Metric Sources
Using Instrumental
- Metrics
- Graphs
- Alerts
- Notices
- Querying
- Where to Start
InstrumentalD
- Installation
- Configuration
- System Metrics
- Service Metrics
- - Docker
  - MySQL
  - Memcached
  - MongoDB
  - Nginx
  - PostgreSQL
  - Redis
- Plugin Scripts
Metrician
API Reference

AWS CloudWatch

Instrumental can automatically import CloudWatch data from your AWS account and generate dashboards and alerts for all critical metrics.

The Instrumental CloudWatch integration currently collects metrics from the following services:

Don't see the service you're looking for? Let us know!

Cost

CloudWatch metrics cost the same as every other metric in Instrumental, which depends on your pricing plan. In addition, enabling this integration may increase your AWS spend. We'll be making API requests on your behalf to fetch metrics, and Amazon will charge you for that usage. This cost is dependent on the number of resources (e.g. EC2 instances) you have and how they change over time.

We provide an estimate of your costs when you first turn on the integration, as well as when changing the integration settings. It's important to note that the estimate is based on a spot-in-time count of your CloudWatch usage, so it may be inaccurate. If your AWS usage changes significantly over time (e.g, via AutoScaling), this estimate may be very inaccurate.

You can control costs by:

Disabling regions, services or metrics. Disabled metrics are never fetched, and thus don't incur costs.
Customizing the fetch interval on metrics or services.

If you have questions about the potential cost of this integration, get in touch.

Security

Instrumental connects to your AWS account using IAM Role Delegation. Instead of giving us secret keys, you'll create a new IAM Role specific to our account, which we'll use when fetching your metrics. That means you have full control over access, and we don't hold any of your secret keys. See below for how to get set up.

The AWS CloudWatch integration only works with regions that support IAM Role Delegation, which excludes the China and GovCloud regions.

Latency

Instrumental fetches your CloudWatch based on your fetch interval settings, as often as every minute at the fastest setting. Because these metrics still need to go through our normal metric ingestion process, CloudWatch metrics can be delayed by up to a minute more than the fetch interval, compared to your other metrics.

Setup Instructions

To enable this integration, you'll need to create a new IAM Role, create a new IAM Policy, and configure the integration in your Instrumental project settings.

1. Create an IAM Policy

This IAM Policy defines what resources we'll be allowed to access within your AWS infrastructure. It will be attached to the Role we create next.

Visit the IAM Policy Dashboard and click the Create Policy button

Select the JSON tab and fill the text box with this:

{
    "Version": "2012-10-17",
    "Statement": [{
        "Action": [
            "cloudwatch:List*",
            "cloudwatch:Get*",
            "application-autoscaling:Describe*",
            "autoscaling:Describe*",
            "apigateway:Get*",
            "cloudfront:List*",
            "cloudsearch:Describe*",
            "dynamodb:List*",
            "ec2:Describe*",
            "elasticfilesystem:Describe*",
            "elasticache:Describe*",
            "elasticloadbalancing:Describe*",
            "kinesis:Describe*",
            "kinesis:List*",
            "kinesisanalytics:List*",
            "kinesisvideo:List*",
            "lambda:List*",
            "rds:Describe*",
            "redshift:Describe*",
            "route53:List*",
            "sns:List*",
            "sqs:List*"
        ],
      "Effect": "Allow",
      "Resource": "*"
    }]
}

This policy gives us read-only access to your CloudWatch data, plus access to read-only endpoints describing your AWS resources. We can then limit your costs by only collecting metrics for active resources.

Click Review Policy
Name the Policy InstrumentalAWSIntegrationPolicy
Click Create Policy

2. Create an IAM Role

This Role will be specific to Instrumental's AWS account, and limited by the Policy you just created.

Visit the IAM Role Dashboard and click the Create Role button
Select the tab Another AWS Account
Set Account ID to 162000585895
Check Require External ID and use your project's ID as the External ID
On the next page, select the Policy from the last step (InstrumentalAWSIntegrationPolicy)
On the next step, name the role InstrumentalAWSIntegrationRole
Click Create Role

3. Configure The Integration

Now that the role is created, it's time to configure your Instrumental project.

Go to the project settings page for your Instrumental project
Click the AWS CloudWatch tab
Use your AWS account id to begin the setup process

Once we can connect to AWS, we'll scan your AWS account for all available CloudWatch metrics, and preconfigure the integration. You'll then have the option to configure specific metrics and services, as well as the fetch interval.

