Platform metrics is a key feature of the Dev Console that enhances observability and simplifies debugging for platforms such as Neo, Extensions, and Vulcan. It serves as a central hub for monitoring performance, identifying issues, and streamlining troubleshooting. Furthermore, it enables you to analyze API behaviors and performance trends.

The Platform Metrics consists of a dashboard which caters to different applications or extensions. The Platform Metrics dashboard gives a high-level overview of all metrics.

Accessing Platform Metrics

To access Platform Metrics,

Navigate to the Logs and Metrics > Platform Metrics from the left pane of the Dev Console.

The following image shows the Platform Metrics page.

Working with Platform Metrics

In the Platform Metrics, you can choose to see the performance based on time range, product, or dashboard.

To view metrics,

1. From the Platform Metrics page, select the date range from the drop-down and click Apply.
   Note: The maximum custom date range you can select is 30 days.

   

2. Select the product from the Select Product drop-down.

   

3. Select the dashboard from the Select DasbBoard drop-down.

4. (Optional)Filter your search by selecting specific API endpoints for the chosen products.

The metrics are displayed based on the selected values.

Drill-Down Flow for Debugging

Each dashboard has different sections that display various metrics to help with debugging. Platform Metrics follows a four-step drill-down process. Below are the drill-down sections and their fields for the Neo Platforms product.

Performance Metrics

The Performance Metrics displays the overall metrics related to Neo Platforms. The fields in this section are:

Total Requests (ABS)

The total number of API calls the Neo platform processes within the selected timeframe. You can view Throughput in a graph.

Latency In ms P(95)

The time within which the Neo platform completes 95% of API calls, measuring worst-case performance. The Percentile appears as a graph.

Error Percentage

The percentage of failed API calls in the Neo platform relative to total requests.

Requests Per Request Uri

The Neo platform tracks API requests to each unique endpoint (Request URI) within the selected timeframe. This helps identify the most frequently accessed endpoints and analyze usage patterns. It includes the following fields:

Requests per status code

The number of API requests grouped by their HTTP status codes within the selected timeframe. It includes the following fields:

Capillary Core Platform

This is the second level of the drill-down. It displays the metrics of the Intouch or Capillary APIs invoked in the Neo Platform. The fields in this section are:

Total Requests (ABS)

The total number of Intouch API invoked within the selected timeframe. You can view Throughput in a graph.

Latency In ms P(95)

The time within which the Intouch APIs complete 95% of API calls, measuring worst-case performance. The Percentile appears as a graph.

Error Percentage

The percentage of failed Intouch API calls relative to total requests.

Parent Api Calls

The Parent API Calls refer to the primary API that triggered the Intouch APIs. It includes the following fields:

Requests Per Request Uri

Requests Per Request URI refers to the number of API requests made to each unique endpoint (Request URI) within a selected timeframe. Below are the fields.

Requests Per Status Code

Requests Per Status Code refers to the number of API requests grouped by their HTTP response status codes within a selected timeframe. Below are the fields.

Error Insights

This is the third level of the drill-down. It provides a detailed analysis of API errors within a selected timeframe. The fields in this section are:

Error Count By Status Code

Error Count By Status Code refers to the number of API errors grouped by their HTTP status codes within a selected timeframe.

Error Message Summary

Error Message Summary is a consolidated view of all error messages generated by API calls within a selected timeframe. Below are the fields.

Error Message By Org Id

Error Message By Org ID" refers to a report or metric that shows the error messages associated with a specific organization (Org ID). The following are the fields.

Error Message By Org Id And Uri

Error Message by Org ID and URI provides a breakdown of API errors grouped by organization ID (Org ID) and request URI (endpoint) within a selected timeframe.

Performance Insight

This is the fourth level of the drill-down. It includes analytics and metrics that assess the efficiency and responsiveness of an API.

Count (>200ms)

Count (>200ms) is the number of API requests that take more than 200 milliseconds to complete.

