Inside the Dell EMC Metro Node (MN-114): Architecture & Hardware Deep Dive

Welcome back to Terminal Notes. When engineering multi-site cloud and on-prem architectures, achieving true high availability (HA) is often the hardest puzzle to solve. Today, we are taking a deep dive into the hardware that makes zero-downtime storage possible across geographical distances: the Dell EMC Metro Node (MN-114).

Widely recognized as the modern successor to the legacy Dell EMC VPLEX, the Metro Node is a purpose-built virtualization appliance that sits between your host servers and your storage arrays.

What Does It Do?
The Metro Node acts as a storage abstraction layer. Its primary function is to provide Active-Active High Availability across metro distances.

By synchronously mirroring data across two different data centers, it ensures a Zero RPO (Recovery Point Objective) and Zero RTO (Recovery Time Objective). If an entire storage array or data center goes offline, the other site seamlessly continues serving I/O without any manual failover scripts or application disruption. It also enables non-disruptive data mobility, allowing admins to migrate massive workloads between arrays in the background.

Hardware Architecture: Under the Hood
The MN-114 is built on a highly reliable enterprise foundation. Here is how the hardware breaks down:

The Chassis: Each node is a 1U rack server (specifically utilizing the Dell PowerEdge R640 architecture).

The Cluster: Because redundancy is the core feature, Metro Nodes are always deployed as a pair. This means a standard local cluster occupies 2U of rack space.

Power Consumption: * Each 1U node draws approximately 271 Watts under typical operating loads.

For the 2U cluster, you should plan for a continuous load of about 542 Watts.

Each node features Dual 750W Redundant Power Supplies. In a standard rack, these must be cross-cabled to separate PDUs to maintain HA.

Connectivity: Network vs. SAN Interfaces
To handle massive data replication and cluster communication, the MN-114 is equipped with dedicated PCIe cards for different traffic types. Tying into our previous discussion on enterprise optics, here is how the ports are configured:

The SAN Ports (Storage Area Network): * Typically handled by a 2-port HBA card populated with Avago 32G SFPs.

These ports use the Fibre Channel protocol to handle the heavy lifting: connecting to your FC switches, talking to the underlying storage arrays (like PowerStore or Unity XT), and presenting virtual volumes to your host servers.

The Network Ports (LAN/WAN): * Handled by a 4-port NIC populated with Dell EMC-branded SFPs (usually running at 10GbE or 25GbE).

These handle Ethernet traffic divided into three critical streams:

Management: Web UI and SSH access.

Local COM: The cross-connect that allows Node A to sync with Node B in the same rack.

WAN COM: The vital link that connects this cluster to the secondary cluster at your disaster recovery site.

Lifecycle Status
If you are managing MN-114s in your environment, it is crucial to track their lifecycle metrics. The MN-114 hit its End of Sale (EOS) on June 30, 2024. However, Dell will continue to provide official mechanical support and software updates until its End of Service Life (EOSL) on June 30, 2029.

If you are planning infrastructure roadmaps, you have a solid runway left on these appliances before a hardware refresh is mandated.

Have you deployed Metro Nodes in your environment? Let me know your thoughts and topology designs in the comments.