Troubleshooting

Did you name the policy correctly?
We need the correct IAM Policy name so we can access your metrics. Check to make sure it's correct (InstrumentalAWSIntegrationPolicy). If you want to use a custom name, you can update the settings of your Integration to reference the new name. We recommend using the default name during the setup process, then changing it later. Let us know if you need help!
Did you set the External ID correctly?
Make sure it's set to your project's ID. The external ID is an important security measure that prevents other Instrumental customers from surreptitiously creating an Integration that accesses your metrics.
Did you incorrectly set the MFA setting?
We are unable to access your metrics if the IAM Role requires MFA.

Metric Naming Convention

Every metric collected by Instrumental's CloudWatch Integration follows this pattern:

{service_namespace}.{region}.{dimension_name}-{dimension_value}.{metric_name}

Service Namespace for each service can be found below.
Region is the AWS Region name, with hyphens converted to underscores, e.g us_east_1.
Dimensions are a CloudWatch term that describes key/value pairs that scope a datapoint to more or less specific contexts. For example, EC2 supports the dimension InstanceId, which scopes a given metric to a specific instance.
Metric Names can be found in the tables below.

Here's an example EC2 metric to demonstrate:

aws_ec2.us_east_1.instance_id-i-000.cpu_utilization

CloudWatch supports multiple concurrent dimensions for some services. In that case, we include all the dimensions. Here's an example Billing & Cost Management metric:

aws_billing.us_east_1.service_name-AWSLambda.currency-USD.requests

Dimensions are included in the metric name in the order they appear in the tables below. Every metric collected will include ONLY 1 set of dimensions. See below for which dimensions are supported for each service.

Some dimension values includes characters that are not allowed in Instrumental metric names (e.g., :). These are converted to _. Check out the metrics documentation for more info on metric name restrictions.

CloudWatch Statistics

CloudWatch supports many statistics, or different mathematical approaches for aggregrating data: Minimum, Maximum, Average, Sum, and SampleCount. However, not every metric supports every statistic, and not every statistic is sensical for every metric. We've chosen the most useful statistic for each metric, and only import that.

Fetch Interval Explained

When you set up the CloudWatch integration, you may notice that you can change the fetch interval for all metrics, for a specific service, or for a specific metric. What does that mean, and why can you change it?

Instrumental must poll the CloudWatch API to get your metric data, and the fetch interval defines how often we make these requests. The more often we fetch the data, the less apparent latency for your CloudWatch metrics in Instrumental. For example, setting the fetch interval to 1 minute means your data will be available in Instrumental after 1-2 minutes; a 10 minute fetch interval means data on graphs after 1-10 minutes.

So why not fetch everything at the lowest fetch interval? The CloudWatch API endpoints we use are metered, so each API requests incurs a small fee on your AWS bill. A smaller fetch interval means a bigger AWS bill, and larger fetch interval means a lower AWS bill. That's why we give you the option to set this value per metric and per service. You may want a few critical ELB metrics at the fastest possible rate and don't mind spending a bit more for that lower latency. Conversely, you may only want CloudFront metrics for historical analysis and thus don't mind a bigger delay.

Supported Services & Metrics

Instrumental's CloudWatch integration currently supports the following services:

API Gateway

All API Gateway metrics begin with the namespace aws_api_gateway and will include one of the following dimension sets:

CloudWatch Dimensions	Metric Name Translation
ApiName	`api_name-{value}`
ApiName, Stage	`api_name-{value}.stage-{value}`
ApiName, Method, Resource, Stage	`api_name-{value}.method-{value}.resource-{value}.stage-{value}`

Supported Metrics

Metric Name	Statistic	Description
latency	Average	Time between when API Gateway receives a request from a client and when it returns a response to the client Named `Latency` in CloudWatch
integration_latency	Average	Time between when API Gateway relays a request to the back end and when it receives a response from the back end Named `IntegrationLatency` in CloudWatch
count	SampleCount	Total number API requests in a given period Named `Count` in CloudWatch
cache_miss_count	Sum	Number of requests served from the back end when API caching is enabled Named `CacheMissCount` in CloudWatch
cache_hit_count	Sum	Number of requests served from the API cache Named `CacheHitCount` in CloudWatch
5xx_error	Sum	Number of server-side errors Named `5XXError` in CloudWatch
4xx_error	Sum	Number of client-side errors Named `4XXError` in CloudWatch

Application ELB

All Application ELB metrics begin with the namespace aws_elb_application and will include one of the following dimension sets:

CloudWatch Dimensions	Metric Name Translation
LoadBalancer	`load_balancer-{value}`
LoadBalancer, AvailabilityZone	`load_balancer-{value}.availability_zone-{value}`
LoadBalancer, AvailabilityZone, TargetGroup	`load_balancer-{value}.availability_zone-{value}.target_group-{value}`

