AWS Compute Optimizer vs ZopNight: What Each One Flags
AWS gives you two native places to find waste: Compute Optimizer and Trusted Advisor's cost checks. Here is what each covers, per AWS documentation, and what ZopNight adds.
AWS Compute Optimizer
Compute Optimizer “analyzes your AWS resources’ configuration and utilization metrics to provide you with rightsizing recommendations and identify idle resources” (user guide). You have to opt in, per account or across an organization from the management account.
What it covers. The supported list is longer than its name suggests: EC2 instances, EC2 Auto Scaling groups, EBS volumes, Lambda functions, ECS services on Fargate, commercial software licenses, Aurora and RDS databases, NAT Gateway, DynamoDB, ElastiCache, MemoryDB, DocumentDB, WorkSpaces and SageMaker.
How far back it looks. By default it reads the last 14 days of CloudWatch metrics. Enhanced infrastructure metrics, a paid preference, extends that to up to 93 days for EC2 instances, instances in Auto Scaling groups, and RDS DB instances (enhanced metrics). Idle checks use their own windows: EBS unattachment and NAT Gateway are judged over 32 days, and WorkSpaces over 63 days (idle recommendations).
What counts as idle. An EC2 instance is idle when “peak CPU utilization is below 5% and your network I/O is less than 5MB/day over the 14-day lookback period”. Idle recommendations refresh daily.
What it costs. AWS states “there is no additional charge for AWS Compute Optimizer”. Enhanced infrastructure metrics cost $0.0003360215 per resource per hour, which AWS puts at about $0.25 for a resource running all of a 31-day month (pricing).
It can now act, narrowly. A newer Automation feature applies “recommended actions”, which AWS calls “a subset of Compute Optimizer’s recommendations”. Two action types are listed: snapshot and delete EBS volumes unattached for 32 or more days, and upgrade previous-generation EBS volume types. You can apply up to 10 at a time or set rules that run daily, weekly or monthly (recommended actions).
Trusted Advisor cost checks
Trusted Advisor has its own cost optimization checks (check reference). Two examples:
- The Low utilization Amazon EC2 instances check flags an instance with 10% or less daily average CPU and 5 MB or less network I/O on at least 4 of the previous 14 days.
- Amazon RDS idle DB instances flags a DB instance with no connection in the last 7 days.
Access depends on your support plan. On Basic Support, AWS says you get “all checks in the Service Limits category and selected checks in the Security and Fault tolerance categories” (Trusted Advisor). Cost checks are not in that list.
Where ZopNight differs
- Part-time waste. Both AWS tools look for resources that are idle across their whole window. A staging database used every weekday is not idle by those rules. ZopNight’s schedule recommendations look for the hours a resource could be off, and a schedule then stops and starts it. See RDS scheduling.
- More levers. ZopNight’s remediation wizard can stop, pause, resize or delete, with an admin approval step for destructive actions. Commitment (RI and CUD) recommendations stay advisory, with the rates and break-even behind each one.
- Other clouds. The same recommendation list covers GCP and Azure, plus Databricks and Snowflake connections.
- Cost as billed. On AWS, ZopNight reads Cost Explorer or, if you point it at one, your CUR 2.0 export in S3, which keeps resource-level history beyond the 14 days of daily resource-level data that Cost Explorer offers (AWS docs).
| ZopNight | AWS Compute Optimizer | |
|---|---|---|
| Clouds | AWS, GCP, Azure | AWS |
| Default lookback | Latest 21 days for schedule recommendations | 14 days; up to 93 with paid enhanced metrics |
| Actions it can take | Start/stop schedules, stop, pause, resize, delete with approval | Snapshot-and-delete unattached EBS volumes, upgrade EBS volume type |
| Off-hours scheduling | Yes | Not among its documented actions |
| Price | See zop.dev/pricing | No charge; enhanced metrics billed per resource-hour |
Choose Compute Optimizer if…
- You are AWS-only and want free rightsizing across EC2, Lambda, Fargate, RDS and more.
- Your workloads have quarterly patterns, and paying about $0.25 a resource a month for 93 days of history is worth it.
- Cleaning up unattached EBS volumes automatically covers most of the action you need.
Choose ZopNight if…
- Most of your waste is non-production resources running nights and weekends, not resources that are idle all the time.
- You want recommendations for AWS, GCP and Azure in one list, each with the metric evidence that triggered it, exportable to Excel for leadership.
- You want a finding to become an action in the same tool: a schedule, a remediation or a Jira ticket.
Questions we get a lot.
If yours isn't here, email us and we'll answer directly.
Is AWS Compute Optimizer free?
Yes. AWS says there is no additional charge for Compute Optimizer. The optional enhanced infrastructure metrics feature, which extends the lookback to up to 93 days, costs $0.0003360215 per resource per hour.
What lookback period does Compute Optimizer use?
14 days by default. With enhanced infrastructure metrics it can use up to 93 days for EC2, Auto Scaling group instances and RDS. EBS unattachment and NAT Gateway idleness are judged over 32 days, and WorkSpaces over 63 days.
Can Compute Optimizer stop idle EC2 instances for me?
Not as of 25 September 2026. Its Automation feature lists two action types, both for EBS volumes: snapshot and delete unattached volumes, and upgrade previous-generation volume types.