Step 2: Create Mapping File with Node Information
In Omnia, nodes are discovered and provisioned based on the groups and functional groups defined in the mapping file. By combining both groups and functional groups, Omnia offers a powerful and flexible approach to managing large-scale node infrastructures, ensuring both logical organization and physical optimization of resources.
A group is based on the physical characteristics of the nodes. It refers to nodes that are located in the same place or have similar hardware. For example, nodes in the same rack or SU (Scalable Unit) might be grouped together, with specific functional groups like Service Kube Node or Slurm Control Node. Groups help with physical organization and management of nodes.
A functional group defines what a node does in the system. It is a way to categorize nodes based on their functionality. Functional groups help group nodes that perform similar tasks, making it easier to manage and assign resources. For example, a node could belong to a functional group such as:
Service Kube Control Plane
Service Kube Node
Slurm Login Node
Slurm Login/Compiler Node
Slurm Control Node
Slurm Compute Node
Minimal OS
Create Mapping File
Omnia supports two methods for discovering target nodes and creating PXE mapping files:
Manual PXE file Mapping: Manually collect PXE NIC information of the nodes to be provisioned and manually define them in the pxe_mapping_file.csv file to be used by Omnia. See Create PXE File Manually for detailed instructions.
OME-based BMC PXE file Generation (Recommended): Use OpenManage Enterprise (OME) to discover the Omnia cluster nodes and generate the PXE mapping file using the
discovery.ymlplaybook. See Create PXE File Using OME for detailed instructions.
Create PXE File Manually
Manually collect PXE NIC information of the nodes to be provisioned and manually define them to Omnia using the pxe_mapping_file.csv file. Provide the file path to the pxe_mapping_file_path variable in /opt/omnia/input/project_default/provision_config.yml.
Each node listed in the mapping file must be assigned with the following values:
FUNCTIONAL_GROUP_NAME, GROUP_NAME, SERVICE_TAG, PARENT_SERVICE_TAG, HOSTNAME, ADMIN_MAC,
ADMIN_IP, BMC_MAC, and BMC_IP.
Refer to the Group Attributes table to assign the appropriate
GROUP_NAME and the Types of Functional Groups table to
assign the correct FUNCTIONAL_GROUP_NAME for each node in the mapping file.
The following is the sample format of a mapping file for x86_64 cluster:
FUNCTIONAL_GROUP_NAME,GROUP_NAME,SERVICE_TAG,PARENT_SERVICE_TAG,HOSTNAME,ADMIN_MAC,ADMIN_IP,BMC_MAC,BMC_IP,IB_NIC_NAME,IB_IP
slurm_control_node_x86_64,grp0,ABCD12,,slurm-control-node1,a1:b2:c3:d4:e5:f6,172.16.107.52,a2:b3:c4:d5:e6:f7,172.17.107.52,InfiniBand.Slot.7-1,192.168.0.100
slurm_node_x86_64,grp1,ABCD34,ABFL82,slurm-node1,b1:c2:d3:e4:f5:a6,172.16.107.43,b2:c3:d4:e5:f6:a7,172.17.107.43,InfiniBand.Slot.7-1,192.168.0.101
slurm_node_x86_64,grp1,ABFG34,ABKD88,slurm-node2,c1:d2:e3:f4:a5:b6,172.16.107.44,c2:d3:e4:f5:a6:b7,172.17.107.44,InfiniBand.Slot.7-1,192.168.0.102
