This topic covers the requirements for creating an Autodesk Flame instance on AWS.
Once the Flame instance is created, you need to configure it. Follow the Steps to configure a Flame instance.
This topic covers the creation of a Flame instance networked with other Flame instances, with or without a Project server. For a standalone Flame instance, see Autodesk Flame Configuration (Standalone).
From the Autodesk Flame AMI you created, deploy the Flame instance. You must set its instance type, storage, the EFA (if using AWS CDI), and security groups:
Instance type: At least G4dn.8xlarge or G5.8xlarge. If you plan on using EFA, only the G4 instance is supported.
Storage:
If you plan on using AWS CDI: You must configure the Flame instance to use the EFA (Elastic Fabric Adapter) network device. You cannot can add an EFA network device after the instance creation.
Security Groups:
Other settings are set according to your requirements, but make sure to set:
When you create the instance, the option Instance volume deletion on termination controls what happens to the storage when you terminate the instance. If the volume is not deleted on instance termination, while you're no longer paying for the terminated instance, you're still paying for the storage.
To avoid paying for storage you no longer need, the instance volume can be deleted automatically when the instance is terminated.
Before selecting Delete on termination, understand that data stored on the instance volume is permanently deleted when the instance is terminated.
After launching your Flame instance for the first time, you must configure it to your storage and network requirements.
Once all of this is done, you connect to Flame with the user you created, using AWS NICE DCV or HP Anyware.
From your local machine, connect to the instance through ssh. Use the flameadmin account and the keypair of the instance.
ssh -i <keypair> flameadmin@<INSTANCE PUBLIC IP>
In a shell, run the command to assign a new machine ID to your instance:
dbus-uuidgen | sudo tee /etc/machine-id
To stop Autodesk Backburner and Autodesk Stone & Wire Services.
sudo systemctl stop adsk_backburner
sudo systemctl stop adsk_sw
sudo hostnamectl set-hostname --static <Your-host-name>
Follow the instructions for your identity management system.
In a shell, enter the following command.
echo Y | sudo /opt/Autodesk/sw/tools/sw_generateFramestoreId
In a shell, enter the following:
sudo /opt/Autodesk/backburner/backburnerConfig
At the prompts, provide the following answers:
Do you want to run the Backburner Manager on this machine? ([y]es/[n]o/[s]kip/[q]uit): nDo you want to run the Backburner Server(s) on this machine? ([y]es/[n]o/[s]kip): yEnter the hostname (or IP address) of the Manager for this Server or localhost to use default. (leave empty to keep 'localhost'): <Your-ProjectServer-hostname>Enter the hostname (or IP address) that the Manager should use to communicate to this Server or localhost to use default. (leave empty to keep 'localhost'): <Your-Flame-name>To restart Autodesk Backburner and Autodesk Stone & Wire Services.
sudo systemctl start adsk_backburner
sudo systemctl start adsk_sw
By default, the Flame instance doesn't have a swap partition. It's missing from the instance because the large number of possible use cases makes it impossible to predict and set the size of the swap partition.
You create the swap partition and size it according to your needs. To help you out, here is an example of a configuration that uses the ephemeral instance storage available to G4dn and G5 instances to set up a swap partition. You want to use the ephemeral storage for the swap partition because A. The swap partition doesn't need saving between restarts and B. Ephemeral storage costs are included in your instance costs.
This method is not optimal as it uses all the available space on the device, making the swap partition needlessly large. Consult AWS documentation for other swap partition creation methods.
Find the instance storage by listing the disks attached to the instance with lsblk:
[flameadmin@vxfhost ~]$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme0n1 259:1 0 50G 0 disk
└─nvme0n1p1 259:2 0 50G 0 part /
nvme1n1 259:0 0 838.2G 0 disk
In this example, nvme1n1 is not mounted, indicating that this is our ephemeral instance storage device (data on this device is not kept when the instance is stopped). This is a good candidate for use as a swap file.
Do not select your media storage!
Verify that the swap partition is always created on the ephemeral storage when the instance starts, add the following to /etc/rc.local
# Create swap partition
mkswap /dev/nvme1n1
swapon /dev/nvme1n1
Verify that the device name matches the ephemeral device name found earlier with lsblk!
Restart the instance with
sudo reboot
Once the instance is restarted, verify that the swap space is now available with lsblk:
[flameadmin@vxfhost ~]$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme0n1 259:1 0 50G 0 disk
└─nvme0n1p1 259:2 0 50G 0 part /
nvme1n1 259:0 0 838.2G 0 disk [SWAP]
The swap partition is identified by the [SWAP] tag. In the previous example, nvme1n1 is now identified as the swap partition.
In a workgroup configuration, your media is stored on a network attached storage (NAS). You've configured the NAS previously, and now you mount it on the instance to use it as your media storage.
In a shell, run the following commands.
Install autofs.
sudo yum install autofs
Create the directory where the media is stored.
sudo mkdir /mnt/StorageMedia
Redirect to the NAS. Use the command sudo vi /etc/auto.direct and add the following line to the file
/mnt/StorageMedia -rw,noatime,nodiratime,bg nas:/mnt/nas/StorageMedia
nas. Different shared storage solutions will use slightly different syntax. You must adapt this line to your setup.Set up the automount. Use the command sudo vi /etc/auto.master and add the following line at the end of the file.
/- auto.direct
Enable and restart autofs.
sudo systemctl enable autofs
sudo systemctl start autofs
What you do here depends on whether or not you're using an identity management system.
If you're using an identity management system, now's the time to add the instance to the system. The procedure to do so is outside the scope of this document, but here are links to the documentation of some common identity management systems.
From now on, connect to the instance with an identity managed user instead of flameadmin.
If you're not using an identity management system, you must create a new user account. Use this account to connect to the instance with AWS NICE DCV or HP Anyware because flameadmin cannot be used with NICE DCV or HP Anyware.
This user account must match in user name, user ID, and user group one of the user accounts defined on the Burn instances for Burn jobs to be successful.
From the root account, open a shell.
Create the user account. Enter:
sudo useradd <username>
Create the user account, making sure to correctly enter ther user's <user id>, <user group>, and <username>.
sudo useradd -u <user id> -g <user group> <username>
Follow the onscreen instructions.
Once passwd is done, give the user administrative privileges:
sudo usermod -aG wheel <username>
sudo command to run commands usually reserved to the root account.