Intel PowerStat and New Relic Integration

Input and output integration overview

Monitor power statistics on Intel-based platforms and is compatible with Linux-based operating systems. It helps in understanding and managing power efficiency and CPU performance.

This plugin allows the sending of metrics to New Relic Insights using the Metrics API, enabling effective monitoring and analysis of application performance.

Integration details

Intel PowerStat

The Intel PowerStat plugin is designed to monitor power statistics specifically on Intel-based platforms running a Linux operating system. It offers visibility into critical metrics such as CPU temperature, utilization, and power consumption, making it essential for power saving initiatives and workload migration strategies. By leveraging telemetry frameworks, this plugin enables users to gain insights into platform-level metrics that help with monitoring and analytics systems in the context of Management and Orchestration (MANO). It facilitates the ability to make informed decisions and perform corrective actions based on the state of the platform, ultimately contributing to better system efficiency and reliability.

New Relic

This plugin writes metrics to New Relic Insights utilizing the Metrics API, which provides a robust mechanism for sending time series data to the New Relic platform. Users must first obtain an Insights API Key to authenticate and authorize their data submissions. The plugin is designed to facilitate easy integration with New Relic’s monitoring and analytics capabilities, supporting a variety of metric types and allowing for efficient data handling. Core features include the ability to add prefixes to metrics for better identification, customizable timeouts for API requests, and support for proxy settings to enhance connectivity. It is essential for users to configure these options according to their requirements, enabling seamless data flow into New Relic for comprehensive real-time analytics and insights.

Configuration

Intel PowerStat

[[inputs.intel_powerstat]]
  # package_metrics = ["current_power_consumption", "current_dram_power_consumption", "thermal_design_power"]
  # cpu_metrics = []
  # included_cpus = []
  # excluded_cpus = []
  # event_definitions = ""
  # msr_read_timeout = "0ms"

New Relic

[[outputs.newrelic]]
  ## The 'insights_key' parameter requires a NR license key.
  ## New Relic recommends you create one
  ## with a convenient name such as TELEGRAF_INSERT_KEY.
  ## reference: https://docs.newrelic.com/docs/apis/intro-apis/new-relic-api-keys/#ingest-license-key
  # insights_key = "New Relic License Key Here"

  ## Prefix to add to add to metric name for easy identification.
  ## This is very useful if your metric names are ambiguous.
  # metric_prefix = ""

  ## Timeout for writes to the New Relic API.
  # timeout = "15s"

  ## HTTP Proxy override. If unset use values from the standard
  ## proxy environment variables to determine proxy, if any.
  # http_proxy = "http://corporate.proxy:3128"

  ## Metric URL override to enable geographic location endpoints.
  # If not set use values from the standard
  # metric_url = "https://metric-api.newrelic.com/metric/v1"

Input and output integration examples

Intel PowerStat

1. Optimizing Data Center Energy Usage: Monitor power consumption metrics across all CPUs in a data center. By capturing real-time data, administrators can identify which servers consume the most power and implement shutdowns or load balancing strategies during low demand periods, effectively reducing operational costs.

2. Dynamic Workload Migration Based on Power Efficiency: Integrate this plugin with a cloud orchestration tool to enable dynamic migration of workloads based on power usage metrics. If a particular server is recorded as consuming excessive power without corresponding output, the orchestrator can seamlessly migrate workloads to more efficient nodes, ensuring optimal resource utilization and lower energy expenses.

3. Monitoring and Alerting Mechanism for Overheating CPUs: Implement an alerting system using the CPU temperature metrics captured by Intel PowerStat. Setting thresholds for temperature can alert system administrators when a CPU is prone to overheating, allowing proactive measures to be taken before hardware damage occurs, ultimately extending the life of the components.

4. Performance Benchmarking for CPU-intensive Applications: Use the metrics provided to benchmark the performance of CPU-intensive applications. By analyzing the cpu_frequency, cpu_temperature, and power metrics under load, developers can optimize application performance and make informed decisions regarding scaling and resource allocation.

New Relic

1. Application Performance Monitoring: Use the New Relic Telegraf plugin to send application performance metrics from a web service to New Relic Insights. By integrating this plugin, developers can collect data such as response times, error rates, and throughput, enabling teams to monitor application health in real-time and resolve issues quickly before they impact users. This setup promotes proactive management of application performance and user experience.

2. Infrastructure Metrics Aggregation: Leverage this plugin to aggregate and send system-level metrics (CPU usage, memory consumption, etc.) from various servers to New Relic. This helps system administrators maintain an comprehensive view of infrastructure performance, facilitating capacity planning and identifying potential bottlenecks. By centralizing metrics in New Relic, teams can visualize trends over time and make informed decisions regarding resource allocation.

3. Dynamic Metric Naming for Multi-tenant Applications: Implement dynamic prefixing with the metric_prefix option to differentiate between different tenants in a multi-tenant application. By configuring the plugin to include a unique identifier per tenant in the metric names, teams can analyze usage patterns and performance metrics per tenant. This provides valuable insights into tenant behavior, supporting tailored optimizations and enhancing service quality across different customer segments.

4. Real-time Anomaly Detection: Combine the New Relic plugin with alerting mechanisms to trigger notifications based on unusual metric patterns. By sending metrics such as request counts and response times, teams can set thresholds in New Relic that, when breached, will automatically alert responsible parties. This user-driven approach supports immediate responses to potential issues before they escalate into larger incidents.

Feedback

Thank you for being part of our community! If you have any general feedback or found any bugs on these pages, we welcome and encourage your input. Please submit your feedback in the InfluxDB community Slack.