- Newest
- Most votes
- Most comments
Hi,
The file system in FSx for Lustre consists of one metadata target (MDT) and several object storage targets (OST) (OST). The metadata for files stored in FSxL is stored in the MDT and the file data is distributed across the OSTs to balance storage capacity, throughput, and IOPS load.
As you can see, the throughput of FSx for Lustre is proportional to the storage capacity, so as you increase storage capacity, you will see an increase in IOPS performance.
However, increasing the storage capacity does not immediately re-stripe the data stored in the OST, and the re-striping is applied naturally as files continue to be created/updated/deleted, so it is difficult to see the performance improvement immediately.
We recommend using the following command to check the layout configuration of a file or directory to determine the current striping status of your filesystem.
lfs getstripe path/to/filename
We also recommend using Persistent Type instead of Scratch Type for high IOPS performance.
For more information, see the links below : https://docs.aws.amazon.com/fsx/latest/LustreGuide/performance.html
Relevant content
- asked 4 years ago
- asked 3 years ago
- asked a year ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated 4 months ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated 8 months ago