Supported Metrics

Metric Name	Statistic	Description
active_connections	Sum	The total number of concurrent TCP connections active from clients to the load balancer and from the load balancer to targets. Named `ActiveConnectionCount` in CloudWatch
client_tls_negotiation_errors	Sum	The number of TLS connections initiated by the client that did not establish a session with the load balancer due to a TLS error. Possible causes include a mismatch of ciphers or protocols or the client failing to verify the server certificate and closing the connection. Named `ClientTLSNegotiationErrorCount` in CloudWatch
consumed_lcus	Maximum	The number of load balancer capacity units (LCU) used by your load balancer. You pay for the number of LCUs that you use per hour. Named `ConsumedLCUs` in CloudWatch
desync_mitigation_mode_noncompliant_requests	Sum	The number of requests that do not comply with RFC 7230. Named `DesyncMitigationMode_NonCompliant_Request_Count` in CloudWatch
dropped_invalid_header_requests	Sum	The number of requests where the load balancer removed HTTP headers with header fields that are not valid before routing the request. Named `DroppedInvalidHeaderRequestCount` in CloudWatch
forwarded_invalid_header_requests	Sum	The number of requests routed by the load balancer that had HTTP headers with header fields that are not valid. Named `ForwardedInvalidHeaderRequestCount` in CloudWatch
grpc_requests	Sum	The number of gRPC requests processed over IPv4 and IPv6. Named `GrpcRequestCount` in CloudWatch
http_fixed_responses	Sum	The number of fixed-response actions that were successful. Named `HTTP_Fixed_Response_Count` in CloudWatch
http_redirects	Sum	The number of redirect actions that were successful. Named `HTTP_Redirect_Count` in CloudWatch
http_redirect_url_limit_exceeded	Sum	The number of redirect actions that couldn't be completed because the URL in the response location header is larger than 8K. Named `HTTP_Redirect_Url_Limit_Exceeded_Count` in CloudWatch
httpcode_elb_3xx	Sum	The number of HTTP 3XX redirection codes that originate from the load balancer. Named `HTTPCode_ELB_3XX_Count` in CloudWatch
httpcode_elb_4xx	Sum	The number of HTTP 4XX client error codes that originate from the load balancer. Client errors are generated when requests are malformed or incomplete. These requests were not received by the target, other than in the case where the load balancer returns an HTTP 460 error code. This count does not include any response codes generated by the targets. Named `HTTPCode_ELB_4XX_Count` in CloudWatch
httpcode_elb_5xx	Sum	The number of HTTP 5XX server error codes that originate from the load balancer. This count does not include any response codes generated by the targets. Named `HTTPCode_ELB_5XX_Count` in CloudWatch
httpcode_elb_500	Sum	The number of HTTP 500 error codes that originate from the load balancer. Named `HTTPCode_ELB_500_Count` in CloudWatch
httpcode_elb_502	Sum	The number of HTTP 502 error codes that originate from the load balancer. Named `HTTPCode_ELB_502_Count` in CloudWatch
httpcode_elb_503	Sum	The number of HTTP 503 error codes that originate from the load balancer. Named `HTTPCode_ELB_503_Count` in CloudWatch
httpcode_elb_504	Sum	The number of HTTP 504 error codes that originate from the load balancer. Named `HTTPCode_ELB_504_Count` in CloudWatch
ipv6_processed_bytes	Sum	The total number of bytes processed by the load balancer over IPv6. This count is included in ProcessedBytes. Named `IPv6ProcessedBytes` in CloudWatch
ipv6_requests	Sum	The number of IPv6 requests received by the load balancer. Named `IPv6RequestCount` in CloudWatch
new_connections	Sum	The total number of new TCP connections established from clients to the load balancer and from the load balancer to targets. Named `NewConnectionCount` in CloudWatch
nonsticky_requests	Sum	The number of requests where the load balancer chose a new target because it couldn't use an existing sticky session. For example, the request was the first request from a new client and no stickiness cookie was presented, a stickiness cookie was presented but it did not specify a target that was registered with this target group, the stickiness cookie was malformed or expired, or an internal error prevented the load balancer from reading the stickiness cookie. Named `NonStickyRequestCount` in CloudWatch
processed_bytes	Sum	The total number of bytes processed by the load balancer over IPv4 and IPv6. This count includes traffic to and from clients and Lambda functions, and traffic from an Identity Provider (IdP) if user authentication is enabled. Named `ProcessedBytes` in CloudWatch
rejected_connections	Sum	The number of connections that were rejected because the load balancer had reached its maximum number of connections. Named `RejectedConnectionCount` in CloudWatch
requests	Sum	The number of requests processed over IPv4 and IPv6. Named `RequestCount` in CloudWatch

