Maintenance Impact on Monitoring

This guide explains how maintenance events impact the status displayed for monitors on your status page, and the precedence rules that determine which status is shown.

Status Override Concept

When a maintenance event is ONGOING, the monitor statuses on your status page are temporarily overridden to show the configured maintenance impact instead of realtime monitoring data.

Key Principle: During maintenance, users see the expected status (configured in maintenance), not the actual monitoring results.

Why Override?

Prevents false alarms during planned work
Communicates expected behavior to users
Distinguishes planned vs unplanned disruptions

Monitor Status Precedence

Kener uses a precedence system to determine which status to display for a monitor. Higher priority statuses override lower ones.

Precedence Order (Highest to Lowest)

1. Maintenance Event (ONGOING)
2. Active Incident
3. Realtime Monitoring Data
4. Default Monitor Status

How it Works:

Check if monitor is in an ONGOING maintenance event → Use maintenance impact
Else, check if monitor is in an OPEN incident → Use incident impact
Else, use latest realtime monitoring data
If no data, use monitor's default status

Detailed Precedence Rules

1. Maintenance Event Status (Highest Priority)

When: Maintenance event status is ONGOING and monitor is affected

Override: Monitor shows the configured monitor_impact from maintenance

Example:

Monitor: API Server
Realtime Status: UP (all checks passing)
Maintenance: ONGOING
Monitor Impact: MAINTENANCE

Displayed Status: MAINTENANCE

Rationale: Users need to know service is under maintenance, even if currently responding.

2. Active Incident Status

When: Incident status is OPEN and monitor is affected

Override: Monitor shows the configured monitor_impact from incident

Example:

Monitor: Database
Realtime Status: UP (checks passing now)
Incident: OPEN
Monitor Impact: DOWN

Displayed Status: DOWN
title: Impact on Monitoring
description: How maintenance events affect displayed monitor status
Rationale: Incident status takes precedence over current checks during active issues.

During an ONGOING maintenance event, maintenance impact can override monitor status shown to users.

Status precedence

Kener resolves status in this order (later overrides earlier):

default status → realtime monitor result → incident impact → maintenance impact

So maintenance has the highest effective priority when active.

Impact values

Monitor: Web Server
Set per monitor in a maintenance:

MAINTENANCE (recommended)
DOWN
DEGRADED
UP

Choose the value that matches expected user impact during the window.

Event lifecycle behavior

SCHEDULED / READY: no override yet
ONGOING: override active
COMPLETED / CANCELLED: override removed

Realtime monitoring still runs

Monitor: New Service
Even during maintenance, checks continue and data is recorded.
Default Status: UP
Maintenance changes displayed/effective status, not monitor execution.
Displayed Status: UP

Practical guidance

Prefer MAINTENANCE for planned work communication.
Use DOWN only when service is expected to be unavailable.
Avoid overlapping maintenances on the same monitor.

Visual: Red, X icon

Meaning: Service is completely unavailable

When to Use:

Service will be offline
No access during maintenance
Total outage expected

User Interpretation: "Service is unavailable"

Example:

Monitor: Database Server
Impact: DOWN
During Event: Shows red "Down for Maintenance"

DEGRADED

Visual: Yellow, warning icon

Meaning: Service is partially available or slow

When to Use:

Service in read-only mode
Reduced capacity
Slower response times expected

User Interpretation: "Service has limited functionality"

Example:

Monitor: API Server
Impact: DEGRADED
During Event: Shows yellow "Degraded Performance"

UP

Visual: Green, checkmark icon

Meaning: Service is fully operational

When to Use:

Rare, for non-disruptive maintenance
Adding monitoring to existing service
Backend work with no user impact

User Interpretation: "Service is operational"

Example:

Monitor: Web Server
Impact: UP
During Event: Shows green "Operational" (rare use case)

Status During Event Lifecycle

Monitor status override behavior changes based on event status:

SCHEDULED Events

Override: ❌ No

Displayed Status: Realtime monitoring data

Visibility: Event shown in "Upcoming Maintenances" section

Example:

Event Status: SCHEDULED
Event Start: In 3 hours
Monitor Status: UP (from realtime checks)
Display: UP (no override yet)

READY Events

Override: ❌ No

Displayed Status: Realtime monitoring data

Visibility: Event shown as "Starting Soon", notification sent

Example:

Event Status: READY
Event Start: In 45 minutes
Monitor Status: UP (from realtime checks)
Display: UP (no override yet)

ONGOING Events

Override: ✅ Yes

Displayed Status: Configured maintenance impact

Visibility: Event prominently displayed, monitors show maintenance status

Example:

Event Status: ONGOING
Monitor Impact: MAINTENANCE
Monitor Realtime: UP
Display: MAINTENANCE (override active)

COMPLETED Events

Override: ❌ No

Displayed Status: Realtime monitoring data (restored)

Visibility: Event shown in "Past Maintenances" section

Example:

Event Status: COMPLETED
Monitor Impact: (was MAINTENANCE)
Monitor Realtime: UP
Display: UP (override removed)

CANCELLED Events

Override: ❌ No

Displayed Status: Realtime monitoring data

Visibility: Event marked as cancelled in history

Multiple Maintenances on Same Monitor

If multiple maintenance events are ONGOING for the same monitor simultaneously:

Precedence: First maintenance event found (by database query order)

Behavior:

Query finds all ONGOING maintenance events for monitor
Use impact from first event returned
Others are ignored for status purposes

Example:

Monitor: API Server

Maintenance A: ONGOING
- Impact: DOWN
- Event ID: 101

Maintenance B: ONGOING
- Impact: DEGRADED
- Event ID: 102

Result: Whichever is returned first by query wins
Recommendation: Avoid overlapping maintenances

Best Practice: Don't schedule overlapping maintenances for the same monitor.

