Create NFS-based file shares and mount them to Linux virtual machines, bare metal servers, or GPU clusters.
Two types of file shares are available:
* **Standard** are general-purpose file shares using a Ceph-based backend. * **VAST** are high-performance file shares, available in selected GPU-enabled regions only.The creation flow for both types starts the same. Follow the instructions for the selected file share type.
## Prepare the network for Bare Metal
To mount the file share on a Bare Metal server, the network must support Bare Metal, and these servers require a dedicated VLAN.
If needed, create a new network and enable the **Bare Metal Network** toggle during configuration.
A Standard NFS file share connects to a private network and can be mounted on Linux Virtual Machines or Bare Metal servers.
2. In the **Basic settings** panel, enter the **File Share name**, specify **Size**, and select **Standard** as the **File Share type**.
3. In the **Network settings** panel, select the **Network** and **Subnet** to use for the file share.
4. In the **Access** panel, the default **Accessible for the selected subnet** option grants read/write access to all addresses on the allocated network. To restrict access by IP address, select **Custom access rules**, click **+ Add rule**, enter the **Source IP or subnet address**, and select the **Access rule mode**.
5. Set **Additional options**, if required.
Connect to the server via the [Gcore Customer Portal](/cloud/virtual-instances/connect/connect-to-your-instance-via-control-panel) or SSH and set up NFS client support.
For Ubuntu and Debian:
```bash theme={null} sudo apt -y install nfs-common ```For CentOS:
```bash theme={null} sudo yum install nfs-utils ```
Mount the file share:
```bash theme={null} mount 0.0.0.0:/shares/share-d54589b8-132f-4de4-ae99-af5c6cdfdb9c /mount/path ```Replace `/mount/path` with the absolute local directory path where the file share should be mounted.
VAST file shares attach to compute resources (VMs, Bare Metal servers, and GPU clusters) using a dedicated network that is assigned automatically at creation.
When VAST share type is selected, controls in **Network settings** and **Access** panels are disabled, and the network is assigned automatically.
The access rules cannot be configured manually — the VAST share is always available in read/write mode within its assigned network only. This network is not visible in the general **Network** tab. It is only available when attaching interfaces to a resource (VM, Bare Metal, or GPU cluster).
VAST file shares always use read/write access; access rules are not supported.
Once the VAST file share is created, it is best to add the VAST network interface while provisioning the compute resource.
**Adding VAST interface while creating a compute resource**While provisioning a compute resource:
1. Scroll down to the **Network settings** panel. 2. Click **Add interface**. 3. Click the interface to configure, and select the **VAST network**. 4. Continue with provisioning the compute resource.The VAST network only becomes available after the file share has been created. It is a third, dedicated network, separate and distinct from public and private networks.
**Adding VAST interface to an existing compute resource**
While the VAST interface can be attached to an already-provisioned compute resource, this requires additional manual network configuration and is not the standard workflow.
**Attach VAST network interface** 1. Go to server **Resource** settings ([VM](/cloud/virtual-instances/create-an-instance), [Bare Metal](/cloud/bare-metal-servers/create-a-bare-metal-server), or [GPU cluster](/edge-ai/ai-infrastructure/create-a-bare-metal-gpu-cluster)). 2. Select the **Networking** tab and click **Add interface**. 3. Click the **Network** drop-down and select the **VAST network**, then click **Add**. 4. Once the interface is added, note the following details for use in subsequent steps: * **IP** address — `198.51.100.25` * **MAC** address — `fa:16:3e:12:34:56` * **CIDR** range — `198.51.102.0/22` **Configure attached VAST network interface**When attaching a VAST subnet to an existing compute resource, use DHCP to automatically configure the network interface. This is the recommended method as it eliminates the need for manual network configuration.
**DHCP configuration for VMs and GPU clusters**Run the DHCP client on the VAST interface. Replace `enp8s0` with the actual interface name:
```bash theme={null} sudo dhclient enp8s0 ``` **DHCP configuration for Bare Metal servers**The DHCP client automatically configures the IP address, subnet mask, gateway, and routing information, following the same [network configuration](/cloud/networking/ip-address/create-and-configure-a-reserved-ip-address) approach used across Gcore Cloud.
