--- title: Amazon S3 description: Amazon S3 is a highly available and scalable cloud storage service. breadcrumbs: Docs > Integrations > Amazon S3 --- # Amazon S3 {% alert level="info" %} Monitor S3 metrics and optimize storage costs at the prefix level with [Storage Monitoring (Preview)](https://www.datadoghq.com/product-preview/storage-monitoring/). {% /alert %} ## Overview{% #overview %} Amazon S3 is a highly available and scalable cloud storage service. Enable this integration to see in Datadog all your S3 metrics. **Note**: This integration requires the permission `s3:GetBucketTagging` to be fully enabled. ## Setup{% #setup %} ### Installation{% #installation %} If you haven't already, set up the [Amazon Web Services integration](https://docs.datadoghq.com/integrations/amazon_web_services.md) first. ### Metric collection{% #metric-collection %} 1. In the [AWS integration page](https://app.datadoghq.com/integrations/amazon-web-services), ensure that `S3` is enabled under the `Metric Collection` tab. 1. Install the [Datadog - Amazon S3 integration](https://app.datadoghq.com/integrations/amazon-s3). ### Log collection{% #log-collection %} #### Enable S3 access logs{% #enable-s3-access-logs %} 1. Go to the S3 bucket. 1. Click **Properties**. 1. Go to the Services Access Logging section and click **Edit**. 1. Select **Enable**. 1. Select the S3 bucket to send the logs to. For more information, see [Enabling Amazon S3 server access logging](https://docs.aws.amazon.com/AmazonS3/latest/user-guide/server-access-logging.html). #### Send logs to Datadog{% #send-logs-to-datadog %} **Note**: Datadog's [automatic trigger setup](https://docs.datadoghq.com/logs/guide/send-aws-services-logs-with-the-datadog-lambda-function.md#automatically-set-up-triggers) is available for S3 buckets only. For CloudWatch log groups, use the [manual trigger setup](https://docs.datadoghq.com/logs/guide/send-aws-services-logs-with-the-datadog-lambda-function.md#collecting-logs-from-cloudwatch-log-group). 1. If you haven't already, set up the [Datadog Forwarder Lambda function](https://docs.datadoghq.com/logs/guide/forwarder.md) in your AWS account. 1. Once the Lambda function is installed, there are two ways to collect your S3 access logs: - Automatically: S3 logs are managed automatically if you grant Datadog access with a set of permissions. See [Automatically Set Up Triggers](https://docs.datadoghq.com/logs/guide/send-aws-services-logs-with-the-datadog-lambda-function.md#automatically-set-up-triggers) for more information on configuring automatic log collection on the Datadog Forwarder Lambda function. - Manually: In the AWS console, add a trigger on the S3 bucket that contains your S3 access logs. See the manual installation steps. #### Manual installation steps{% #manual-installation-steps %} 1. If you haven't already, set up the [Datadog Forwarder Lambda function](https://docs.datadoghq.com/logs/guide/forwarder.md) in your AWS account. 1. Once set up, go to the Datadog Forwarder Lambda function. In the Function Overview section, click on **Add Trigger**. 1. Select the **S3** trigger for the Trigger Configuration. 1. Select the S3 bucket that contains your S3 logs. 1. Leave the event type as `All object create events`. 1. Click **Add** to add the trigger to your Lambda. Go to the [Log Explorer](https://app.datadoghq.com/logs) to start exploring your logs. For more information on collecting AWS Services logs, see [Send AWS Services Logs with the Datadog Lambda Function](https://docs.datadoghq.com/logs/guide/send-aws-services-logs-with-the-datadog-lambda-function.md). ## Data Collected{% #data-collected %} ### Metrics{% #metrics %} | | | | | **aws.s3.4xx\_errors**(count) | The total number of HTTP 4xx client error status code requests made to a bucket | | **aws.s3.5xx\_errors**(count) | The total number of HTTP 5xx server error status code requests made to a bucket | | **aws.s3.all\_requests**(count) | The total number of HTTP requests made to a bucket, regardless of type. | | **aws.s3.bucket\_size\_bytes**(gauge) | The amount of data in bytes stored in a bucket in the Standard storage class, Standard - Infrequent Access (Standard_IA) storage class, or the Reduced Redundancy Storage (RRS) class.*Shown as byte* | | **aws.s3.bytes\_downloaded**(count) | The total number bytes downloaded from the bucket.