login_compiler_node_x86_64,grp8,ABCD78,,login-compiler-node1,d1:e2:f3:a4:b5:c6,172.16.107.41,d2:e3:f4:a5:b6:c7,172.17.107.41,InfiniBand.Slot.7-1,192.168.0.103
login_compiler_node_x86_64,grp8,ABFG78,,login-compiler-node2,e1:f2:a3:b4:c5:d6,172.16.107.42,e2:f3:a4:b5:c6:d7,172.17.107.42,InfiniBand.Slot.7-1,192.168.0.104
service_kube_control_plane_x86_64,grp3,ABFG79,,service-kube-control-plane1,f1:a2:b3:c4:d5:e6,172.16.107.53,f2:a3:b4:c5:d6:e7,172.17.107.53,,InfiniBand.Slot.7-1,192.168.0.105
service_kube_control_plane_x86_64,grp4,ABFH78,,service-kube-control-plane2,11:22:33:44:55:66,172.16.107.54,12:23:34:45:56:67,172.17.107.54,,InfiniBand.Slot.7-1,192.168.0.106
service_kube_control_plane_x86_64,grp4,ABFH80,,service-kube-control-plane3,aa:bb:cc:dd:ee:01,172.16.107.55,ab:bc:cd:de:ef:12,172.17.107.55,,InfiniBand.Slot.7-1,192.168.0.107
service_kube_node_x86_64,grp5,ABFL82,,service-kube-node1,33:44:55:66:77:88,172.16.107.56,34:45:56:67:78:89,172.17.107.56,InfiniBand.Slot.7-1,192.168.0.108
service_kube_node_x86_64,grp5,ABKD88,,service-kube-node2,55:66:77:88:99:aa,172.16.107.57,56:67:78:89:aa:bb,172.17.107.57,InfiniBand.Slot.7-1,192.168.0.109
The following is the sample format of a mapping file for x86_64 and aarch64 cluster:
FUNCTIONAL_GROUP_NAME,GROUP_NAME,SERVICE_TAG,PARENT_SERVICE_TAG,HOSTNAME,ADMIN_MAC,ADMIN_IP,BMC_MAC,BMC_IP,IB_NIC_NAME,IB_IP
slurm_control_node_x86_64,grp0,ABCD12,,slurm-control-node1,a1:b2:c3:d4:e5:f6,172.16.107.52,a2:b3:c4:d5:e6:f7,172.17.107.52,InfiniBand.Slot.7-1,192.168.0.100
slurm_node_aarch64,grp1,ABCD34,ABFL82,slurm-node1,b1:c2:d3:e4:f5:a6,172.16.107.43,b2:c3:d4:e5:f6:a7,172.17.107.43,InfiniBand.Slot.7-2,192.168.0.101
slurm_node_aarch64,grp2,ABFG34,ABKD88,slurm-node2,c1:d2:e3:f4:a5:b6,172.16.107.44,c2:d3:e4:f5:a6:b7,172.17.107.44,NIC.InfiniBand.1-3,192.168.0.102
login_compiler_node_aarch64,grp8,ABCD78,,login-compiler-node1,d1:e2:f3:a4:b5:c6,172.16.107.41,d2:e3:f4:a5:b6:c7,172.17.107.41,InfiniBand.PCIe.Slot.8-1,192.168.0.103
login_node_aarch64,grp9,ABFG78,,login-node1,e1:f2:a3:b4:c5:d6,172.16.107.42,e2:f3:a4:b5:c6:d7,172.17.107.42,NIC.InfiniBand.1-1,192.168.0.104
service_kube_control_plane_x86_64,grp3,ABFG79,,service-kube-control-plane1,f1:a2:b3:c4:d5:e6,172.16.107.53,f2:a3:b4:c5:d6:e7,172.17.107.53,,
service_kube_control_plane_x86_64,grp4,ABFH78,,service-kube-control-plane2,11:22:33:44:55:66,172.16.107.54,12:23:34:45:56:67,172.17.107.54,,
service_kube_control_plane_x86_64,grp4,ABFH80,,service-kube-control-plane3,aa:bb:cc:dd:ee:01,172.16.107.55,ab:bc:cd:de:ef:12,172.17.107.55,,
service_kube_node_x86_64,grp5,ABFL82,,service-kube-node1,33:44:55:66:77:88,172.16.107.56,34:45:56:67:78:89,172.17.107.56,,
service_kube_node_x86_64,grp5,ABKD88,,service-kube-node2,55:66:77:88:99:aa,172.16.107.57,56:67:78:89:aa:bb,172.17.107.57,,
os_x86_64,grp6,ABEF56,,os-node1,77:88:99:aa:bb:cc,172.16.107.60,78:89:aa:bb:cc:dd,172.17.107.60,,
os_aarch64,grp7,ABEF78,,os-node2,99:aa:bb:cc:dd:ee,172.16.107.61,9a:ab:bc:cd:de:ef,172.17.107.61,,
Note
Ensure that nodes belonging to the same group have the same parent. In the mapping file, node entries with the same
GROUP_NAMEmust have the same parent specified in thePARENT_SERVICE_TAGcolumn.The header fields mentioned above are case sensitive.