Classic ELB

All Classic ELB metrics begin with the namespace aws_elb and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
LoadBalancerName, AvailabilityZone	`load_balancer_name-{value}.availability_zone-{value}`

Supported Metrics

Metric Name	Statistic	Description
latency	Average	Average of the total time elapsed, in seconds, from the time the load balancer sent the request to a registered instance until the instance started to send the response headers Named `Latency` in CloudWatch
unhealthy_hosts	Average	Average number of unhealthy instances registered with your load balancer Named `UnHealthyHostCount` in CloudWatch
healthy_hosts	Average	Average number of healthy instances registered with your load balancer Named `HealthyHostCount` in CloudWatch
surge_queue_length	Maximum	Total number of requests (HTTP listener) or connections (TCP listener) that are pending routing to a healthy instance. The maximum size of the queue is 1,024. Named `SurgeQueueLength` in CloudWatch
spillover_count	Sum	Total number of requests that were rejected because the surge queue is full Named `SpilloverCount` in CloudWatch
request_count	Sum	Number of requests completed or connections made Named `RequestCount` in CloudWatch
http_code_elb_5xx	Sum	Number of HTTP 5XX error codes generated by the load balancer Named `HTTPCode_ELB_5XX` in CloudWatch
http_code_elb_4xx	Sum	Number of HTTP 4XX error codes generated by the load balancer Named `HTTPCode_ELB_4XX` in CloudWatch
http_code_backend_5xx	Sum	Number of 5XX HTTP response codes generated by registered instances Named `HTTPCode_Backend_5XX` in CloudWatch
http_code_backend_4xx	Sum	Number of 4XX HTTP response codes generated by registered instances Named `HTTPCode_Backend_4XX` in CloudWatch
http_code_backend_3xx	Sum	Number of 3XX HTTP response codes generated by registered instances Named `HTTPCode_Backend_3XX` in CloudWatch
http_code_backend_2xx	Sum	Number of 2XX HTTP response codes generated by registered instances Named `HTTPCode_Backend_2XX` in CloudWatch
backend_connection_errors	Sum	Number of connections that were not successfully established between the load balancer and the registered instances Named `BackendConnectionErrors` in CloudWatch

CloudFront

All CloudFront metrics begin with the namespace aws_cloudfront and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
Region, DistributionId	`region-{value}.distribution_id-{value}`

Supported Metrics

Metric Name	Statistic	Description
5xx_error_rate	Average	Percentage requests for which the HTTP status code is 5xx Named `5xxErrorRate` in CloudWatch
4xx_error_rate	Average	Percentage requests for which the HTTP status code is 4xx Named `4xxErrorRate` in CloudWatch
total_error_rate	Average	Percentage requests for which the HTTP status code is 4xx or 5xx Named `TotalErrorRate` in CloudWatch
bytes_uploaded	Sum	Number of bytes uploaded to your origin with CloudFront using POST and PUT requests Named `BytesUploaded` in CloudWatch
bytes_downloaded	Sum	Number of bytes downloaded by viewers for GET, HEAD, and OPTIONS requests Named `BytesDownloaded` in CloudWatch
requests	Sum	Number of requests for all HTTP methods and for both HTTP and HTTPS requests Named `Requests` in CloudWatch

EBS

All EBS metrics begin with the namespace aws_ebs and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
VolumeId	`volume_id-{value}`

Supported Metrics

Metric Name	Statistic	Description
burst_balance	Average	Percentage of I/O credits (for gp2) or throughput credits (for st1 and sc1) remaining in the burst bucket. For gp2, st1, and sc1 volumes only. Named `BurstBalance` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_consumed_read_write_ops	Sum	Total amount of read and write operations (normalized to 256K capacity units). Provisioned IOPS SSD volumes only. Named `VolumeConsumedReadWriteOps` in CloudWatch
volume_throughput_percentage	Average	Percentage of I/O operations per second (IOPS) delivered of the total IOPS provisioned for an Amazon EBS volume. Provisioned IOPS SSD volumes only. Named `VolumeThroughputPercentage` in CloudWatch
volume_queue_length	Average	Number of read and write operation requests waiting to be completed Named `VolumeQueueLength` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_idle_time	Sum	Total number of seconds in a specified period when no read or write operations were submitted Named `VolumeIdleTime` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_total_write_time	Sum	Total number of seconds spent by all write operations that completed in a specified period Named `VolumeTotalWriteTime` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_total_read_time	Sum	Total number of seconds spent by all read operations that completed in a specified period Named `VolumeTotalReadTime` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_write_ops	Sum	Total number of write operations on the volume Named `VolumeWriteOps` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_read_ops	Sum	Total number of read operations on the volume Named `VolumeReadOps` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_write_bytes	Sum	Total number of bytes written to the volume Named `VolumeWriteBytes` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_read_bytes	Sum	Total number of bytes read from the volume Named `VolumeReadBytes` in CloudWatch Only available from CloudWatch at 5 minute resolution

EC2

AWS Elastic Cloud Compute. Detailed monitoring MUST be enabled on your instances for EC2 metrics to work as expected.