*Shown as byte* | | **aws.s3.bytes\_pending\_replication**(gauge) | The total number of bytes of objects pending replication*Shown as byte* | | **aws.s3.bytes\_uploaded**(count) | The total number bytes uploaded to the bucket.*Shown as byte* | | **aws.s3.delete\_requests**(count) | The number of HTTP DELETE requests made for objects in a bucket. This also includes Delete Multiple Objects requests. | | **aws.s3.first\_byte\_latency**(gauge) | The average per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.maximum**(gauge) | The maximum per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.minimum**(gauge) | The minimum per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.p50**(gauge) | The 50 percentile per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.p90**(gauge) | The 90 percentile per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.p95**(gauge) | The 95 percentile per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.p99**(gauge) | The 99 percentile per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.first\_byte\_latency.p99.99**(gauge) | The 99.99 percentile per-request time from the complete request being received by a bucket to when the response starts to be returned.*Shown as millisecond* | | **aws.s3.get\_requests**(count) | The number of HTTP GET requests made for objects in a bucket. This doesn't include list operations. | | **aws.s3.head\_requests**(count) | The number of HTTP HEAD requests made to a bucket. | | **aws.s3.list\_requests**(count) | The number of HTTP requests that list the contents of a bucket. | | **aws.s3.number\_of\_objects**(gauge) | The total number of objects stored in a bucket for all storage classes except for the GLACIER storage class. | | **aws.s3.operations\_failed\_replication**(gauge) | The average number of operations that failed to replicate for a given replication rule*Shown as operation* | | **aws.s3.operations\_failed\_replication.sum**(count) | The total number of operations that failed to replicate for a given replication rule*Shown as operation* | | **aws.s3.operations\_failed\_replication.samplecount**(count) | The total number of replication operations*Shown as operation* | | **aws.s3.operations\_pending\_replication**(gauge) | The number of operations pending replication*Shown as operation* | | **aws.s3.post\_requests**(count) | The number of HTTP POST requests made to a bucket. | | **aws.s3.put\_requests**(count) | The number of HTTP PUT requests made for objects in a bucket. | | **aws.s3.replication\_latency**(gauge) | The maximum number of seconds by which the replication destination Region is behind the source Region*Shown as second* | | **aws.s3.total\_request\_latency**(gauge) | The average elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.maximum**(gauge) | The maximum elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.minimum**(gauge) | The minimum elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.p50**(gauge) | The 50 percentile elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.p90**(gauge) | The 90 percentile elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.p95**(gauge) | The 95 percentile elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.p99**(gauge) | The 90 percentile elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | | **aws.s3.total\_request\_latency.p99.99**(gauge) | The 99.99 percentile elapsed per-request time from the first byte received to the last byte sent to a bucket*Shown as millisecond* | #### Daily storage metrics{% #daily-storage-metrics %} AWS reports `aws.s3.bucket_size_bytes` and `aws.s3.number_of_objects` one time per day. These metrics might not appear for up to 24 hours after integration setup, and they reflect the previous day's values. After deleting objects or emptying a bucket, `aws.s3.number_of_objects` might continue reporting the previous count until the next daily update from AWS. #### Request metrics{% #request-metrics %} S3 request metrics (such as `aws.s3.get_requests`) require configuration for each S3 bucket in the AWS console. They are not enabled by default. See [Monitoring metrics with Amazon CloudWatch](https://docs.aws.amazon.com/AmazonS3/latest/dev/cloudwatch-monitoring.html) for setup instructions. ### Events{% #events %} The Amazon S3 integration does not include any events. ### Service Checks{% #service-checks %} The Amazon S3 integration does not include any service checks. ## Troubleshooting{% #troubleshooting %} Need help? Contact [Datadog support](https://docs.datadoghq.com/help/).