Hardware symmetry check ensures nodes selected for an Azure Stack HCI cluster are supported and have symmetrical hardware configurations as recommended by Dell EMC.
Azure stack HCI cluster works best when the hardware configurations of all the selected nodes are compliant with Dell EMC recommendations.
The Integrated Deploy and Update feature in OMIMSWAC helps you to maintain hardware symmetry and update nodes while creating an Azure Stack HCI cluster using AX nodes (running Azure Stack HCI OS) in Windows Admin Center. The Hardware symmetry check runs a set of rules on nodes and helps you to align their hardware configurations with Dell EMC recommendations.
Before you create an Azure Stack HCI cluster, ensure to run and comply with all the rules of Hardware symmetry checks.
Hardware symmetry checks are intended to do the following:
- Find hardware or configuration issues before an Azure stack HCI cluster goes into production.
- Ensure that the Azure stack HCI cluster you deploy is reliable and the cluster performance is optimal.
This topic explains the Hardware symmetry rules and provides examples of supported and unsupported configurations. For information about supported and validated components required for hardware symmetry, see
AS HCI support matrix. If any of the hardware symmetry checks fail with a "Critical" or "Warning" error, review the recommendations and show details and contact the
Dell support team to resolve the issue before proceeding to the next step.
The critical error states that this aspect of nodes configuration is not supported. You must correct the issue before you can deploy an symmetric Azure Stack HCI cluster. And Warning states that this aspect of nodes configuration can be supported for cluster deployment, but might result in sub-optimal cluster performance. Therefore, it should be reviewed.
Example configurations
Here are some supported and unsupported configurations:
Not supported: different models between nodes
The first two nodes use AX-640 model but the third node uses AX-740xd.
Node 1
| Node 2 | Node 3 |
AX-640
| AX-640
| AX-740xd
|
This is not supported. All nodes should have the same model node.
Supported: no more than two media types
Supported configuration for two media types are as follows:
Node 1 | Node 2 | Node 3 | Node 4 | Node 5 |
NVMe+SSD
| NVMe+HDD
| SSD+HDD
| All NVMe
| All SSD
|
Not supported: at least minimum number of drives
If two media types exist:
Node 1 | Node 2 | Node 3 |
2 × SSD for cache
| 2 × SSD for cache
| 2 × SSD for cache
|
3 × HDD for capacity
| 3 × HDD for capacity
| 3 × HDD for capacity
|
This is not supported. Nodes with two media types drives should have 2 SSD drives for cache and 4 HDD drives for capacity.
Supported: at least minimum number of drives
If two media types exist, supported configurations are:
Node 1 | Node 2 | Node 3 |
2 × SSD for cache
| 2 × SSD for cache
| 2 × SSD for cache
|
4 × HDD for capacity
| 4 × HDD for capacity
| 4 × HDD for capacity
|
This is supported. Nodes with two media types drives should have 2 cache drives (SSD/NVMe/AEP) and 4 capacity drives (HDD/SSD/NVMe).
Not supported: drives with different bus protocol
The first two nodes use SSD drive with SAS bus protocol but the third node uses SSD drive with SATA bus protocol.
Node 1 | Node 2 | Node 3 |
SSD with SAS protocol
| SSD with SAS protocol
| SSD with SATA protocol
|
The first two nodes use SSD and HDD drives with SAS bus protocol but the third node uses HDD drive with SATA bus protocol.
Node 1 | Node 2 | Node 3 |
SSD with SAS protocol
| HDD with SAS protocol
| HDD with SATA protocol
|
These are not supported. Drives across nodes should have the same bus protocol for symmetrical configuration.
Not supported: drives with same capacity and different count
The first two nodes use 2TB SSD and the last node uses 3TB SSD. Every node has total 4 SSD.
Node 1 | Node 2 | Node 3 |
4 × 2 TB SSD
| 4 × 2 TB SSD
| 4 × 3 TB SSD
|
This is not supported. All drives of each media type (SSD/NVMe/HDD) should be of the same count and capacity.
Not supported: at least one same RDMA network adapter present in all nodes
The first two nodes use Qlogic network adapter and the last node uses Mellanox network adapter.
Node 1 | Node 2 | Node 3 |
Qlogic network adapter
| Qlogic network adapter
| Mellanox network adapter
|
This is not supported. One network adapters should be common across all nodes.