Amazon EC2 Observability | Germain Documentation

☁️ Amazon EC2 Observability

👀 Overview

GermainUX monitors the availability, performance, and capacity of Amazon EC2 instances and their attached EBS volumes.

Telemetry can be collected from:

Source
Amazon CloudWatch
EC2 status checks
EBS volume metrics
Operating-system commands and metrics
Processes and services
Application and system logs
Application endpoints
Synthetic transactions

This provides a combined view of AWS infrastructure, operating-system health, application availability, and user or business impact.

🏗️ Monitoring Architecture

The GermainUX Engine connects to configured AWS services and monitored resources. Depending on the required visibility, it can:

Action
Retrieve EC2 and EBS metrics exposed through CloudWatch.
Monitor Linux hosts through SSH.
Monitor Windows hosts through approved Windows management interfaces.
Read accessible application and system logs.
Test application endpoints and business transactions.
Execute approved diagnostic or corrective actions.

The Engine requires network access to GermainUX, the appropriate AWS endpoints, and any EC2 resources monitored directly.

📋 Prerequisites

Before configuring EC2 monitoring:

Deploy and register a GermainUX Engine.
Identify the AWS accounts, regions, instances, and volumes to monitor.
Create an AWS identity with the minimum required read permissions.
Configure approved credentials in GermainUX.
Confirm that the Engine can reach the required AWS endpoints.
Install and configure the CloudWatch Agent if guest operating-system metrics are required through CloudWatch.
Configure SSH or Windows access if direct operating-system monitoring is required.
Synchronize clocks across the monitored systems.

Do not use an AWS account root user or broadly privileged identity for monitoring.

💽 EC2 Instance Monitoring

GermainUX can collect available CloudWatch metrics such as:

Measurement	Description
CPU utilization	Percentage of allocated compute capacity in use
Network input and output	Bytes received and transmitted
Network packets	Packets received and transmitted
Disk read and write operations	Instance-store disk operations, where available
Disk read and write bytes	Instance-store data transfer, where available
Instance status check	Combined instance health
System status check	Health of the AWS infrastructure supporting the instance
Attached EBS status check	Reachability of attached EBS volumes, when available
CPU credit balance	Remaining CPU credits for burstable instances
CPU credit usage	CPU credits consumed
CPU surplus credits	Surplus credit activity for applicable unlimited-mode instances

Metrics depend on the instance type, operating system, AWS configuration, and CloudWatch data available for the resource.

Standard EC2 CloudWatch metrics do not include guest memory or filesystem utilization. Collect those measurements using the CloudWatch Agent or direct operating-system monitoring.

🗄️ EBS Volume Monitoring

GermainUX can monitor available EBS metrics including:

Metric
Read and write operations
Read and write bytes
Operation latency
Queue length
Idle time
Throughput utilization
IOPS utilization
Burst balance
Volume status checks

🔋 EBS Burst Balance

EBS Burst Balance indicates the percentage of burst credits remaining for volume types that use a burst-credit model.

A sustained decline can indicate that the volume is consuming burst capacity faster than it recovers. When the balance is depleted, storage performance may be limited to the volume’s baseline capability.

Configure alerts before the balance reaches a critical level and correlate it with:

Correlate With
IOPS
Throughput
Queue length
Latency
Application response time
Database performance
Business-transaction duration

Burst Balance is not applicable to every EBS volume type. For example, gp3 volumes use provisioned performance rather than the gp2 burst-credit model.

🖥️ Operating-System Monitoring

For Linux or Windows running on EC2, GermainUX can monitor available guest-level measurements such as:

Measurement
CPU and load
Physical and virtual memory
Swap or paging
Filesystem utilization
Disk activity
Network interfaces
Process availability
Process CPU and memory
Service state
System uptime
System and application logs

Direct operating-system monitoring requires appropriate connectivity and a dedicated least-privilege account.

⚙️ Process and Service Monitoring

Configure GermainUX to monitor processes and services required by the application, including:

Process / Service
Web and application servers
Databases
Integration services
Message brokers
Background workers
Scheduled services
Custom application processes

Alerts can identify stopped processes, unexpected process counts, excessive resource consumption, or repeated restarts.

