PLEASE NOTE THIS IS AN ARCHIVED POST - Netuitive has since become Metricly, and the tool has matured greatly since the time this was written!
The number of tools targeting the DevOps community is expanding quickly, as evidenced by The Period Table of DevOps Tools. One would think this is good news, right?
The fact is DevOps shops have one primary objective: To advance their own applications and the business that depends upon them. Reaching this goal requires an integrated stack of capabilities that are generic as well as customer-specific.
The good news is that the open source community is aggressively addressing many of the required generic capabilities. Here is a chart to highlight some examples:
|Feature Capability||Requirements and Use Cases||DevOps Tool Provider|
|Instrumentation||Agents to monitor systems infrastructure, middleware, and applications||Collectd|
|Data Repositories||Services that leverage a wide range of technologies for storing, organizing, querying, and managing time-series data||OpenTSDB|
|Configuration Management||Facilitate and automate configuration changes||Chef|
|Graphing and UI Frameworks||Dashboards and widgets that specialize in visualization of data||Graphite|
|Analytics||Algorithms that transform raw data into results that discover anomalies, assign categories, or make predictions||R|
|Log Management||Index, transform, and search logs||ElasticSearch|
|Capacity Planning & Billing Analysis||Specialized reporting for measuring capacity and analyzing cloud costs||Netflix/ice for AWS|
While each of these providers is advancing on their own technology frameworks, timeframes, and licensing terms, integration of these tools into a DevOps stack is left to the end-user. To further the workload, these components have to play nice together before a DevOps shop can begin to layer in their unique context to yield customized and directly beneficial results. Consequently, DevOps engineers are required to spend significant time and resources on integrating and maintaining their tools as opposed to advancing their own business objectives.
This is déjà vu all over again. Think back to the early days of enterprise systems management… to the 1990s and early 2000s. Management tool frameworks were created based on corporate acquisitions by large framework vendors such as CA, BMC, HP, Compuware, and IBM. These tools were expensive to license and even more expensive (and frustrating!) to install and customize.
Today, it’s a new world. Over the last decade, open source projects such those mentioned above have reached a level of breadth, scale, and maturity that render them more than capable to replace their predecessor proprietary products. New monitoring vendors have the unique opportunity to leverage three major forces:
- Lessons learned from the old enterprise management frameworks
- Integration of open source technologies to which significant value can be added
- The DevOps mindset which provides an operational foundation for collaboration across all stakeholders
For DevOps users, this maps to incredible benefits such as spending less time investing in the development of agents, analytics, presentations, and core platform technologies. This allows resources to be dedicated to expanding the features and functionalities of your own business instead of operational upkeep. DevOps teams are in a position to demand extended functionality from their tool providers. Simply combining several, readily available, open source technologies is not enough to justify charging for a DevOps monitoring service.
The new breed of DevOps tool vendors must deliver not only a sound, scalable, and resilient architecture that leverages core, modern, open source technologies; they must offer value-added capabilities.
In terms of DevOps monitoring tools, you should expect:
- Extensible integration of open source agents to form sensor networks that insure adequate data in real-time
- Flexible visualization and reporting managed by a modern, user-friendly GUI
- Advanced analytics with customizable policies that encapsulate best practices for anomaly detection, cost identification and containment, and actionable notification
- Comprehensive tools to enrich and features to meet unique business requirements
DevOps teams’ catalytic powers extend well beyond the data center and into engineering pits and software development labs around the globe. It is where exciting and innovative features and functions integrate with open source technologies to create a new breed of tools.
At the end of the day, you should expect nothing less from DevOps monitoring tools than to provide capabilities that are greater than the sum of their open source parts.
Metricly coaches users throughout their cloud journey to organize, plan, analyze, and optimize their public cloud resources.Try Metricly Free