**Manual configuration (alternative method)**For custom network settings, configure the interface manually instead.
Connect to the server via the Customer Portal or SSH and configure the attached VAST network interface.
1. Use `ip addr` to list all available interfaces. 2. Identify the VAST interface using the MAC address from the previous step. 3. Note the **interface name** for use in the following steps. Interface names follow the format `enp8s0` or `enp4s0`. 4. Configure the interface as described below for the relevant instance type.In all commands below, replace `198.51.100.25` with the IP address from step 2, and replace `198.51.100.1` with the gateway IP — take the first three octets of the CIDR range from step 2 and append `.1`.
**Configuring VAST interface for VMs and GPU clusters**Replace `enp8s0` with the interface name from step 3.
Set interface MTU:
```bash theme={null} sudo ip link set enp8s0 mtu 9000 ```Activate the interface:
```bash theme={null} sudo ip link set enp8s0 up ```Set interface CIDR:
```bash theme={null} sudo ip addr add 198.51.100.25/22 dev enp8s0 ```Add static route to the VAST network:
```bash theme={null} sudo ip route add 172.19.252.0/22 via 198.51.100.1 dev enp8s0 ``` **Configuring VAST interface for Bare Metal servers**Bare Metal servers require different configuration depending on the network hardware.
Connect to the server via the Customer Portal or SSH and set up VAST NFS client support.
1. Install NFS client tools: ```bash theme={null} sudo apt-get update && sudo apt-get install nfs-common -y ``` 2. Install build tools and headers for kernel modules: ```bash theme={null} sudo apt install dkms debhelper dh-dkms build-essential linux-headers-$(uname -r) -y ``` 3. Install VAST NFS package: ```bash theme={null} curl -sSf https://vast-nfs.s3.amazonaws.com/download.sh | bash -s -- ```Use an existing directory for mounting the share, or create a new one:
```bash theme={null} sudo mkdir -p /mount/path ```Mount the VAST file share:
* Replace `ndp1-2-vast.cloud.gc.onl:/manila/manila-01234567-89ab-cdef-0123-456789abcdef` with the connection point from the VAST file share overview tab. * Replace `/mount/path` with the absolute local directory path where the file share should be mounted. ```bash theme={null} sudo mount -o vers=3,nconnect=56,remoteports=dns,spread_reads,spread_writes,noextend ndp1-2-vast.cloud.gc.onl:/manila/manila-01234567-89ab-cdef-0123-456789abcdef /mount/path ```The contents of the VAST file share are now accessible in the specified directory.
This process is the same for both Standard and VAST-based file shares:
NFS file shares provide shared persistent storage accessible from any VM on the same private subnet. Each step below creates one resource — network, subnet, file share — and ends with the connection point used in the mount command.
Before picking a region, check that file shares are available there — look for `has_sfs: true` in the regions response:
```bash theme={null} curl "https://api.gcore.com/cloud/v1/regions" \ -H "Authorization: APIKey $GCORE_API_KEY" ```Provision an NFS file share with per-CIDR access rules. The [`gcore_cloud_file_share`](https://registry.terraform.io/providers/G-Core/gcore/latest/docs/resources/cloud_file_share) resource and its access rule companion are declared in a single configuration.
## Create a file shareDeclare the file share and access rule using an existing network and subnet.
```hcl theme={null} resource "gcore_cloud_file_share" "example" { project_id = var.project_id region_id = var.region_id name = "my-share" protocol = "NFS" size = 5 # GiB; minimum is 1 network = { network_id = var.network_id subnet_id = var.subnet_id } # terraform import gcore_cloud_file_share.example 'After `terraform apply`, `connection_point` holds the NFS server address and share path for the mount command, in the format `
Edit the `size` field and run `terraform apply` — size can only be increased. Standard shares must be unmounted before resizing.
```hcl theme={null} resource "gcore_cloud_file_share" "example" { # ... (unchanged fields) size = 10 # increased from 5 } ``` ```bash theme={null} terraform apply ``` ## Delete a file shareRemove the resource blocks — Terraform deletes the access rule first, then the file share on the next `terraform apply`.
```hcl theme={null} # Remove or comment out both blocks: # resource "gcore_cloud_file_share_access_rule" "example" { ... } # resource "gcore_cloud_file_share" "example" { ... } ``` ```bash theme={null} terraform apply ```