route and path of every network request. Instana includes distributed tracing with its AutoTraceTM feature.
Instana offers a beautiful dashboard with advanced filtering capabilities to slice-and-dice monitoring data with just a few clicks. They call this “Unbounded analytics,” and it’s a way of going off the beaten track to look for hidden insights in the monitoring data.
Another interesting feature of Instana is its automated root-cause analysis. With this feature, Instana brings all incidents, warnings, and red flags together on a single screen for a comprehensive view.
As part of IBM, Instana will likely make it to the OpenShift cloud platform. This is now IBM’s key area of focus as it looks to be a hybrid cloud management platform. Instana’s capabilities to monitor across the entire spectrum of cloud platforms is key to this acquisition. IBM is stopping at nothing in its quest to build the best hybrid cloud system, and Instana looks like a worthy addition to the suite.
Splunk acquires Plumbr
There are two agents — a browser agent and a Java agent. The browser agent attaches to browsers and monitors browser user experiences. The Java agent attaches to a backend JVM and sends performance metrics from infrastructure. Users can choose to use just one or both agents.
The monitoring data is sent to Plumbr servers, which can run either as a SaaS or on-premises on your own infrastructure. Starting as a Java-only offering, Plumbr has since expanded to support PHP, Python, and nodeJS.
Plumbr focuses on the end-user experience and can replicate what an end-user sees accurately. Their browser agent is what enables this feature called real user monitoring. The browser agent logs all user activities and reports them. This capability fits well with Splunk’s vision for digital experience management (DEM).
Plumbr has some unique features, for example API call tracing. Typically, in a microservices application, requests touch multiple nodes across services built with varied programming languages. It can be hard to keep track of these requests, especially when errors occur. Plumbr brings visibility into requests by tracing them end-to-end across the entire path they take. This information is invaluable when troubleshooting and brings deep visibility.
Plumbr differentiates between root-cause analysis and root-cause detection. It says the former is not enough but it is what most monitoring tools do. Plumbr specializes in root-cause detection, and in doing so, helps resolve issues much faster.
Alerts can be sent directly as push notifications or to other monitoring tools like Slack or PagerDuty.
When resolving multiple issues, Plumbr can prioritize incidents, so you always focus on the most important ones. Plumbr calls this impact analysis. It is a way to rank issues based on their impact on end-users.
Splunk has a broad monitoring offering that includes everything from application performance monitoring (APM), to logging, to tracing, to alerting, and everything in-between. For Plumber, Splunk opens up a world of clients that have already bought into Splunk’s vision for monitoring and observability.
Splunk acquires Flowmill
Right on the heels of their Plumbr acquisition, Splunk announced another acquisition, this time of Flowmill. This startup is into network performance monitoring (NPM) for cloud applications.
There is a lot of activity in the cloud networking space with the advent of service meshes. Cloud-native apps are powered by containers, and monitoring them presents many challenges. There are open-source solutions like Prometheus and Jaeger that look to bring better observability. Flowmill’s offering in this space is to unify and simplify network monitoring in the cloud.
The secret sauce of Flowmill is eBPF, a Linux kernel feature that helps trace the networking activity of containers.
Flowmill integrates with the operating system, the cloud provider, and Kubernetes to pull performance data. A Flowmill agent is installed on every host VM or server. Once done, the agent tracks all containers and applications on the host. It does this without requiring any further changes to application code or the containers running within each host. Flowmill has two additional agents that are specific to Kubernetes and AWS. With these agents, Flowmill gathers and consolidates network performance data across the entire system.
Some of the important metrics Flowmill collects include data on throughput, packet loss, connection failures, DNS timeouts, and roundtrip time. It also reports container-specific metrics like the Kubernetes namespace, version number, and L4 ports.
Flowmill brings greater visibility into services by spotting hidden dependencies, uncovering traffic spikes and their causes, and highlighting unwanted service to service connections. It delivers these metrics in easy-to-read dashboards complete with visualizations. Flowmill delivers an interesting “Cost Dashboard” that shows the actual cost of network traffic within the application and between the app and external services. You can then optimize the most expensive services to save on costs. With the ambiguity in cloud spending, this feature alone can pay for the investment in Flowmill.
With this acquisition, Splunk adds to its network and DevOps monitoring chops and continues to offer and broader and deeper observability and monitoring suite.
DevOps monitoring: No longer an afterthought
Whether it’s a cloud vendor platform like IBM or a monitoring-only vendor like Splunk, it’s clear that monitoring is an essential part of cloud-native computing. As an SMB or enterprise customer looking to scale your cloud operations, this presents an interesting choice. You can go with an all-in-one platform like IBM/Red Hat that offers more than just monitoring or to opt for a specialized platform like Splunk. At every level of the stack, that can be the question — whether to consolidate or distribute your tools and vendors. Whichever route you choose, know that there are powerful options readily available. DevOps monitoring need not be difficult or an afterthought with such powerful and broad-based monitoring solutions available.
Featured image: Pixabay