The IP addresses provided in the mapping file are not validated by Omnia. Ensure that the correct IP addresses are provided. Incorrect IP addresses can cause unexpected failures.
The service tags provided in the mapping file are not validated by Omnia. Ensure that correct service tags are provided. Incorrect service tags can cause unexpected failures.
The hostnames provided should not contain the domain name of the nodes.
All fields mentioned in the mapping file are mandatory.
The ADMIN_MAC and BMC_MAC addresses provided in
pxe_mapping_file.csvshould refer to the PXE NIC and BMC NIC on the target nodes respectively.Target servers should be configured to boot in PXE mode with the appropriate NIC as the first boot device.
Note
Minimal OS Functional Groups: The os_x86_64 and os_aarch64 functional groups provide a clean operating system baseline designed for downstream platform software installation. These groups include only essential OS packages and LDMS telemetry packages, with no schedulers, container runtimes, or orchestration software. Use these groups when you need a clean OS environment without conflicts from pre-installed components.
Additional Packages Support: Administrators can optionally include additional packages by creating additional_packages.json files in input/config/{arch}/rhel/10.0/. For detailed instructions on configuring additional packages, see Add Additional Packages. When present, these packages are included in the Minimal OS images alongside the base and LDMS packages. If the file is absent or empty, images build successfully with the standard Minimal OS package set only.
Create PXE File Using OME
OME-based BMC discovery is the recommended method for discovering target nodes. This mechanism leverages OpenManage Enterprise to automatically discover servers through their BMC/iDRAC interfaces, reducing manual configuration effort.
Note
In Dell Omnia deployments integrated with OpenManage Enterprise (OME), server identification and mapping during PXE boot rely on information retrieved from OME and iDRAC inventory. Depending on the DNS environment, the DnsName value may match the intended iDRAC hostname, or may return a reverse DNS name (e.g., pool‑<IP‑based>), which may not align with naming conventions required for cluster configuration. Due to differences between iDRAC configuration and OME‑reported hostnames, users must explicitly define GROUP_NAME and PARENT_SERVICE_TAG in the pxe_mapping_file to ensure accurate PXE provisioning and cluster setup in Omnia.
Prerequisites
Before proceeding with OME discovery, ensure the following:
OpenManage Enterprise is installed and accessible
All target servers have iDRAC configured with network connectivity
OME has discovered the devices (servers are visible in OME inventory)
You have administrative access to OME
Ensure that servers have the correct NIC order and configuration to match your intended IP assignment scheme. When Omnia performs OME-based discovery, it uses the following NIC selection logic:
Admin IP: The first discoverable NIC (typically the first Ethernet interface) will be used to generate the admin IP address in the PXE mapping file
InfiniBand IP: The first discoverable InfiniBand NIC will be used to generate the InfiniBand IP address in the PXE mapping file
You must verify NIC ordering in the server BIOS or iDRAC settings before discovery.
Procedure
In OpenManage Enterprise, discover the cluster nodes that you want to provision with Omnia. For more information on discovering devices in OME, see the OpenManage Enterprise User Guide.
After discovering the nodes, create static groups for each Omnia functional group type supported in Omnia. For more information on groups and functional group support in Omnia, see Groups and Functional Groups.
slurm_control_node_x86_64slurm_node_x86_64login_compiler_node_x86_64service_kube_control_plane_x86_64service_kube_node_x86_64slurm_node_aarch64login_node_aarch64login_compiler_node_aarch64os_aarch64
To create static groups in OME:
In the left navigation menu, navigate to CUSTOM GROUPS > Static Groups
Click the ellipsis (…) next to Static Groups and select Create Group
Provide the group name exactly matching the functional group name
Add a description for the group.
Click Finish
Repeat this process for each functional group type you plan to use in your Omnia deployment.
After creating the static groups for each functional group type, add the discovered nodes to the corresponding static groups. To add the devices to the static groups:
Select the static functional group from the list.
Click Add Devices.
In the Add Devices to Group <static group name> dialog box, select the servers that belong to a specific functional group.