All EC2 metrics begin with the namespace aws_ec2 and will include one of the following dimension sets:

CloudWatch Dimensions	Metric Name Translation
InstanceId	`instance_id-{value}`
AutoScalingGroupName	`auto_scaling_group_name-{value}`

Supported Metrics

Metric Name	Statistic	Description
status_check_failed_system	Sum	Whether the instance has passed the system status check Named `StatusCheckFailed_System` in CloudWatch
status_check_failed_instance	Sum	Whether the instance has passed the instance status check Named `StatusCheckFailed_Instance` in CloudWatch
status_check_failed	Sum	Whether the instance has passed both the instance status check and the system status check Named `StatusCheckFailed` in CloudWatch
network_out	Sum	Bytes sent out on all network interfaces by the instance Named `NetworkOut` in CloudWatch
network_in	Sum	Bytes received on all network interfaces by the instance Named `NetworkIn` in CloudWatch
disk_write_bytes	Sum	Bytes written to all instance store volumes available to the instance Named `DiskWriteBytes` in CloudWatch
disk_read_bytes	Sum	Bytes read from all instance store volumes available to the instance Named `DiskReadBytes` in CloudWatch
disk_write_ops	Sum	Completed write operations from all instance store volumes available to the instance Named `DiskWriteOps` in CloudWatch
disk_read_ops	Sum	Completed read operations from all instance store volumes available to the instance Named `DiskReadOps` in CloudWatch
cpu_utilization	Average	Percentage of EC2 compute units currently in use Named `CPUUtilization` in CloudWatch

EC2 Auto Scaling

Note: EC2 Auto Scaling Group metrics must be enabled in the AWS Console.

All EC2 Auto Scaling metrics begin with the namespace aws_auto_scaling and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
AutoScalingGroupName	`auto_scaling_group_name-{value}`

Supported Metrics

Metric Name	Statistic	Description
group_total_instances	Average	Total number of instances in the Auto Scaling group (instances that are in service, pending, and terminating) Named `GroupTotalInstances` in CloudWatch
group_terminating_instances	Average	Number of instances that are in the process of terminating Named `GroupTerminatingInstances` in CloudWatch
group_standby_instances	Average	Number of instances that are in a Standby state (instances still running but not actively in service) Named `GroupStandbyInstances` in CloudWatch
group_pending_instances	Average	Number of instances that are pending (instances not yet in service) Named `GroupPendingInstances` in CloudWatch
group_in_service_instances	Average	Number of instances that are running. Does not include instances that are pending or terminating. Named `GroupInServiceInstances` in CloudWatch
group_desired_capacity	Average	Number of instances that the Auto Scaling group attempts to maintain Named `GroupDesiredCapacity` in CloudWatch
group_max_size	Average	Maximum size of the Auto Scaling group Named `GroupMaxSize` in CloudWatch
group_min_size	Average	Minimum size of the Auto Scaling group Named `GroupMinSize` in CloudWatch

EFS

All EFS metrics begin with the namespace aws_efs and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
FileSystemId	`file_system_id-{value}`

Supported Metrics

Metric Name	Statistic	Description
total_io_bytes	Sum	Total number of bytes for file system operations, including data read, data write, and metadata operations Named `TotalIOBytes` in CloudWatch
permitted_throughput	Average	Maximum amount of throughput a file system is allowed. Named `PermittedThroughput` in CloudWatch
percent_io_limit	Average	Shows how close a file system is to reaching the I/O limit of the General Purpose performance mode Named `PercentIOLimit` in CloudWatch
metadata_io_bytes	Sum	Total number of bytes associated with metadata operations Named `MetadataIOBytes` in CloudWatch
data_write_io_bytes	Sum	Total number of bytes associated with write operations Named `DataWriteIOBytes` in CloudWatch
data_read_io_bytes	Sum	Total number of bytes associated with read operations Named `DataReadIOBytes` in CloudWatch
client_connections	Sum	Number of client connections to a file system Named `ClientConnections` in CloudWatch
burst_credit_balance	Average	Balance of burst credits that a file system has Named `BurstCreditBalance` in CloudWatch

ElastiCache

All ElastiCache metrics begin with the namespace aws_elasticache and will include the following dimension set:

CloudWatch Dimensions	Metric Name Translation
CacheClusterId, CacheNodeId	`cache_cluster_id-{value}.cache_node_id-{value}`