🔗 Application and Endpoint Monitoring

Infrastructure availability does not guarantee that the hosted application is usable. GermainUX can also monitor:

What
HTTP and HTTPS endpoint availability
Response time
Status codes
API transactions
Authentication flows
Database queries
Application logs
Browser user experience
Business processes
Synthetic user scenarios

Correlating EC2 and EBS telemetry with application performance helps determine whether infrastructure conditions are affecting users or business transactions.

📄 Log Monitoring

The GermainUX Engine can monitor accessible logs for:

Log Type
Application errors
Service startup and shutdown
Authentication failures
Resource warnings
Disk and filesystem problems
Network errors
Database failures
Out-of-memory conditions
Deployment events

Logs can be read from accessible files or integrated sources. Configure parsing, categorization, privacy, and retention rules for each format.

Do not collect credentials, tokens, personal information, or other unnecessary sensitive data from logs.

🔧 Configure EC2 Monitoring

Go to Workspace > Settings > Wizards.
Select the Amazon EC2 or AWS monitoring configuration.
Select the GermainUX Engine.
Configure the approved AWS credentials.
Select the AWS account and region.
Select the instances and EBS volumes to monitor.
Choose the required metrics and collection frequency.
Add operating-system, process, log, or endpoint monitoring where required.
Configure KPIs, SLAs, alerts, and dashboards.
Save and enable the configuration.
Confirm that telemetry appears in GermainUX.

The available configuration fields may vary by GermainUX version and AWS monitoring scope.

🏷️ Resource Identification

Use consistent AWS and GermainUX attributes to identify resources, including:

Attribute
AWS account
Region
Availability Zone
Instance ID
Instance name
Instance type
Volume ID
Application
Environment
Business service
Owner or support team

AWS tags can provide operational context when available through the configured integration. Avoid relying only on hostnames, which may change when instances are replaced.

Recommended Alerts

Configure alerts for conditions such as:

Condition
EC2 instance unavailable
Failed instance or system status check
Sustained high CPU utilization
Low CPU credit balance
Unexpected network-traffic changes
High EBS latency or queue length
Low EBS Burst Balance
High IOPS or throughput utilization
Low filesystem capacity
Memory pressure or excessive paging
Required process or service stopped
Application endpoint unavailable
Repeated application or system errors
Missing monitoring data

Use sustained-duration and recurrence conditions to avoid alerts from brief, harmless spikes.

📊 Dashboards and Reports

EC2 dashboards can include:

Dashboard Item
Instance availability
Status-check failures
CPU utilization
CPU credit balance
Network activity
EBS IOPS and throughput
EBS latency and queue length
EBS Burst Balance
Memory and filesystem utilization
Process and service availability
Application endpoint performance
Log errors
Related user or business-process impact

Scheduled reports can summarize performance, availability, capacity, SLA violations, trends, and changes compared with the previous period.

🤖 Automation

When an issue is detected, GermainUX can initiate approved actions through the Engine, including:

Action
Sending an alert
Creating or updating a ticket
Calling a webhook or HTTP endpoint
Running an authorized diagnostic script
Collecting additional evidence
Restarting an approved service
Validating recovery

Actions that stop or start EC2 instances, detach volumes, modify infrastructure, or affect production services must be explicitly authorized and protected by appropriate safeguards.

🔒 Security

Apply the following controls:

Control
Use a dedicated AWS identity.
Grant only the required read permissions.
Prefer temporary credentials or assigned roles where supported.
Protect and rotate stored credentials.
Restrict direct host access by network policy.
Use dedicated least-privilege operating-system accounts.
Encrypt connections.
Audit automated actions.
Avoid collecting unnecessary sensitive data.

✅ Validation

After enabling monitoring:

Confirm that the Engine can access the configured AWS account and region.
Verify that the expected instances and volumes are discovered.
Compare selected values with CloudWatch.
Confirm that instance and volume identifiers are correct.
Verify operating-system and process telemetry, when configured.
Test an alert condition safely.
Confirm that alerts reach the appropriate recipients.
Verify that dashboards show the expected application and environment context.

ℹ️ Get Help

The Germain Team can help you set this up. Contact GermainUX Support.

Component: Engine, JS Profiler, Mobile App, RPA Bot Recorder, RUM JS

Feature Availability: 2022.1 or later