You can now view and customize the detection logic behind Hud’s Heads-Up alerts, giving you control over what triggers an alert and when.

Details

Explore all Heads-Up rules directly in the Hud Web App to understand what each rule monitors.

Adjust detection logic by configuring thresholds such as degradation percentage, invocation counts, and other rule-specific parameters.

Override default settings at the service level to better match the behavior and priorities of each service.

Why It Matters

Alerting is most effective when it reflects your system and your tolerance for change. Custom rule configuration helps reduce irrelevant alerts and ensures you are notified only when something truly matters to your team.

How to Access

Go to the Heads-Up Rules page in the Hud Web App and configure rules per environment and service.

Hud now treats Kafka and SQS as first-class message queues, giving you a unified way to explore, filter, and understand queue-based flows across the system.

Details

The Kafka Topics page is now includes both Kafka and Amazon SQS.

Function-related tables now include SQS flows, with the same fields and structure as existing entities, plus an SQS type indicator.

Why It Matters

Many systems rely on multiple messaging technologies. A unified view across Kafka and SQS makes it easier to reason about asynchronous flows, trace function involvement, and analyze performance without treating each queue type as a separate mental model.

How to Access

• Open Queues page from the main navigation to view Kafka and SQS together.
• On Function pages, use Filter by Flow or review tables to see where functions participate in SQS-based flows.

You can now see exactly where a function participates across endpoints and topics, including when it was first and last invoked within each flow.

Details

• Added a new top-level Exists in Flows table on the Function page.
• The table lists all endpoints and topics where the function appears.
• Each row includes service, entry point, type, runs per day within the flow, distribution of total runs, first tracked invocation, and last tracked invocation.
• Functions that are the first invoked in a flow are marked with a Directly called from tag, including an explanatory tooltip.
• Supports free-text search by entry point name and applies to both endpoint and Kafka flows.

Why It Matters

This view makes it easier to understand how and when a function is used across different flows. It adds historical and structural context that helps teams trace behavior changes and investigate issues more effectively.

How to Access

Open any Function page in the Hud web app. The Exists in Flows table appears at the top of the page.

You can now filter metrics tables using numeric thresholds, making it easier to focus on meaningful results and exclude low-signal functions or endpoints.

Details

• Added min and max threshold filters for runs per day and average duration.
• Filters are available across Functions tables on the All Functions page, Service pages, Endpoint pages, and Topic pages.
• Threshold filtering is also supported in Endpoints and Topics tables, including E2E duration for Kafka.
• Filters work consistently with table sorting, filter presets, and filter updates or removals.
• Analytics events track filter usage, including filter type, comparison operator, and selected values.

Why It Matters

Sorting alone can surface functions with very low traffic or trivial durations that are not actionable. Threshold filters help you narrow the table to entities with enough volume or impact to be worth investigating, leading to more relevant insights and better prioritization.

How to Access

Open any supported metrics table in the Hud web app and use the table filters to set minimum or maximum values for runs per day or average duration.

P90 duration is now available across tables and graphs, giving you a clearer view of typical performance without being skewed by rare outliers. You can also control which duration lines are shown in graphs to improve readability when scales differ.

Details

P90 duration added to duration graphs across Functions, Endpoints, Queues, and the IDE Research tab.

New P90 + trend column added to Functions tables, including the Service page Functions table.

Graph customization now lets you show or hide individual duration lines (for example Avg or P90) to adjust the y-axis scale and focus on the metric you care about.

Why It Matters

P90 highlights common performance patterns while excluding extreme anomalies, making it easier to spot real trends and prioritize issues. Being able to hide graphs also improves investigation when one metric dominates the scale, so insights are easier to read and compare.

How to Access

1. Open any Functions, Endpoints, or Topics page to view P90 in duration graphs.
2. Sort Functions tables by the new P90 + trend column.
3. Use the graph customization pane to toggle which duration lines are displayed.

Whenever your endpoint hits an HTTP 500, we’ll send a heads-up alert and highlight the downstream function causing trouble 🔍

Why it matters:

Real-time: Get notified the moment an error appears

Root Source: See which function is acting up in a glance

For more info on other alert types, check out our docs: Heads-Up Alerts Overview.

View those and stay tuned for more Heads-Up updates in Hud’s channel

"My Services" lets you configure which services you’re focusing on, so that by default you’ll only see their data across all of Hud’s products - whereas all the other services will be hidden by default.

To get started, go to the Services page and click the ⭐ next to each service you’re interested in. You can easily remove this filter when viewing functions, endpoints and heads-up issues so you can view all services or specific other services.

Let’s go configure your services

Hud x JetBrains: We are proud to announce Hud's extension for JetBrains IDEs is out of beta!

You can install it now from JetBrains Marketplace.

You can now see all Hud's Heads-Ups that were detected in the past week.

Navigate to the root cause function in the code directly from the heads-up. Select your service to filter the issues.

Look for the Hud logo on the left to launch it.

Introducing a shiny new alert: Post-Deployment Endpoint Degradation with Root Cause Analysis

We now automatically detect production deployments (new or updated functions) and alert on performance degradation at intervals of 2m, 10m, 30m and hourly up to 24 hours post-deployment.

Just like our classic alerts, we flag endpoints with at least 100% degradation (double the average duration) by comparing post-deployment performance to the week prior. We also pinpoint likely downstream functions causing the issue. 🔍

We’d love to hear your feedback and ideas!

Introducing Heads-Up page on the Web!

You can now view all your recent Heads-Ups in the web app or directly on your IDE!

Check out now the Heads-Up table on the web!

Meet the New Endpoints Page

See how your endpoints perform in production, detect trends, and access functions invoked in the endpoint’s flow.

Check out your endpoints data at app.hud.io/explore/endpoints