Click Finish
Repeat this process for all functional groups, ensuring each server is assigned to the correct static group based on its intended role in the Omnia cluster.
Note
When you run the discovery.yml playbook, devices that are not assigned to any Omnia-supported custom static group will be considered as slurm_node_aarch64 in the auto-generated PXE mapping file.
After creating the static groups in OME, configure the
discovery_config.ymlfile with OME connection details and discovery parameters. The following table lists the parameters fordiscovery_config.yml:
Parameter |
Mandatory/Optional |
Details |
|---|---|---|
|
Optional |
|
|
Conditional Mandatory |
|
Execute the
discovery.ymlplaybook with thediscovery_mechanism=omeparameter to generate the PXE mapping file automatically:ssh omnia_core cd /omnia/discovery ansible-playbook discovery.yml -e "discovery_mechanism=ome"
The discovery.yml file will automatically create the PXE mapping file in the /opt/omnia/input/project_default/ directory. For example bmc_pxe_mapping_file_<timestamp>.csv with the discovered nodes from OME. The user can verify and edit the mapping file if necessary.
Groups
Nodes that are located in the same place or similar hardware can be grouped together. To do so, update the mapping file with all necessary attributes for the nodes, based on their role within the cluster. Each group will have following attributes as indicated in the table below:
Attribute |
Mandatory/Conditional mandatory/Optional |
Description |
|---|---|---|
Group Name - |
Mandatory |
|
Parent of the node- “parent’’ |
Conditional Mandatory |
|
Functional Groups
Nodes with similar functional roles or functionalities can be grouped together. The following table lists the functional groups available in Omnia.
Note
At least one functional group is mandatory, and you must not change the name of functional groups.
Ensure that the group nodes intended for a specific role must be associated with the corresponding functional group and must not be associated under multiple functional groups.
The functional groups are case-sensitive.
Omnia supports HA functionality for the
service_cluster. For more information, click here.To set up a service cluster, the
service_kube_nodemust be present in the mapping file.
Functional Group Name |
Layer |
Details |
|---|---|---|
Slurm control plane - |
Management |
|
Slurm compute node - |
Compute |
|
Slurm compute node - |
Compute |
|
Service Cluster Kubernetes control plane- |
Management |
|
Service Cluster Kubernetes worker node - |
Management |
|
Slurm Login node - |
Management |
|
Slurm Login node - |
Management |
|
Slurm Login and Compiler node - |
Management |
|
Slurm Login and Compiler node- |
Management |
|
Minimal OS compute node - |
Compute |
|
Minimal OS compute node - |
Compute |
|
Recommended Software by Functional Groups
Caution
Ensure that the software_config.json file contains all required inputs for the software to be deployed on each functional group. For more information, see Input parameters for Local Repositories.
The following table lists the functional groups along with the recommended software to be deployed on each group.
Functional Group Name |
Recommended Software |
|---|---|
service_kube_control_plane_x86_64 |
service_k8s.json |
service_kube_node_x86_64 |
service_k8s.json |
slurm_control_node_x86_64 |
slurm_custom.json, openldap.json, ldms.json |
slurm_node_x86_64 |
slurm_custom.json, openldap.json, ldms.json |
slurm_node_aarch64 |
slurm_custom.json, openldap.json, ldms.json |
login_node_x86_64 |
slurm_custom.json, openldap.json, ldms.json |
login_node_aarch64 |
slurm_custom.json, openldap.json, ldms.json |
login_compiler_node_x86_64 |
slurm_custom.json, openldap.json, ucx.json, openmpi.json, ldms.json |
login_compiler_node_aarch64 |
slurm_custom.json, openldap.json, ucx.json, openmpi.json, ldms.json |
os_x86_64 |
default_packages.json, ldms.json |
os_aarch64 |
default_packages.json, ldms.json |
Note
The os_x86_64 and os_aarch64 functional groups support optional additional packages via additional_packages.json files. Create these files in input/config/{arch}/rhel/10.0/ to include custom packages like podman, diagnostic tools, or monitoring agents. If no additional packages are needed, the images build successfully with the standard package.
If you have any feedback about Omnia documentation, please reach out at omnia.readme@dell.com.