Realtime Monitoring During Maintenance

Key Point: Realtime monitoring continues during maintenance events.

What Happens:

Monitor checks continue to run on schedule
Data is recorded in monitoring_data table
Status page shows maintenance impact (override)
Backend data available for analysis

Why Continue Monitoring?

Verify service comes back after maintenance
Detect unexpected issues during maintenance
Maintain complete monitoring history
Automatic transition when maintenance ends

Example:

During ONGOING maintenance event:

Monitor Checks:
- 3:00 AM: DOWN (expected, service offline)
- 3:15 AM: DOWN
- 3:30 AM: DOWN
- 3:45 AM: UP (service restored)
- 4:00 AM: UP

Status Page Display:
- 3:00-4:00 AM: MAINTENANCE (override)
- After 4:00 AM: UP (realtime, override removed)

Backend Data: All checks recorded

Impact on Uptime Calculations

Maintenance windows affect uptime calculations based on your monitor configuration:

Default Behavior

Maintenance Status Impact:

MAINTENANCE impact: Excluded from downtime (treated as UP)
DOWN impact: Included in downtime
DEGRADED impact: Depends on include_degraded_in_downtime setting

Rationale: Planned maintenance shouldn't penalize uptime metrics.

Configuration

Monitor setting: include_degraded_in_downtime

"YES" - DEGRADED counts as downtime
"NO" - DEGRADED counts as uptime

Examples

Scenario 1: Maintenance Impact = MAINTENANCE

Monitor: API Server
Maintenance: 1 hour MAINTENANCE impact
Realtime Data: DOWN during maintenance
Uptime Calculation: Excluded from downtime
Result: Uptime maintained at ~100%

Scenario 2: Maintenance Impact = DOWN

Monitor: Database
Maintenance: 1 hour DOWN impact
Realtime Data: DOWN during maintenance
Uptime Calculation: Counted as downtime
Result: Uptime reduced

Scenario 3: Maintenance Impact = DEGRADED

Monitor: API Server
Maintenance: 1 hour DEGRADED impact
include_degraded_in_downtime: NO
Uptime Calculation: Counted as uptime
Result: Uptime maintained

Status Page Display

Home Page

Ongoing Maintenances Section:

Listed at top of page
Shows maintenance title and affected monitors
Monitors display maintenance impact status
Countdown/duration shown

Monitor List:

Monitors show override status during ONGOING events
Badge/icon indicates under maintenance
Tooltip explains maintenance in progress

Monitor Detail Page

Status Banner:

Shows current status (maintenance override if applicable)
"Under Maintenance" banner when override active
Link to maintenance event details

Status History Chart:

Maintenance periods highlighted
Different visualization for maintenance vs incidents
Clearly distinguishable from actual downtime

Uptime Stats:

Calculated according to configuration
Maintenance windows may be excluded
Footnote explains calculation method

Maintenance Events Page

Dedicated page showing:

All ongoing maintenances
Upcoming maintenances (next N days)
Past maintenances (last N days)
Filterable by monitor, date range

Troubleshooting

Problem: Monitor still shows DOWN during ONGOING maintenance

Solutions:

Verify event status is exactly "ONGOING" (not READY or SCHEDULED)
Check monitor is in maintenance_monitors table for this event
Verify maintenance status is ACTIVE
Check current time is between start and end time

Problem: Maintenance ended but monitor still shows MAINTENANCE

Solutions:

Wait for status update scheduler (runs every minute)
Check event status transitioned to COMPLETED
Verify current time is past end_date_time
Check for overlapping maintenance events

Problem: Monitor shows realtime status during maintenance

Solutions:

Verify event status is ONGOING (not SCHEDULED or READY)
Check monitor_tag matches exactly
Verify maintenance is ACTIVE
Check status precedence (incident may override maintenance)

Problem: Uptime affected by maintenance window

Solutions:

Check monitor impact level (DOWN counts as downtime)
Use MAINTENANCE impact to exclude from downtime
Verify uptime calculation formula in monitor settings
Review include_degraded_in_downtime setting

Best Practices

Use MAINTENANCE Impact:

✅ Impact: MAINTENANCE
Reason: Clear communication, excluded from downtime

Use DOWN Only When Truly Offline:

✅ Impact: DOWN for complete outage
❌ Impact: DOWN for partial degradation

Avoid Overlapping Maintenances:

✅ Schedule maintenances at different times
❌ Multiple ONGOING events for same monitor

Match Impact to Reality:

✅ Database offline → DOWN
✅ Read-only mode → DEGRADED
✅ Backend update, no user impact → MAINTENANCE or UP

Communicate Clearly:

Title: "Database Upgrade to v2.0"
Impact: DOWN
Description: "Database will be completely offline during this window"

Next Steps

Maintenance Events - Learn about event lifecycle and status transitions
Creating and Managing Maintenances - How to configure monitor impacts
RRULE Patterns - Advanced scheduling patterns

Maintenance Impact on Monitoring

Status Override Concept

Monitor Status Precedence

Precedence Order (Highest to Lowest)

Detailed Precedence Rules

1. Maintenance Event Status (Highest Priority)

2. Active Incident Status

Status precedence

Impact values

Event lifecycle behavior

Realtime monitoring still runs

Practical guidance

Related guides

DEGRADED

UP

Status During Event Lifecycle

SCHEDULED Events

READY Events

ONGOING Events

COMPLETED Events

CANCELLED Events

Multiple Maintenances on Same Monitor

Realtime Monitoring During Maintenance

Impact on Uptime Calculations

Default Behavior

Configuration

Examples

Status Page Display

Home Page

Monitor Detail Page

Maintenance Events Page

Troubleshooting

Best Practices

Next Steps