Drilling Into Your Data

The Drill Down feature enhances your ability to investigate, analyze, and resolve issues without needing to switch contexts. Instead of sifting through large volumes of data indiscriminately, focus your efforts on drilling into relevant and focused subsets of data in order to save your team time and effort.

Clicking on any of Middleware's interactive tables will produce popouts that connect you to all information related to that data point that Middleware has collected. These popouts will allow you to drill down to the root cause regardless of which data point you start from.

Metrics

Metrics are compact statistical representations of system state and resource utilization. They are quantifiable data points within your software system, like throughput, resource utilization, and error rates.

Metrics provide a high-level view of the system in question, whereas traces and logs provide you with a more granular look into your system-level data. Access metrics by drilling into the Logs page:

Events

Events are discrete occurrences that transpire at a particular moment in time. They can be as broad as an uncaught exception, or specific like a user logout or button click.

Access events by drilling into the APM page from the Span List:

Spans

Spans are structured log messages that represent a single operation in a trace. They allow you to break down your traces into discrete operations.

Multiple connected spans form a trace, with a single root span that describes the entire operation. Spans are useful for identifying system performance bottlenecks and dependencies across your infrastructure in a concise, structured, and human-readable format. Access spans by drilling into the APM Page:

Traces

Traces are the entire path of a transaction or request as they flow through various components of your distributed system. Traces are particularly useful for identifying the root cause of a seemingly isolated issue in a broader context, such as latency cascading and dependency failures.

Distributed Traces look at entire operations and how they form transactions. Distributed Traces are composed of spans, with a start and end time and location where they occur. Access traces by drilling into the APM or Logs Page:

Trace Attributes

Trace Attributes are key-value pairs that contain metadata associated with the span you are investigating.

Error Traces

Error traces are records of the sequence of function calls and their associated parameters and variables that occur when an error or exception occurs in your software application. They are useful for correlating other types of data collected by your observability product, such as performance metrics, log messages, or system events.

Logs

Logs are timestamped records of text that contain information about events, activities, and messages generated by your distributed software architecture. Logs contain metadata that can be structured or unstructured.

Logs are at the core of most drill-down practices and are useful when drilling into resolving event-based issues, post-mortem analysis related to outages, and monitoring complex systems patterns and potentially abnormal behavior. Access logs by drilling into the Logs Page:

Next Steps

• Log Monitoring Overview
• Infrastructure Monitoring
• Intro to APM
• Real User Monitoring (RUM)
• Continuous Profiling