Skip to main content
Plan requirement: which grouping methods you can pick depends on your subscription — No grouping is available on every plan, Rule-based grouping requires On-call Standard or higher, and Intelligent grouping requires On-call Pro (selectable by default on-premises). Alert storm warning and inhibit rules require an On-call Standard or higher subscription. Silence rules are available on all plans. Learn more
An alert storm can bring hundreds of similar notifications. Noise reduction groups these alerts into one incident, so you handle it once instead of wading through repeats.

Why noise reduction

Core concepts

Before understanding noise reduction, you need to understand the relationship between three core objects:
Key understanding:
  • One alert can contain multiple events (events with the same alert_key that arrive within the Event Aggregation window, plus the corresponding recovery events)
  • One incident can contain multiple alerts (similar alerts grouped together)
  • Noise reduction happens at two stages: Event Aggregation controls “Event → Alert”; Alert Grouping controls “Alert → Incident”

Noise reduction process

When a monitoring system pushes alerts to Flashduty On-call, the system automatically executes the following workflow:
1

Receive Event

Determine whether to merge into an existing alert, otherwise create a new alert.
2

Process New Alert

Determine whether to merge into an existing incident, otherwise create a new incident.
3

Trigger Notification

New incidents notify relevant personnel according to escalation rules.
4

Subsequent Alerts Merge

Subsequent alerts merge into existing incidents without repeated notifications.
Alert Noise Reduction Flowchart

Let AI configure noise reduction for you

The header of the Alert grouping and Flapping detection pages carries a Let AI tune noise reduction button on the right. It opens AI SRE in the side panel, takes this channel along as a context reference, and prefills a request for the page you are on: /channel-noise is the matching built-in Skill, reachable only from this entry, and it tunes one channel per session. You can send the prefilled request as it is, or rewrite it first, for example to narrow the look-back window or to name the grouping dimensions you must keep. After that you keep talking in plain language to refine what you asked for.
Prerequisite: AI SRE is enabled for the account and you have permission to use AI SRE conversations. The button is hidden when either is missing.
The agent reads the channel’s recent alerts (the look-back window is adjustable, up to 30 days) and recommends values; on the Alert grouping page it also gives an expected incident count from the page’s grouping preview. Nothing is written straight away: you can review the recommended values it fills into the form and click Save, or let it save directly, in which case a confirmation card lists each setting with its before and after value (for example, “Dimensions: labels.check → labels.check + labels.service”) and only your confirmation writes it. The change takes effect immediately. /channel-noise changes Alert grouping and Flapping detection only. For Event aggregation, silence rules, inhibit rules and drop rules it recommends what to change and names the page to change it on, but never writes. For how a reference tag travels with a message and how sessions are managed, see Console chat.

Event aggregation

Go to Channel Details → Configuration → Noise Reduction → Event Aggregation to configure. Event Aggregation controls the “Event → Alert” merge. When the monitoring system keeps pushing events with the same alert_key, it decides whether they merge into the existing alert or each create a new one. An event merges into an existing alert only when all four conditions hold. If any one of them fails, Flashduty creates a new alert:
  • Same channel: the event and the target alert belong to the same channel
  • Same alert_key: the event describes the same alert lifecycle
  • Target alert still active: recovered or ended alerts no longer accept events
  • Still inside the aggregation window: after the window expires, a matching alert_key still creates a new alert
New channels have Event Aggregation enabled by default, with a window of 24 hours (1440 minutes).

Configuration

alert_key is the identifier used for alert correlation and deduplication. It is either reported by the upstream integration or generated automatically by integration rules.

Versus alert grouping

Alert grouping

Go to Channel Details → Configuration → Noise Reduction → Alert Grouping to configure. Alert grouping merges multiple similar alerts into a single incident for unified assignment and notification. When an alert storm hits, you only need to handle one incident instead of hundreds of repeated notifications.
A new channel’s grouping method is derived from your subscription. The create wizard’s Alert Grouping step pre-selects it and marks that card with a Default badge, and saving the channel writes that configuration: free plans default to No grouping, Standard to Rule-based grouping (default grouping field labels.check), and Pro and on-premises deployments to Intelligent grouping (default fields for the similarity judgement: title, description, labels.service, labels.resource). With these defaults the aggregation window is until the incident closes, alert storm warning is off (default threshold 100 once enabled), and strict grouping is off. You can pick another available method in the wizard, or change it later under Noise Reduction → Alert Grouping. For the wizard step, see Create a channel.

