As Data Center application demands continue to evolve at a rapid pace, today’s IT infrastructure and operations personnel need to adopt fast-moving technologies with the same dexterity. Automation of the existing infrastructure, migration/upgrade toward fabric architectures as well as the ability to seamlessly adopt the cloud, are key requirements, to fulfill the business needs of faster time to market, service agility, and network automation. In this day and age, we strongly believe that the network manager, controller, and orchestrator plays an important role in this transition by delivering fabric automation, enhanced visibility, and operational simplicity.
Cisco’s Data Center Network Manager version 11 presents the next giant stride towards addressing these market requirements in conjunction with the Cisco Open Nexus and MDS family of switches. DCNM’s journey started about ten years ago, primarily as a solution for SAN centric device management. Since 2017, DCNM has evolved significantly to become a state-of-the-art single pane of glass solution for all types of LAN and SAN network deployments. DCNM now delivers a comprehensive management solution for VXLAN-EVPN programmable fabrics, classic LAN deployments, SAN deployments, and IP Fabrics for Media.
DCNM 11 presents a paradigm shift, on the automation front, for devices running NX-OS. In this age of intent-based networking, DCNM provides the ability to manage legacy deployments with the same ease as new fabric deployments based on VXLAN BGP EVPN. This allows data center network administrators to employ a single tool for network management and automation as they transition their existing access-aggregation-core network toward a spine-leaf based fabric.
DCNM can be introduced in the midst of an existing network deployment without having to destroy and re-create the entire network deployment. Unlike most tools, DCNM presents an equally powerful management solution for both greenfield and brownfield deployments. The drift in the network configuration due to out-of-band changes is reported by the configuration compliance module in DCNM thereby allowing a sync top-down or bottom-up, depending on the user’s choice.
A novel microservices based architecture with an application framework, enables DCNM to host native as well as 3rd party applications. This holistic framework provides the entire life cycle management of the application, including support for a scale-out clustered mode that allows compute elements to be dynamically added to the DCNM “mothership” to address the needs of a growing network. An example application, hosted on the DCNM that provides compute visibility from the network edge, all the way to the VMs, is depicted below:
Cisco DCNM 11 provides a controller-centric, top-down approach for Day 0, 1 and 2 operations of multiple data center networks that may have been deployed across multiple geographic locations. This includes support for provisioning underlay and overlay configuration aligned with Cisco published best practices. Since its launch, DCNM 11 has delighted our beachhead customers in an unprecedented way. In many forums, customers have personally expressed their excitement in using the product; thereby validating the architecture, design, and implementation choices made by the DCNM team. I feel very privileged to be part of the team that is delivering this powerful yet easily consumable solution.
We are at a point in our journey where we would like to describe and discuss various aspects of DCNM. We will be kicking off a series of blogs, with the first one being a brief introduction to the novel Data Center Fabric Builder. Never before has it been this easy to build a new fabric with just a few clicks. Some highlights of the fabric builder include a one-click vPC pair setup, customizable fabric policies based on python, per switch and per fabric configuration deployment history and built in per fabric, per device configuration compliance. Stay tuned for additional details….
Thank you Venu,
Interested so see further blogs on this topic specially about " customizable fabric policies based on python"and per switch, per fabric deployment features.
Creating a fabric for NX devices on single data center was always easy, however crafting multiple fabric across geographic region and still being able to push customized network & security policies and being able to visualize any consistency drift is most difficult state-of-the-art and kudos to DCNM v10 being able to nail this down using right set of architecture, design and implementation strategy. Excited to learn more on the fabric builder, consistency drift and security features….