Api Details

API Details refers to specific information about an API request, method, status code, response time, request ID, and other relevant metadata. The following are the fields:

Creating a Custom Dashboard

Creating a custom dashboard is a simple process that requires minimal manual UI development. Follow these steps to set up a dashboard:

1. Identify the data requirements for your desired dashboard. Ensure the necessary metrics are available.
2. Define the dashboard layout and arrange the charts to present information.
3. Determine how many charts your dashboard needs and assign a unique ID to each one.
4. Define the configuration for each chart in a JavaScript array. Specify the following:
   • Chart ID – Assign a unique identifier to the chart.
   • Data Mapping – Link the chart ID to a query in your database. Use dynamic variables (e.g., order IDs, app names) to make the query adaptable.
   • Display Name – Set the header that appears on the dashboard.
   • Chart Type – Choose from scalar, table, line chart, or area chart. Each chart supports only one type.
   • (Optional) Metadata & Customization (Optional) – Configure additional properties:
     • Format timestamps.
     • Convert large numbers into readable formats.
     • Enable pagination for tables with large datasets.
5. (Optional) Enable filters to let users refine the displayed data. Configure the filter parameters and ensure a query fetches filter options dynamically.
6. Call the initializeDashboard function in your code and pass the chart and filter configurations.
7. The dashboard automatically generates and displays the charts based on your configurations. You don’t need to make manual XML or CSS changes.

Using Dashboard Filters

Dashboard filters let users refine the displayed data based on specific criteria. Filters improve data analysis by allowing users to focus on relevant subsets without modifying queries or configurations manually.

Key Features

• Optional Configuration - Filters are not mandatory for every dashboard. Dashboard creators decide whether to enable them based on the dashboard’s requirements.
• Configurable Availability - Filters appear only if configured. Dashboard designers specify whether to include filters and which data attributes they apply to.
• Dynamic Filter Options - Filter options update dynamically based on real-time data queries. This ensures you always have relevant filtering options instead of predefined, static lists.
• Filtering Purpose - Filters help you refine displayed data, making it easier to analyze specific metrics.
  For example, you can filter by API endpoint to focus on a specific service's performance.
• Filter Parameter Configuration - During dashboard creation, developers specify which columns or parameters you can filter.
• Query-Based Filtering - Each filter runs a database query to fetch available options. This keeps the filter selections accurate and up to date.
• User-Friendly Interaction - Once configured, filters appear as dropdown menus or selection boxes. You can select filter values, and the dashboard updates automatically to display relevant data.

How to Use Dashboard Filters

1. If enabled, a filter section appears on the dashboard.
2. Select the desired parameter or column to filter by.
3. Choose a value from the dynamically loaded filter options. For example: a specific API endpoint.
4. The dashboard updates automatically to reflect the selected filter.

Connect+ platform metrics

The Connect+ platform metrics dashboard in the Dev Console gives you end-to-end visibility into the performance of your dataflows. When you select Connect Plus as the product, two dashboards are available from the dashboard drop-down:

• Connect Plus Files — Monitors file-based dataflows that ingest data from FTP or SFTP sources.
• Connect Plus Stream — Monitors stream-based dataflows that ingest data from Kafka or similar event stream sources.

Both dashboards share the same five tabs. The tab names reflect the data source type of the selected dashboard. For example, the first tab is named Files in the Connect Plus Files dashboard and Stream in the Connect Plus Stream dashboard.

Use the date range selector at the top of the page to set a window of up to 30 days. All five tabs reflect data for the selected range.

The sections below describe each tab. All descriptions apply equally to both dashboards. Where tab names are referenced, the Files dashboard naming is used as the example.

Overview tab

The overview tab is the default view when you open a Connect+ dashboard. It gives you a high-level picture of dataflow activity for the selected time range—summary KPIs at the top, a searchable list of in-progress and recent runs, top error codes, and time-series charts for throughput and latency.

For a full description of the overview tab, see Files tab.

Neo Insights tab

The Neo Insights tab shows the performance of Neo API calls triggered by your dataflows. Use it to understand how the Neo layer is behaving—throughput, latency, error rates, and which dataflows are generating the most Neo traffic.

