# CloudWatch alarm for DynamoDB AccountProvisionedWriteCapacityUtilization

CloudWatch alarms · Updated 16 November 2026

**Short answer:** AccountProvisionedWriteCapacityUtilization is the share of your Region's write capacity quota that provisioned tables and indexes have claimed. AWS recommends alarming when its Maximum exceeds 80 for 2 of 2 five-minute periods, which leaves room to raise the quota before tables can no longer scale their writes.

## Why write capacity is usually the tighter quota

The metric adds up the write capacity units (WCUs) provisioned on every provisioned-mode table and global secondary index in the Region and compares the total with the account's write quota. Several things make writes eat that quota faster than reads:

- **Every GSI needs its own WCUs**, sized to absorb the writes it receives from the base table, so a table with four indexes can need several times its own write capacity in total.
- **Global tables** replicate each write into every replica Region, so each replica needs write capacity for its own traffic plus what it receives.
- **Write-heavy batch jobs** prompt teams to raise auto scaling maximums "just in case", and those maximums are what auto scaling will try to reach.

When the quota is reached, `UpdateTable` calls that try to add WCUs fail, auto scaling stalls, and writes beyond the current capacity are throttled.

## Recommended alarm settings

| Setting | Value |
|---|---|
| Namespace | `AWS/DynamoDB` |
| Metric | `AccountProvisionedWriteCapacityUtilization` |
| Dimensions | none (account and Region level) |
| Statistic | `Maximum` |
| Period | 300 seconds |
| Datapoints to alarm | 2 |
| Evaluation periods | 2 |
| Comparison operator | `GreaterThanThreshold` |
| Threshold | 80.0 (percent) |

From the CloudWatch [recommended alarms](https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/Best_Practice_Recommended_Alarms_AWS_Services.html) list. The metric has no dimensions; there is one value per account and Region.

## Tuning

Keep 80% unless you have a reason to need more warning. Add up the auto scaling *maximum* write capacity of your largest tables and indexes; if those maximums together would exceed the quota, the alarm should fire earlier, around 60–70%, because a single busy evening could push several tables to their maximum at once. Check the quota with `aws dynamodb describe-limits`.

## Create it

```
aws cloudwatch put-metric-alarm \
  --alarm-name ddb-account-write-capacity-80pct \
  --namespace AWS/DynamoDB \
  --metric-name AccountProvisionedWriteCapacityUtilization \
  --statistic Maximum \
  --period 300 \
  --evaluation-periods 2 \
  --datapoints-to-alarm 2 \
  --threshold 80 \
  --comparison-operator GreaterThanThreshold \
  --alarm-actions arn:aws:sns:eu-west-2:123456789012:ops-alerts \
  --region eu-west-2
```

Repeat in every Region with provisioned tables, including global-table replica Regions.

## When it fires

1. **Rank tables and indexes by provisioned WCUs.** `aws dynamodb describe-table --table-name orders --query "[Table.ProvisionedThroughput.WriteCapacityUnits, Table.GlobalSecondaryIndexes[].[IndexName,ProvisionedThroughput.WriteCapacityUnits]]"` for each candidate.
2. **Remove indexes nobody queries.** An unused GSI still consumes write capacity on every base-table write.
3. **Lower inflated auto scaling minimums** that keep capacity high around the clock.
4. **Move bursty write workloads to on-demand mode** where the cost profile allows.
5. **Request an increase** to the account-level write capacity quota in Service Quotas, ideally before a planned load test or migration.

If tables are already throttling, the table-level [WriteThrottleEvents alarm](https://cloudconsole.dev/alarms/dynamodb-writethrottleevents) will be firing too; fix the quota first or auto scaling cannot help.

## Questions

### What happens when DynamoDB reaches the account write capacity quota?

You cannot provision more write capacity in that Region, so auto scaling cannot raise busy tables and their excess writes are throttled.

### Do global secondary indexes count towards the account write quota?

Yes. Each provisioned GSI has its own write capacity, and it is included in the account total.

### Why does the alarm use a five-minute period?

The metric reflects provisioned settings, which change only when capacity is updated, so one-minute resolution adds nothing.

## When this happens away from your desk

This alarm gives you hours or days of notice, so it is a planning signal rather than a pager event. CloudConsole, a phone app for AWS, is built to show it with the tables that use the most capacity, so you can pass it to the right owner.

[About the CloudConsole app](https://cloudconsole.dev/aws-console-app) · [Download on the App Store](https://apps.apple.com/app/cloudconsole-for-aws/id6835419710)

## Sources

- [Recommended alarms for AWS services (Amazon CloudWatch User Guide)](https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/Best_Practice_Recommended_Alarms_AWS_Services.html)
- [Troubleshooting throttling issues in Amazon DynamoDB](https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/TroubleshootingThrottling.html)
- [Service Quotas and AWS service limits](https://docs.aws.amazon.com/general/latest/gr/aws_service_limits.html)

## Related

- [CloudWatch alarm for DynamoDB AccountProvisionedReadCapacityUtilization](https://cloudconsole.dev/alarms/dynamodb-accountprovisionedreadcapacityutilization)
- [CloudWatch alarm for DynamoDB WriteThrottleEvents (table)](https://cloudconsole.dev/alarms/dynamodb-writethrottleevents)
- [DynamoDB WriteThrottleEvents alarm for a global secondary index](https://cloudconsole.dev/alarms/dynamodb-gsi-writethrottleevents)
- [ProvisionedThroughputExceededException](https://cloudconsole.dev/errors/dynamodb-provisionedthroughputexceeded)
- [aws dynamodb describe-table](https://cloudconsole.dev/cli/dynamodb-describe-table)
- [CloudWatch alarm for DynamoDB AgeOfOldestUnreplicatedRecord](https://cloudconsole.dev/alarms/dynamodb-ageofoldestunreplicatedrecord)

---

Source: https://cloudconsole.dev/alarms/dynamodb-accountprovisionedwritecapacityutilization

CloudConsole is an independent iPhone and Android app for Amazon Web Services by Cloudconcur Ltd. It is not affiliated with Amazon Web Services.
