Managing Storage for VCF Automation (VCFA) Content Library Transfer Spooling Area

When a user uploads an ISO/OVF/OVA from their desktop to VCF Automation (VCFA) Content Library, the file is temporarily stored in the Transfer Spooling Area (TSA) that resides locally within VCFA before it is finally transferred to the destination vCenter Server.

By default, VCFA uses an embedded object store backed by SeaweedFS to hold the temporary user files, which are stored in 64MB chunks. The size of the SeaweedFS volumes will be based on the VCFA deployment mode (single node vs multi-node for HA). A total of three volumes (3 replicas) will always be configured regardless of the deployment mode for VCFA and the total usable storage capacity is sum of these three volumes.

• Single Node, each volume is 75GB
• Multi-Node (Medium), each volume is 150GB
• Multi-Node (Large), each volume is 200GB

The SeaweedFS volumes can be resized non-disruptively by going into VCF Operations and navigating to Fleet Management->Lifecycle->Management->Components and then select your VCF Automation instance and perform the storage resize operation.

Enter the desired capacity for the Shared Storage Data and VCFA will then automatically resize the volume without any user impact.

Alternatively, VCFA can be configured with a more scalable option that leverages an external NFS server for the TSA function to accommodate larger file uploads.

To configure the TSA storage backing from the internal SeaweedFS to an external NFS server, we need to issue an API request when logged into VCFA Appliance. Create a shell script called configure_nfs.sh which should contain the following and update the three variables at the top of the script:

NFS_SERVER_IP="192.168.30.29"
NFS_SERVER_MOUNT_PATH="/volume1/vcfa"
NFS_SERVER_MOUNT_SIZE_GIB="500"

### DO NOT EDIT BEYOND HERE ###

K8S_TOKEN=$(kubectl get secrets synthetic-checker-krp -n vmsp-platform -ojsonpath={.data.token} | base64 -d)
NODE_IP=$(kubectl get nodes -ojsonpath='{.items[0].status.addresses[?(@.type=="InternalIP")].address}')

REQUEST_RESPONSE=$(curl -k -s -X POST -H "Authorization: Bearer $K8S_TOKEN" https://${NODE_IP}:30005/webhooks/prelude/tenant-manager/configure-nfs \
-d "{
        \"host\": \"$NFS_SERVER_IP\",
        \"path\": \"$NFS_SERVER_MOUNT_PATH\",
        \"size\": \"$NFS_SERVER_MOUNT_SIZE_GIB\"
    }"
curl -q -L -k -X GET -H "Authorization: Bearer $K8S_TOKEN"   https://${NODE_IP}:30005/$(echo $REQUEST_RESPONSE | jq -r .statusURI) | jq .

Once the script has been successfully executed, you should see a 200 response code that indicates the API endpoint has received the request and the reconfiguration is now in progress.

The /configure-nfs webhook will first check whether the NFS server and volume is accessible by spinning up a few helper containers, which we can see by running the following command:

kubectl -n prelude get pods | grep tenant

You can run the following command to ensure that the container is able to update the ownership of the NFS volume:

kubectl -n prelude logs tenant-manager-0 -c prepare-dir-ownership

We can confirm the NFS mount was successful by running the following command on the tenant-manager pod:

kubectl -n prelude exec -it tenant-manager-0 -- mount | grep nfs

Lastly, the Tenant Manager pod should now be running after the storage reconfiguration:

kubectl -n prelude get pods -l app.kubernetes.io/instance=tenant-manager

When uploading a new file, we should now see that the temporary file is now being stored on our NFS server

To configure the TSA storage backing from an external NFS server to the internal SeaweedFS, we simply issue an API request when logged into VCFA Appliance with an empty string for the host property. Create a shell script called configure_seaweedfs.sh which should contain the following:

K8S_TOKEN=$(kubectl get secrets synthetic-checker-krp -n vmsp-platform -ojsonpath={.data.token} | base64 -d)
NODE_IP=$(kubectl get nodes -ojsonpath='{.items[0].status.addresses[?(@.type=="InternalIP")].address}')

REQUEST_RESPONSE=$(curl -k -X POST -H "Authorization: Bearer $K8S_TOKEN" https://${NODE_IP}:30005/webhooks/prelude/tenant-manager/configure-nfs -d '{"host":""}')
curl -q -L -k -X GET -H "Authorization: Bearer $K8S_TOKEN"   https://${NODE_IP}:30005/$(echo $REQUEST_RESPONSE | jq -r .statusURI) | jq .

You can monitor the storage reconfiguration similar to configuring an NFS server and /configure-nfs webhook does provide an overall status that you can perform GET request by running the following:

K8S_TOKEN=$(kubectl get secrets synthetic-checker-krp -n vmsp-platform -ojsonpath={.data.token} | base64 -d)
NODE_IP=$(kubectl get nodes -ojsonpath='{.items[0].status.addresses[?(@.type=="InternalIP")].address}')

curl -k -X GET -H "Authorization: Bearer $K8S_TOKEN" https://${NODE_IP}:30005/webhooks/prelude/tenant-manager/configure-nfs