Supported Metrics

Metric Name	Statistic	Description
stream_based_cmds	Average	Average number of commands that are stream-based per minute Named `StreamBasedCmds` in CloudWatch
string_based_cmds	Average	Average number of commands that are string-based per minute Named `StringBasedCmds` in CloudWatch
sorted_set_based_cmds	Average	Average number of commands that are sorted set-based per minute Named `SortedSetBasedCmds` in CloudWatch
set_type_cmds	Average	Average number of write types of commands per minute Named `SetTypeCmds` in CloudWatch
set_based_cmds	Average	Average number of commands that are set-based per minute Named `SetBasedCmds` in CloudWatch
list_based_cmds	Average	Average number of commands that are list-based per minute Named `ListBasedCmds` in CloudWatch
key_based_cmds	Average	Average number of commands that are key-based per minute Named `KeyBasedCmds` in CloudWatch
hyper_log_log_based_cmds	Average	Average number of HyperLogLog-based commands per minute Named `HyperLogLogBasedCmds` in CloudWatch
hash_based_cmds	Average	Average number of commands that are hash-based per minute Named `HashBasedCmds` in CloudWatch
get_type_cmds	Average	Average number of read-only type commands per minute Named `GetTypeCmds` in CloudWatch
replication_lag	Average	How far behind, in seconds, the replica is in applying changes from the primary node. Only applicable for a node running as a read replica. Named `ReplicationLag` in CloudWatch
replication_bytes	Sum	The number of bytes that the primary is sending to all of its replicas. Representative of the write load on the replication group. Named `ReplicationBytes` in CloudWatch
new_connections	Sum	Total number of connections that have been accepted by the server during this period Named `NewConnections` in CloudWatch
engine_cpu_utilization	Average	CPU utilization of the Redis process. Provides more precise visibility into the load of the Redis process itself. Named `EngineCPUUtilization` in CloudWatch
cache_misses	Sum	Number of unsuccessful read-only key lookups in the main dictionary Named `CacheMisses` in CloudWatch
cache_hits	Sum	Number of successful read-only key lookups in the main dictionary Named `CacheHits` in CloudWatch
bytes_used_for_cache	Average	Total number of bytes allocated by Redis for all purposes, including the dataset, buffers, etc Named `BytesUsedForCache` in CloudWatch
active_defrag_hits	Sum	Number of value reallocations performed by the active defragmentation process Named `ActiveDefragHits` in CloudWatch
touch_misses	Average	Number of items that have been touched, but were not found. Specific to Memcached 1.4.14. Named `TouchMisses` in CloudWatch
touch_hits	Average	Number of keys that have been touched and were given a new expiration time. Specific to Memcached 1.4.14. Named `TouchHits` in CloudWatch
slabs_moved	Average	Total number of slab pages that have been moved. Specific to Memcached 1.4.14. Named `SlabsMoved` in CloudWatch
expired_unfetched	Average	Number of expired items reclaimed from the LRU which were never touched after being set. Specific to Memcached 1.4.14. Named `ExpiredUnfetched` in CloudWatch
evicted_unfetched	Average	Number of valid items evicted from the least recently used cache (LRU) which were never touched after being set. Specific to Memcached 1.4.14. Named `EvictedUnfetched` in CloudWatch
curr_config	Average	Current number of configurations stored. Specific to Memcached 1.4.14. Named `CurrConfig` in CloudWatch
cmd_touch	Average	Cumulative number of touch requests. Specific to Memcached 1.4.14. Named `CmdTouch` in CloudWatch
cmd_config_set	Average	Cumulative number of config set requests. Specific to Memcached 1.4.14. Named `CmdConfigSet` in CloudWatch
cmd_config_get	Average	Cumulative number of config get requests. Specific to Memcached 1.4.14. Named `CmdConfigGet` in CloudWatch
bytes_used_for_hash	Average	Number of bytes currently used by hash tables. Specific to Memcached 1.4.14. Named `BytesUsedForHash` in CloudWatch
reclaimed	Average	Number of expired items the cache evicted to allow space for new writes. Named `Reclaimed` in CloudWatch
incr_misses	Average	Number of increment requests the cache has received where the key requested was not found. Specific to Memcached. Named `IncrMisses` in CloudWatch
incr_hits	Average	number of increment requests the cache has received where the key requested was found. Specific to Memcached. Named `IncrHits` in CloudWatch
