All Integrations
CloudCloudWatch Metric Streams + IAM

AWS EC2 Integration

Monitor EC2 instances with auto-discovered host metrics, CPU, memory, and network telemetry. Fleet-wide visibility with AI anomaly detection and automatic coverage for Auto Scaling Groups.

Setup

How It Works

01

Deploy CloudFormation Stack

Launch the TigerOps CloudFormation template to create the IAM role and Systems Manager association that installs the TigerOps agent across your EC2 fleet automatically.

02

Auto-Discover EC2 Instances

TigerOps uses the AWS EC2 API to enumerate your instances and applies tag-based filtering. New instances launched from Auto Scaling Groups are discovered and monitored within seconds.

03

Stream CloudWatch + Agent Metrics

Combine CloudWatch namespace metrics (AWS/EC2) with in-instance agent metrics for CPU steal, memory, disk I/O, and process-level data not available via CloudWatch.

04

Set Fleet-Wide Alert Policies

Define CPU, memory, and disk alert thresholds once and apply them to every EC2 instance via tag selectors. TigerOps scales alert coverage automatically as your fleet grows.

Capabilities

What You Get Out of the Box

Auto-Discovery via EC2 Tags

TigerOps continuously queries the EC2 API and groups instances by tag (Environment, Team, Service). New instances are automatically enrolled into the correct dashboards and alert policies.

Full Host Metrics via Agent

CPU (user, system, steal, iowait), memory (used, cached, buffers), disk throughput, network bytes in/out, and per-process CPU and memory — all metrics not available via CloudWatch alone.

Auto Scaling Group Visibility

Track ASG desired vs. running capacity, scale-in/scale-out events, and per-instance health. TigerOps correlates scaling events with application latency and error rate changes.

Instance Type Benchmarking

Compare CPU and memory utilisation across instance types and sizes. TigerOps surfaces over-provisioned instances and suggests right-sizing opportunities based on 30-day usage patterns.

EBS Volume Metrics

Per-volume read/write IOPS, throughput, and latency alongside instance metrics. Detect EBS volume saturation before it causes application timeouts.

AI Fleet Anomaly Detection

TigerOps baselines normal behaviour per instance group and flags statistically unusual instances. Catch a single misbehaving instance in a 500-node fleet without manual threshold tuning.

Configuration

Agent Install via SSM + CloudFormation

Install the TigerOps agent fleet-wide using AWS Systems Manager State Manager.

ec2-agent-setup.sh
# Deploy TigerOps EC2 integration via CloudFormation
aws cloudformation deploy \
  --template-url https://tigerops-cfn.s3.amazonaws.com/ec2-integration.yaml \
  --stack-name tigerops-ec2 \
  --capabilities CAPABILITY_IAM \
  --parameter-overrides \
    TigerOpsApiKey=${TIGEROPS_API_KEY} \
    TagKey=Environment \
    TagValue=production \
    ScrapeInterval=15s

# Or install agent manually on a single instance
curl -sSL https://install.atatus.net/agent | sudo bash -s -- \
  --api-key=${TIGEROPS_API_KEY} \
  --hostname=$(hostname) \
  --tags="env:production,team:platform"

# tigerops-agent.yaml — collector config
receivers:
  hostmetrics:
    collection_interval: 15s
    scrapers:
      cpu:
      memory:
      disk:
      network:
      process:
        mute_process_name_error: true

exporters:
  prometheusremotewrite:
    endpoint: https://ingest.atatus.net/api/v1/write
    headers:
      Authorization: "Bearer ${TIGEROPS_API_KEY}"

service:
  pipelines:
    metrics:
      receivers: [hostmetrics]
      exporters: [prometheusremotewrite]
FAQ

Common Questions

Can TigerOps monitor EC2 instances without installing an agent?

Yes. TigerOps can ingest CloudWatch EC2 metrics (CPU, network, disk) via Metric Streams with no agent required. For memory, swap, and process-level metrics you need the lightweight TigerOps agent, which is installed via AWS Systems Manager Run Command or user data.

How does TigerOps handle ephemeral EC2 instances from Auto Scaling?

TigerOps uses EC2 instance lifecycle events via EventBridge to detect launches and terminations in real time. Metrics from terminated instances are retained for their configured retention period and excluded from live dashboards automatically.

Does TigerOps support Windows EC2 instances?

Yes. The TigerOps agent supports both Linux (deb/rpm) and Windows (MSI installer). Windows instances expose CPU, memory, disk, and network metrics through WMI, with the same fleet-wide dashboard and alert experience as Linux instances.

What is the overhead of the TigerOps agent on EC2?

The TigerOps agent uses under 0.1% CPU and approximately 30 MB of memory on a typical instance. It collects and ships metrics every 15 seconds with local buffering to handle brief network interruptions without data loss.

Can I monitor EC2 instances across multiple AWS accounts?

Yes. TigerOps supports multi-account setups via cross-account IAM roles. Deploy the CloudFormation stack in each account and TigerOps aggregates metrics from all accounts into a unified fleet view with per-account filtering.

Get Started

Full Visibility Across Your Entire EC2 Fleet

Auto-discovery, host-level metrics, and AI anomaly detection for every instance. Deploy in minutes.