Managing throughput capacity - FSx for Lustre

Managing throughput capacity

Every FSx for Lustre file system has a throughput capacity that is configured when you create the file system. The throughput of an FSx for Lustre file system is measured in megabytes per second
 per tebibyte (MB/s/TiB). Throughput capacity is one factor that determines the speed at which the file server hosting the file system can serve file data. Higher levels of throughput capacity also come with higher levels of I/O operations per second (IOPS) and more memory for caching of data on the file server. For more information, see Amazon FSx for Lustre performance.

You can modify the throughput tier of a persistent SSD-based file system by increasing or decreasing the value of the file system's throughput per unit of storage. Valid values depend on the deployment type of the file system, as follows:

  • For Persistent_1 SSD-based deployment types, valid values are 50, 100, and 200 MB/s/TiB.

  • For Persistent_2 SSD-based deployment types, valid values are 125, 250, 500, and 1000 MB/s/TiB.

You can view the current value of the file system's throughput per unit of storage, as follows:

  • Using the console – On the Summary panel of the file system details page, the Throughput per unit of storage field shows the current value.

  • Using the CLI or API – Use the describe-file-systems CLI command or the DescribeFileSystems API operation, and look for the PerUnitStorageThroughput property.

When you modify your file system's throughput capacity, behind the scenes, Amazon FSx switches out the file system's file servers. Your file system will be unavailable for a few minutes during throughput capacity scaling. You are billed for the new amount of throughput capacity once it is available to your file system.

Considerations when updating throughput capacity

Here are a few important items to consider when updating throughput capacity:

  • Increase or decrease – You can increase or decrease the amount of throughput capacity for a file system.

  • Update increments – When you modify throughput capacity, use the increments listed in the Update throughput tier dialog box.

  • Time between increases – You can't make further throughput capacity changes on a file system until 6 hours after the last request, or until the throughput optimization process has completed, whichever time is longer.

  • Deployment type – You can only update the throughput capacity of persistent SSD-based deployment types.

When to modify throughput capacity

Amazon FSx integrates with Amazon CloudWatch, enabling you to monitor your file system's ongoing throughput usage levels. The performance (throughput and IOPS) that you can drive through your file system depends on your specific workload’s characteristics, in addition to your file system’s throughput capacity, storage capacity, and storage type. For information about how to determine your file system's current throughput, see How to use Amazon FSx for Lustre metrics. For information about CloudWatch metrics, see Monitoring with Amazon CloudWatch.