get_misses	Average	Number of get requests the cache has received where the key requested was not found. Specific to Memcached. Named `GetMisses` in CloudWatch
get_hits	Average	Number of get requests the cache has received where the key requested was found. Specific to Memcached. Named `GetHits` in CloudWatch
evictions	Average	Number of non-expired items the cache evicted to allow space for new writes. Named `Evictions` in CloudWatch
delete_misses	Average	Number of delete requests the cache has received where the requested key was not found. Specific to Memcached. Named `DeleteMisses` in CloudWatch
delete_hits	Average	Number of delete requests the cache has received where the requested key was found. Specific to Memcached. Named `DeleteHits` in CloudWatch
decr_misses	Average	Number of decrement requests the cache has received where the requested key was not found. Specific to Memcached. Named `DecrMisses` in CloudWatch
decr_hits	Average	Number of decrement requests the cache has received where the requested key was found. Specific to Memcached. Named `DecrHits` in CloudWatch
curr_items	Average	Number of items in the cache Named `CurrItems` in CloudWatch
curr_connections	Average	Number of connections connected to the cache. ElastiCache uses two to three of the connections to monitor the cluster in each case. Named `CurrConnections` in CloudWatch
cmd_get	Average	The number of get commands the cache has received. Specific to Memcached. Named `CmdGet` in CloudWatch
cmd_set	Average	The number of set commands the cache has received. Specific to Memcached. Named `CmdSet` in CloudWatch
cmd_flush	Average	Number of flush commands the cache has received. Specific to Memcached. Named `CmdFlush` in CloudWatch
cas_misses	Average	Number of CAS requests the cache has received where the key requested was not found. Specific to Memcached. Named `CasMisses` in CloudWatch
cas_hits	Average	Number of CAS requests the cache has received where the requested key was found and the CAS value matched. Specific to Memcached. Named `CasHits` in CloudWatch
cas_badval	Average	Number of CAS (check and set) requests the cache has received where the CAS value did not match the CAS value stored. Specific to Memcached. Named `CasBadval` in CloudWatch
bytes_written	Average	Number of bytes that have been written to the network by the cache node. Specific to Memcached. Named `BytesWrittenOutFromMemcached` in CloudWatch
bytes_cached	Average	Number of bytes used to store cache items. Specific to Memcached. Named `BytesUsedForCacheItems` in CloudWatch
bytes_read	Average	Number of bytes that have been read from the network by the cache node. Specific to Memcached. Named `BytesReadIntoMemcached` in CloudWatch
swap_usage	Average	Amount of swap used on the host Named `SwapUsage` in CloudWatch
network_bytes_out	Average	Average number of bytes per minute the host has written to the network Named `NetworkBytesOut` in CloudWatch
network_bytes_in	Average	Average number of bytes per minute the host has read from the network Named `NetworkBytesIn` in CloudWatch
freeable_memory	Average	Amount of free memory available on the host Named `FreeableMemory` in CloudWatch
cpu_utilization	Average	Percentage of CPU utilization Named `CPUUtilization` in CloudWatch
network_bandwidth_in_allowance_exceeded	Average	The number of packets shaped because the inbound aggregate bandwidth exceeded the maximum for the instance Named `NetworkBandwidthInAllowanceExceeded` in CloudWatch
network_bandwidth_out_allowance_exceeded	Average	The number of packets shaped because the outbound aggregate bandwidth exceeded the maximum for the instance Named `NetworkBandwidthOutAllowanceExceeded` in CloudWatch
network_conntrack_allowance_exceeded	Average	The number of packets shaped because connection tracking exceeded the maximum for the instance and new connections could not be established. This can result in packet loss for traffic to or from the instance Named `NetworkConntrackAllowanceExceeded` in CloudWatch
network_link_local_allowance_exceeded	Average	The number of packets shaped because the PPS of the traffic to local proxy services exceeded the maximum for the network interface Named `NetworkLinkLocalAllowanceExceeded` in CloudWatch
network_packets_per_second_allowance_exceeded	Average	The number of packets shaped because the bidirectional packets per second exceeded the maximum for the instance. Named `NetworkPacketsPerSecondAllowanceExceeded` in CloudWatch