Grouping modes

The Alert Grouping page offers three mutually exclusive methods. A choice takes effect only after you save it: In the create wizard, methods your subscription doesn’t include are greyed out, with the tooltip “Your current subscription does not support this grouping method”. The API field group_method carries the same choice: i (intelligent grouping), p (rule-based grouping), n (no grouping); see Incident webhook. Clicking another method card shows that method’s settings right below, with no confirmation dialog, and any draft you filled in before switching is kept. Nothing takes effect until you click Save.

Common configuration

When new alerts are highly similar to active incidents, automatically merge into the incident. The system uses machine learning to calculate semantic similarity between alerts, requiring no manual rule definition.
1

Select Grouping Mode

Pick the Intelligent grouping card under Grouping method at the top of the page (greyed out on Standard, which needs Pro or higher)
2

Select information used for matching

Expand Advanced settings (described as “Similarity fields and storm alerts”), then select the fields under Information used for matching. Default fields include: title, description, labels.service, labels.resource. You can add or remove fields based on your needs, up to a maximum of 10 fields. Click Reset to restore default settings.
The system calculates alert similarity based on the selected fields’ content. When similarity reaches the threshold, new alerts automatically merge into existing incidents.
Intelligent Grouping Configuration

Intelligent Grouping Configuration

Preview rule grouping

After selecting Rule-based grouping and filling in the configuration, click Preview at the bottom to see how the current draft groups a sample of recent alerts. The result includes event, alert, and incident counts, plus the alerts in each incident. Previewing does not save the configuration. Click Save to apply the rule after reviewing the results.

Configuration limits

Grouping configuration has these limits:

Grouping effect

After setting grouping by Alert Check Item, Flashduty groups 5 events into 3 alerts, and the 3 alerts into 1 incident:
The screenshot below shows another incident example. In incident details, you can inspect grouping relationships:
  • Click alert title to view alert timeline and associated events
  • Click event point to view specific event content
Grouping Effect

Grouping Effect Display

Flapping detection

When the same incident triggers and recovers frequently, the system marks it as “flapping” status to avoid notification bombardment. Go to Channel Details → Configuration → Noise Reduction → Flapping Detection:
New channels enable flapping detection in Alert Only mode: Flashduty alerts once when it detects flapping, and later incidents still notify per their policy. The defaults are a 60-minute observation window and 4 state changes. Mute duration applies only in Alert Then Silence mode and defaults to 120 minutes; when you switch the mode to Alert Then Silence, Flashduty pre-fills the mute duration with 120 minutes.
Flapping detection modes

Configurable parameters

What counts as the same incident: Flashduty uses the alert_key of the first alert associated with each incident as the flapping identity. Two incidents are counted as the same problem inside the observation window only when that identity matches. This judgement does not use the grouping dimensions configured on the Alert Grouping page. The alert_key is reported by the alert integration or generated by that integration’s rules — integrations differ, and you cannot change it on this page.The Current effect card at the top of the page states the active mode in one line (for example, “Alert only · alerts once when flapping is detected”).

Silence rules

During maintenance windows or known issue periods, silence rules can suppress alert notifications for specific conditions. Go to Channel Details → Configuration → Noise Reduction → Silence Rules. The Search name or description box above the list filters it by rule name and description, case-insensitively. It does not match the values inside the conditions, and the list shows No matching results when nothing matches.

Silence time

Time input for One-time Silence: a duration input on the left (default 1d) plus an absolute time-range picker on the right. The two controls stay in sync:
  • Duration accepts shorthand such as 30m, 1h, 12h, 1d, 1w, 2w. Changing the duration recomputes the end time from the current moment.
  • You can also pick the start and end directly on the right; the duration input updates accordingly.
  • Start time must be earlier than end time, and neither field may be empty.
Auto-delete on expiration: one-time silence rules can opt into this switch. When it’s on, Flashduty deletes the rule 24 hours after its end time, as it does for Quick Silence. When it’s off (the default), the expired rule stays in the list, inactive, for you to reuse or delete.

