Amazon CloudWatch metrics for Amazon EFS - Amazon Elastic File System
Bytes reported in CloudWatch

Amazon CloudWatch metrics for Amazon EFS

Amazon EFS metrics use the EFS namespace and provide metrics for a single dimension, FileSystemId. A file system's ID can be found in the Amazon EFS console, and it takes the form of fs-abcdef0123456789a.

The AWS/EFS namespace includes the following metrics.

TimeSinceLastSync

Shows the amount of time that has passed since the last successful sync to the destination file system in a replication configuration. Any changes to data on the source file system that occurred before the TimeSinceLastSync value have been successfully replicated. Any changes on the source that occurred after TimeSinceLastSync might not be fully replicated.

Units: Seconds

Valid statistics: Minimum, Maximum, Average

PercentIOLimit

Shows how close a file system is to reaching the I/O limit of the General Purpose performance mode.

Units: Percent

Valid statistics: Minimum, Maximum, Average

BurstCreditBalance

The number of burst credits that a file system has. Burst credits allow a file system to burst to throughput levels above a file system’s baseline level for periods of time.

The Minimum statistic is the smallest burst credit balance for any minute during the period. The Maximum statistic is the largest burst credit balance for any minute during the period. The Average statistic is the average burst credit balance during the period.

Units: Bytes

Valid statistics: Minimum, Maximum, Average

PermittedThroughput

The maximum amount of throughput that a file system can drive.

  • For file systems using Elastic throughput, this value reflects the maximum write throughput of the file system.

  • For file systems using Provisioned throughput, if the amount of data stored in the EFS Standard storage class allows your file system to drive a higher throughput than you provisioned, this metric reflects the higher throughput instead of the provisioned amount.

  • For file systems in Bursting throughput mode, this value is a function of the file system size and BurstCreditBalance.

The Minimum statistic is the smallest throughput permitted for any minute during the period. The Maximum statistic is the highest throughput permitted for any minute during the period. The Average statistic is the average throughput permitted during the period.

Note

Read operations are metered at one-third the rate of other operations.

Units: Bytes per second

Valid statistics: Minimum, Maximum, Average

MeteredIOBytes

The number of metered bytes for each file system operation, including data read, data write, and metadata operations, with read operations metered at one-third the rate of other operations.

You can create a CloudWatch metric math expression that compares MeteredIOBytes to PermittedThroughput. If these values are equal, then you are consuming the entire amount of throughput allocated to your file system. In this situation, you might consider changing the file system's throughput mode to get higher throughput.

The Sum statistic is the total number of metered bytes associated with all file system operations. The Minimum statistic is the size of the smallest operation during the period. The Maximum statistic is the size of the largest operation during the period. The Average statistic is the average size of an operation during the period. The SampleCount statistic provides a count of all operations.

Units:

  • Bytes for Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

TotalIOBytes

The actual number of bytes for each file system operation, including data read, data write, and metadata operations. This is the actual amount that your application is driving, and not the throughput the file system is being metered at. It might be higher than the numbers shown in PermittedThroughput.

The Sum statistic is the total number of bytes associated with all file system operations. The Minimum statistic is the size of the smallest operation during the period. The Maximum statistic is the size of the largest operation during the period. The Average statistic is the average size of an operation during the period. The SampleCount statistic provides a count of all operations.

Note

To calculate the average operations per second for a period, divide the SampleCount statistic by the number of seconds in the period. To calculate the average throughput (bytes per second) for a period, divide the Sum statistic by the number of seconds in the period.

Units:

  • Bytes for Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

DataReadIOBytes

The actual number of bytes for each file system read operation.

The Sum statistic is the total number of bytes associated with read operations. The Minimum statistic is the size of the smallest read operation during the period. The Maximum statistic is the size of the largest read operation during the period. The Average statistic is the average size of read operations during the period. The SampleCount statistic provides a count of read operations.

Units:

  • Bytes for Minimum, Maximum, Average, and Sum.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

DataWriteIOBytes

The actual number of bytes for each file system write operation.

The Sum statistic is the total number of bytes associated with write operations. The Minimum statistic is the size of the smallest write operation during the period. The Maximum statistic is the size of the largest write operation during the period. The Average statistic is the average size of write operations during the period. The SampleCount statistic provides a count of write operations.

Units:

  • Bytes are the units for the Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

MetadataIOBytes

The actual number of bytes for each metadata operation.

The Sum statistic is the total number of bytes associated with metadata operations. The Minimum statistic is the size of the smallest metadata operation during the period. The Maximum statistic is the size of the largest metadata operation during the period. The Average statistic is the size of the average metadata operation during the period. The SampleCount statistic provides a count of metadata operations.

Units:

  • Bytes are the units for the Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

MetadataReadIOBytes

The actual number of bytes for each metadata read operation.

The Sum statistic is the total number of bytes associated with metadata read operations. The Minimum statistic is the size of the smallest metadata read operation during the period. The Maximum statistic is the size of the largest metadata read operation during the period. The Average statistic is the average size of metadata read operations during the period. The SampleCount statistic provides a count of metadata read operations.

Units:

  • Bytes are the units for the Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

MetadataWriteIOBytes

The actual number of bytes for each metadata write operation.

The Sum statistic is the total number of bytes associated with metadata write operations. The Minimum statistic is the size of the smallest metadata write operation during the period. The Maximum statistic is the size of the largest metadata write operation during the period. The Average statistic is the average size of metadata write operations during the period. The SampleCount statistic provides a count of metadata write operations.

Units:

  • Bytes are the units for the Minimum, Maximum, Average, and Sum statistics.

  • Count for SampleCount.

Valid statistics: Minimum, Maximum, Average, Sum, SampleCount

ClientConnections

The number of client connections to a file system. When using a standard client, there is one connection per mounted Amazon EC2 instance.

Note

To calculate the average ClientConnections for periods greater than one minute, divide the Sum statistic by the number of minutes in the period.

Units: Count of client connections

Valid statistics: Sum

StorageBytes

The size of the file system in bytes, including the amount of data stored in the EFS storage classes. This metric is emitted to CloudWatch every 15 minutes.

The StorageBytes metric has the following dimensions:

  • Total is the metered size (in bytes) of data stored in the file system, in all storage classes. For EFS Infrequent Access and EFS Archive storage classes, files smaller than 128KiB are rounded to 128KiB.

  • Standard is the metered size (in bytes) of data stored in the EFS Standard storage class.

  • IA is the metered size (in bytes) of data stored in the EFS Infrequent Access storage class.

  • IASizeOverhead is the difference (in bytes) between the actual size of data in the EFS Infrequent Access storage class (indicated in the IA dimension) and the amount by which it is rounded to 128KiB, if the size of the file is smaller than 128KiB.

  • Archive is the metered size (in bytes) of data in the EFS Archive storage class. This number reflects the actual size of data stored in Archive storage, before rounding small files to 128KiB.

  • ArchiveSizeOverhead is the difference (in bytes) between the actual size of data in the EFS Archive storage class (indicated in the Archive dimension) and the amount by which it is rounded to 128KiB, if the size of the file is smaller than 128KiB.

Units: Bytes

Valid statistics: Minimum, Maximum, Average

Note

StorageBytes is displayed on the Amazon EFS console File system metrics page using base 1024 units (kibibytes, mebibytes, gibibytes, and tebibytes).

Bytes reported in CloudWatch

Amazon EFS CloudWatch metrics are reported as raw bytes. Bytes are not rounded to either a decimal or binary multiple of the unit. Keep this in mind when calculating your burst rate using the data you get from the metrics. For more information about bursting, see Bursting throughput.