A TIR Shared File System (SFS) is storage that several TIR Instances can mount at the same time, each with full read and write access. The workers of a training job can mount it too. Use it for a shared workspace that distributed work reads and writes together, instead of copying files between instances by hand.
Where to find it
Section titled “Where to find it”In TIR, open your project. In the side menu, under Storage, select Shared File System.
What you can do
Section titled “What you can do”Create and connect
Section titled “Create and connect”- Create an SFS with a name and a size.
- Mount an SFS on an instance to read and write its data from the instance.
- Mount an SFS on a training job so every worker in the job shares it.
Lifecycle
Section titled “Lifecycle”- Increase the size of an SFS as your data grows.
- Delete an SFS when you are done with it.
Monitor
Section titled “Monitor”- Monitor an SFS: which instances have it mounted, and how much of its space is used.
SFS, PFS, and datasets
Section titled “SFS, PFS, and datasets”TIR also offers datasets and Parallel File System (PFS), shared storage you mount the same way as an SFS. Each suits a different kind of work:
| Storage | Choose it when |
|---|---|
| SFS | Several instances, or every worker of a training job, need to read and write the same files together |
| PFS | Many workers need to read and write the same files in parallel at once |
| Datasets | You mainly read data, often in sequence, rather than write to it from many places at once |
For how SFS and PFS compare in size, region, and bandwidth, see What is the difference between a PFS and an SFS?.
Where to start
Section titled “Where to start”Start with Get started with TIR Shared File System, which creates an SFS and mounts it on an instance. For help with a message, Troubleshooting TIR Shared File System has the fixes, and the TIR Shared File System FAQ has short answers.