Lambda

All Lambda metrics begin with the namespace aws_lambda and will include one of the following dimension sets:

CloudWatch Dimensions	Metric Name Translation
No Dimensions
FunctionName, Resource	`function_name-{value}.resource-{value}`

Supported Metrics

Metric Name	Statistic	Description
duration	Average	Measures the average elapsed wall clock time from when the function code starts executing as a result of an invocation to when it stops executing Named `Duration` in CloudWatch
unreserved_concurrent_executions	Average	Emitted as an aggregate metric for all functions in the account only. Represents the sum of the concurrency of the functions that do not have a custom concurrency limit specified. Named `UnreservedConcurrentExecutions` in CloudWatch
concurrent_executions	Average	Emitted as an aggregate metric for all functions in the account, and for functions that have a custom concurrency limit specified Named `ConcurrentExecutions` in CloudWatch
iterator_age	Average	Measures the age of the last record for each batch of records processed Named `IteratorAge` in CloudWatch
throttles	Sum	Measures the number of Lambda function invocation attempts that were throttled due to invocation rates exceeding the concurrent limits (error code 429) Named `Throttles` in CloudWatch
dead_letter_errors	Sum	Incremented when Lambda is unable to write the failed event payload to your configured Dead Letter Queues Named `DeadLetterErrors` in CloudWatch
errors	Sum	Number of invocations that failed due to errors in the function (response code 4XX) Named `Errors` in CloudWatch
invocations	Sum	Number of times a function is invoked in response to an event or invocation API call. Includes successful and failed invocations, but does not include throttled attempts. Equals the billed requests for the function. Named `Invocations` in CloudWatch

RDS

All RDS metrics begin with the namespace aws_rds and will include one of the following dimension sets:

CloudWatch Dimensions	Metric Name Translation
DBInstanceIdentifier	`db_instance_identifier-{value}`
DBClusterIdentifier	`db_cluster_identifier-{value}`

Supported Metrics

Metric Name	Statistic	Description
volume_write_iops	Sum	Number of billed write I/O operations from a cluster volume. Applies to Aurora MySQL and Aurora PostgreSQL. Named `VolumeWriteIOPs` in CloudWatch Only available from CloudWatch at 5 minute resolution
volume_read_iops	Sum	Number of billed read I/O operations from a cluster volume. Applies to Aurora MySQL and Aurora PostgreSQL. Named `VolumeReadIOPs` in CloudWatch Only available from CloudWatch at 5 minute resolution
burst_balance	Average	Percent of General Purpose SSD (gp2) burst-bucket I/O credits available Named `BurstBalance` in CloudWatch
write_throughput	Average	Average number of bytes written to disk per second. Does not apply to Aurora MySQL. Named `WriteThroughput` in CloudWatch
write_latency	Average	Average amount of time taken per disk I/O operation. Does not apply to Aurora MySQL. Named `WriteLatency` in CloudWatch
write_iops	Average	Average number of disk write I/O operations per second. Does not apply to Aurora MySQL. Named `WriteIOPS` in CloudWatch
transaction_logs_generation	Average	Size of transaction logs generated per second Named `TransactionLogsGeneration` in CloudWatch
transaction_logs_disk_usage	Average	Disk space used by transaction logs. Only applies to PostgreSQL, including Aurora. Named `TransactionLogsDiskUsage` in CloudWatch
swap_usage	Average	Amount of swap space used on the DB instance. Does not apply to Aurora MySQL. Named `SwapUsage` in CloudWatch
replica_lag	Average	Amount of time a Read Replica DB instance lags behind the source DB instance Named `ReplicaLag` in CloudWatch
read_throughput	Average	Average number of bytes read from disk per second. Does not apply to Aurora MySQL. Named `ReadThroughput` in CloudWatch
read_latency	Average	Average amount of time taken per disk I/O operation. Does not apply to Aurora MySQL. Named `ReadLatency` in CloudWatch
read_iops	Average	Average number of disk read I/O operations per second. Does not apply to Aurora MySQL. Named `ReadIOPS` in CloudWatch
network_transmit_throughput	Average	Outgoing (Transmit) network traffic on the DB instance, including both customer database traffic and Amazon RDS traffic used for monitoring and replication in Bytes/second Named `NetworkTransmitThroughput` in CloudWatch
network_receive_throughput	Average	Incoming (Receive) network traffic on the DB instance, including both customer database traffic and Amazon RDS traffic used for monitoring and replication in Bytes/second Named `NetworkReceiveThroughput` in CloudWatch
maximum_used_transaction_ids	Maximum	Maximum transaction ID that has been used. Applies only to PostgreSQL, including Aurora. Named `MaximumUsedTransactionIDs` in CloudWatch
free_storage_space	Average	Amount of available storage space Named `FreeStorageSpace` in CloudWatch
freeable_memory	Average	amount of available random access memory Named `FreeableMemory` in CloudWatch
disk_queue_depth	Average	Number of outstanding IOs (read/write requests) waiting to access the disk. Does not apply to Aurora MySQL. Named `DiskQueueDepth` in CloudWatch
database_connections	Average	Number of database connections in use Named `DatabaseConnections` in CloudWatch
cpu_utilization	Average	Percentage of CPU utilization Named `CPUUtilization` in CloudWatch
bin_log_disk_usage	Average	Amount of disk space occupied by binary logs, in bytes. Applies only to MySQL, including Aurora. Named `BinLogDiskUsage` in CloudWatch