ServiceNow Instance Observer provides a comprehensive set of alerts designed to monitor platform health, performance, and user experience. These alerts are categorized for easy consumption and actionability.

Transactions

Monitors application transactions for anomalies, spikes, or degradations in performance such as:

• Transaction Decrease: Detects a drop in total transaction volume
• Transaction Decrease Node: Identifies transaction volume drop per node
• Transaction Increase: Flags unexpected transaction surges
• Transaction Increase Node: Highlights node-level transaction spikes
• Response Time: Triggers when system-wide response time increases
• Response Time Node: Flags nodes with degraded response times
• Database Response Time: Monitors database-level latency impacting transactions
• Slow Queries Per Second: Identifies the volume of slow database queries affecting responsiveness

Node health (CPU, memory, or garbage collection)

Tracks node infrastructure health to avoid bottlenecks or failures:

• Node CPU time: High CPU usage alert for a node
• Node memory: Monitors memory consumption patterns
• Node garbage collection time: Tracks JVM GC delays
• Load balancer container CPU utilization: Flags CPU overload on LB containers
• Load balancer container memory utilization: Detects memory exhaustion on LB containers

Database performance and health

Covers critical database indicators to verify query health and data reliability:

• Database host health CPU: High CPU on primary DB host
• Shards host health CPU: Resource issues on shard hosts
• Read replica host health (CPU): Read-replica CPU anomalies
• Standby replication lag: Lag in standby DB replication
• InnoDB row lock: Frequency of row lock waits
• Primary database growth: Flags abnormal growth in primary DB
• Database table growth: Specific table-level growth indicators

Inbound and outbound email

Promotes timely delivery and ingestion of email-based communications:

• Outbound email: Delays or failures in outbound email processing
• Inbound email: Issues in ingesting incoming emails

Scheduler and job execution

Helps detect issues in the job execution life cycle:

• Scheduler stuck: Scheduler not progressing or blocked
• Long-running jobs: Jobs exceeding typical run time
• Specific long-running jobs: Custom job monitoring
• Thread running: Threads running unusually long or in high volume

Session and user activity

Tracks user login behavior across instance and nodes:

• User session logged in – Instance: Log in activity across instance
• User session logged in – Node: Node-wise session metrics

Event queue and semaphore management

Critical for debugging platform event handling and job execution throttling:

• Default semaphore mean: Semaphore wait time trends
• Default semaphore QDepth: Depth of queued semaphore requests
• Integrated semaphore: Monitors integrated semaphore contention
• Event queue check: Tracks backlog in event queues
• Specific queue for events: Custom event queue monitoring
• High priority event queue: Monitors mission-critical event queues
• ECC queue: External communication channel backlog alerts

Asynchronous Messaging Bus (AMB)

Internal messaging bus observability for real-time app behavior:

• AMB send queue depth: Size of outgoing message queue
• AMB send in use: Utilization of AMB sending capacity

Historical or list data volume

Monitors growth of historical or list data that can impact performance:

History list length: Flags excessive record count in history tables.

Application host health

Monitors health at the application layer:

Application host health CPU: Application-tier CPU overload alerts.

AI/ML or intelligent alerting

Includes alerts generated via AI/ML-based behavior analysis:

Auriga Intelligent: AI-driven anomaly or pattern detection alerts.