Managing provisioned throughput capacity - FSx for Lustre
Considerations when updating throughput capacityWhen to modify throughput capacity

Managing provisioned throughput capacity

Every FSx for Lustre file system has a throughput capacity that is configured when you create the file system. For file systems using SSD or HDD storage, the throughput capacity is measured in megabytes per second per tebibyte (MBps/TiB). For file systems using Intelligent-Tiering storage, the throughput capacity is measured in megabytes per second (MBps) for the file system. 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 MBps/TiB.

  • For Persistent 2 SSD-based deployment types, valid values are 125, 250, 500, and 1000 MBps/TiB.

You can modify the throughput capacity of an Intelligent-Tiering file system by increasing the value of the total throughput capacity for the file system. Valid values are 4,000 MBps or increments of 4,000 MBps, up to a maximum of 2,000,000 MBps.

You can view the current value of the file system's throughput capacity 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 for SSD-based file systems while the Throughput capacity field shows the current value for Intelligent-Tiering file systems.

  • 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 on SSD file systems or adds new file servers on Intelligent-Tiering file systems. 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.

Topics

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 an SSD-based file system. You can only increase the amount of throughput capacity for an Intelligent-Tiering file system.

  • Update increments – When you modify throughput capacity, use the increments listed in the Update throughput tier dialog box for SSD-based file systems or in the Update throughput capacity dialog box for Intelligent-Tiering file systems.

  • 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 or Intelligent-Tiering deployment types. You cannot modify the throughput capacity of EFA-enabled SSD-based file systems.

When to modify throughput capacity

For the SSD read cache default mode (Proportional to throughput capacity), Amazon FSx automatically provisions 5 GiB of data storage for every MBps of throughput capacity you provision. As you scale your file system’s throughput capacity, Amazon FSx automatically scales your SSD data cache by attaching additional cache storage to any newly added file servers. Note that any time you scale throughput capacity beyond 500 MBps, the contents of your cache will be retained.

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 class. For information about how to determine your file system's current throughput, see How to use Amazon FSx for Lustre CloudWatch metrics. For information about CloudWatch metrics, see Monitoring with Amazon CloudWatch.