Summary metrics

The top of the tab shows three summary metrics for the selected time range.

Throughput chart

The Throughput chart plots the number of Neo API requests per minute over the selected time range. Use it to identify periods of high ingestion activity and correlate spikes with latency or error increases.

Percentile chart

The Percentile chart plots three latency lines over time.

A sustained gap between P95 and P99 indicates that a small number of requests experience significantly higher latency than the majority.

4xx and 5xx error % trend chart

The 4xx and 5xx Error % Trend chart shows how client-side and server-side error rates change over time. The two lines are:

• 4xx Error % — Percentage of Neo calls that returned a client error.
• 5xx Error % — Percentage of Neo calls that returned a server error.

Use this chart to detect whether error rates are trending upward and to identify the date when errors started.

Parent API calls

The Parent API Calls table lists the Neo API endpoints invoked by your dataflows, along with the dataflow that made each call.

Requests per dataflow (Referer)

The Requests Per Dataflow (Referer) table summarizes Neo API traffic by dataflow, using the request header referer value to identify the originating dataflow.

Core Insights tab

The Core Insights tab shows the performance of Intouch (Capillary Core) API calls made by your dataflows. These are the downstream Capillary platform APIs that Connect+ calls after transforming the source data. Use this tab to distinguish between issues in the Neo layer and issues in the Intouch API layer.

Summary metrics

Throughput chart

The Throughput chart shows the volume of Intouch API calls per minute over the selected time range.

Percentile chart

The Percentile chart plots P90, P95, and P99 latency lines for Intouch API calls over time. See the Neo Insights tab section for the definition of each percentile line.

4xx and 5xx error % trend chart

The 4xx and 5xx Error % Trend chart shows how Intouch API error rates change over the selected time range, with separate lines for 4xx and 5xx errors.

Parent API calls

The Parent API Calls table lists the Intouch API endpoints invoked by your dataflows.

Requests per request URI

The Requests Per Request URI table groups Intouch API traffic by the originating dataflow, using the request header referer value.

Requests per status code

The Requests Per Status Code table groups all Intouch API calls by their HTTP response code.

Errors tab

The Errors tab gives you a complete breakdown of every error that occurred during dataflow processing. Use it to identify which error codes and dataflows are contributing the most failures, trace individual error records to their request IDs, and navigate directly into a specific failed record for deeper investigation.

Error count by status code

The top of the tab shows the total error count for the two most prominent HTTP error codes within the selected time range. Counts are displayed as large figures alongside the HTTP status code.

Trend chart

The Trend chart shows how 4xx and 5xx error counts change over time. The two lines are:

• 4xx Error Count — The number of records that returned a 4xx client error on each date.
• 5xx Error Count — The number of records that returned a 5xx server error on each date.

Use this chart to identify when a spike in errors started and whether it is driven by client-side or server-side issues.

Error message summary

The Error Message Summary table lists individual error records. Each row represents one failed record.

Error count by dataflow name

The Error Count By Dataflow Name table shows the total number of errors per dataflow within the selected time range.

Debug table

The Debug Table provides the most granular view of individual failed records. Use it when you need to trace a specific error to its exact request, block, and processing time.

Performance Summary tab

The Performance Summary tab focuses on slow-processing records—specifically, requests that took more than two seconds to complete. Use it to identify processing bottlenecks at the block level and find which dataflows consistently produce long-running records. Filter the view by dataflow name, block name, lineage ID, or error code using the filter bar at the top of the tab.

Count (>2 seconds)

The summary metric at the top of the tab shows the total number of records that took more than two seconds to process across all dataflows within the selected time range.

Trend chart

The Trend chart shows how the volume of slow records has changed over time. Peaks in the chart indicate periods of degraded processing performance.

Request details

The Request Details table lists individual slow records, one row per request.

Blocks statistics

The Blocks Statistics table aggregates processing data at the block level across all slow records. Use it to identify which blocks are responsible for the highest volume of slow processing or failures.