Silence conditions

Define which alerts to silence. You can combine multiple conditions. Combination Logic:
  • AND: All conditions must be met to silence
  • OR: Any condition met triggers silence
See Configure Filter Conditions for details.

Silence behavior

The next time you create a silence rule, Flashduty preselects the behavior you chose last.

Quick silence

Quickly create temporary silence rules based on existing incidents. Operation Path: Incident Details → More Actions → Quick Silence
  • Rule name defaults to “Quick Silence - #short-ID”, with the incident title included in the description
  • Effective scope is the incident’s channel (cannot be changed)
  • One-time silence is the default and lasts 1 day. Type a custom duration on the left (such as 30m, 1h, 12h, 1d, 1w, or 2w), or pick an absolute time range on the right. Flashduty deletes a one-time rule after it expires
  • You can instead select recurring silence and configure the silenced time windows with the standard recurring schedule
  • Conditions default to severity plus filtered labels. Flashduty excludes two kinds of labels. System format rules exclude label values that are JSON, numeric, datetime-formatted, or overly long. Account exclusion rules exclude labels globally or per integration (see Quick-silence label exclusion rules)
Quick Silence

Quick Silence Configuration

When you create a quick silence, the drawer lists the labels Flashduty excluded and why. Only the main account or account admins can change account exclusion rules. Flashduty builds the matching conditions from the source incident’s labels. Conditions that are too precise can let similar alerts through, so review them before you create the rule.
When repeatedly using quick silence on the same incident, it edits the original rule rather than creating a new one.

Copy a silence rule

Silence rules are listed in a table; click a row to expand it and see the rule’s details. In that list, open More actions (⋯) on the target rule and choose Copy (not offered for rules created by Quick Silence) to set the new rule’s name, description and target channel: The copy is a new, independent rule that keeps the source rule’s silence conditions, silence behavior, auto-delete setting and schedule, but is no longer linked to the source incident. After copying to another channel, click Go to view to open the new rule.

Inhibit rules

When a root cause alert exists, automatically inhibit related secondary alerts. For example: When a Critical level incident exists, inhibit Warning/Info level incidents for the same check item.

Configuration path

The Search name or description box above the inhibit rule list filters it by rule name and description, case-insensitively. It does not match the values inside the conditions, and the list shows No matching results when nothing matches.

Inhibit conditions

Flashduty inhibits a new alert when all of these hold:
  • The new alert meets the conditions.
  • A matching active alert exists from the last 10 minutes. Any active, unrecovered alert can be the source, acknowledged or not.
  • Both alerts share the equal items.

Inhibit behavior

Configuration example

Scenario: When Critical level alerts exist, inhibit Warning/Info level alerts for the same check item.

FAQ

No. The incident takes its title from the first alert that triggered it. You can edit the title at any time, and new alerts don’t change it.
  • Manually created incidents: No, labels list always remains empty
  • Auto-triggered incidents: Possibly, incident labels stay consistent with the first alert; if that alert’s labels change, incident labels update accordingly
Yes. Alert labels always stay consistent with the latest merged event. However, if the new event is a recovery event, the alert keeps existing labels and only adds labels that didn’t exist before.
Up to about 5000.
The channel’s Event Aggregation setting decides whether an event can merge into an existing alert. It governs the “Event → Alert” stage. It’s on by default with a 24-hour window, which you can set from 1 to 1440 minutes, extend to 7 days on request, or turn off:
  • Event Aggregation enabled: events with the same alert_key merge into the same alert within the window; events arriving after the window expires create a new alert
  • Event Aggregation disabled: every event creates an independent alert, with no merging
Note that Alert Grouping controls the merge window at the “Alert → Incident” stage (24 hours by default, extendable to 30 days). It is a separate concern from whether events merge into an alert — don’t conflate the two.

Configure Escalation Rules

Define alert notification and escalation rules

Configure Filter Conditions

Learn about condition matching syntax