> For the complete documentation index, see [llms.txt](https://docs.pharos.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/monitoring/log-monitoring/event-records.md).

# Event Records

Understand the data presented at the event record level and when these are created.

When a monitoring rule's conditions are met, Pharos creates an event record that captures the context of what triggered the rule. These event records serve as both a historical log of monitoring activations and a starting point for incident response.

## When Events Are Created

Event creation occurs the moment a monitoring rule determines that its configured thresholds have been breached. This evaluation happens continuously as logs are ingested, so an event can be created within seconds of the problematic pattern emerging.

Consider a spike rule configured to trigger when 50 errors occur within a 5-minute window. As logs stream in, the monitoring engine tracks the count. At the moment the 50th qualifying log arrives within any rolling 5-minute period, an event is created immediately. The event captures not just that moment, but the full context of the logs that contributed to the threshold breach.

For user impact rules, the same principle applies but with unique user counting. When the configured number of distinct users have experienced errors within the time window, the event is created, capturing which users were affected and what errors they encountered.

{% hint style="info" %}
**Near Real-Time Detection** Event creation happens as close to real-time as the platform allows. For most scenarios, you'll see events created within 1-2 minutes of the triggering condition being met, enabling rapid incident response.
{% endhint %}

## Event Record Data

Each event record contains several categories of information designed to give you immediate insight into what happened:

**Temporal Data** shows when the event was created and the time period it covers. The event visualizes this through an occurrence chart that displays when the triggering logs arrived. This chart often shows a clear pattern—perhaps everything was fine, then suddenly errors spiked, breaching the threshold. This temporal visualization helps you understand whether this was a sudden acute problem or a gradual increase that finally crossed the line.

**Log Data** is the core of the event record. The event maintains references to all logs that triggered the monitoring rule during the evaluation window. You can review these logs directly from the event, examining their details to understand what types of errors occurred. If the monitoring rule used grouping, you'll see how the logs were distributed across those groups.

**Rule Context** connects the event to its source: the event holds a lookup to the monitoring rule that created it, and the Rule Name and Rule Description shown on the event are live formula fields over that lookup—they display the rule's *current* name and description (with a fallback when the description is blank), so renaming a rule or editing its description later changes what past events display. This helps when reviewing events later—you can see exactly what conditions the system detected. If you need to adjust the rule based on what you learn from the event, having this context readily available is valuable.

**Statistical Summary** provides aggregate information about the logs in the event. You can see total log counts, how many unique users were affected, what types of errors occurred, and how the logs break down by category or other dimensions. This summary gives you a quick assessment of severity without needing to dig into individual logs immediately.

## Accessing Related Issues

While monitoring rules operate on logs, those logs have associated issues. From an event record, you can navigate to the logs, and from each log, you can access its parent issue. This connection is crucial for incident response.

When a monitoring event fires, part of the response process involves examining the issues underlying those logs. Perhaps several issues you'd previously triaged as low priority are all involved in this event, suggesting they need priority reassessment. Or maybe the event reveals a new pattern you hadn't recognized when looking at issues individually.

The event serves as a bridge between the log-level analysis that triggered it and the issue-level triaging and action you'll take in response. You're not meant to respond at the log level—that's too granular. Instead, you use the event to identify which issues need attention, then follow your normal issue management workflow.

{% hint style="warning" %}
**Events Don't Replace Issue Triaging** Event records highlight problems that need attention, but the actual remediation work happens at the issue level. Think of events as triggering your incident response process, which then flows into your existing issue management practices.
{% endhint %}

## Event Dashboard and List Views

Events appear in several places within Pharos. The homepage dashboard includes an events section that shows recent events, providing quick visibility into whether any monitoring rules have triggered. This makes events part of your regular operational awareness.

The Events tab provides a full list view where you can sort and filter events by creation date, related rule, or other criteria. This list view is typically the starting point for reviewing what monitoring has detected. You can quickly scan to see if any events require immediate attention or if they're related to known issues being worked.

Since events are created when anomalies are detected, having them surface prominently ensures they don't get lost among the general flow of logs and issues. They represent exception conditions worthy of special attention.

## Event Notifications and Ticketing

Events can trigger notifications just like issues can. When you configure a monitoring rule, you can specify that events created by that rule should generate Teams messages, Slack notifications, or other alerts. This ensures the right people are notified immediately when a monitoring condition is met.

Events can also automatically create tickets in your project management system. This is particularly useful for monitoring rules that detect critical conditions—the event fires, a ticket is created with all the relevant context, and someone is assigned to investigate. The ticket can reference the event record, which maintains the full log data and analysis that triggered it.

The combination of event creation, notification, and ticketing automation enables a fully automated incident detection and response workflow. When something goes wrong at scale, your team knows about it immediately and has the information needed to begin investigation without manual discovery processes.

## Multiple Events from the Same Root Cause

It's possible for a single underlying problem to trigger multiple events if it matches criteria for different monitoring rules. For example, a deployment issue might trigger both a general spike rule and a specific user impact rule. This is generally acceptable because it ensures all relevant monitoring perspectives are captured.

When reviewing events, you may identify that several events are related to the same root cause. This information feeds into your incident response—you're dealing with one problem manifesting in multiple ways, not several independent problems. The event records help you piece together this broader picture.

## Historical Context and Trending

Event records persist even after the immediate incident is resolved. This historical data becomes valuable for understanding patterns over time. You might notice that certain monitoring rules trigger repeatedly around release windows, suggesting a need for better testing or more careful deployment practices.

Reviewing past events also helps with threshold tuning. If a monitoring rule frequently creates events for situations that turn out to be non-issues, you can adjust the thresholds to be less sensitive. Conversely, if you discover critical problems that didn't trigger monitoring, you can create new rules or lower thresholds to catch similar situations in the future.

{% hint style="success" %}
**Events Enable Learning** Each event represents an opportunity to learn something about your org's behavior and health. Review events regularly, even ones that didn't require urgent action, to refine your monitoring strategy and improve your understanding of what "normal" looks like.
{% endhint %}
