Federated multi-controller management dashboard for Arista DMF and CCF deployments across multi-pod, multi-site environments
Arista Multi-Cloud Director (MCD) is a federated management platform that presents a single operational dashboard across multiple independent Arista DANZ Monitoring Fabric (DMF) and Converged Cloud Fabric (CCF) controller instances deployed across different data centre pods, sites, or cloud regions. In large-scale deployments where a single DMF or CCF controller instance is scoped to one pod or one data centre, and the organisation runs dozens of such pods across multiple sites, MCD provides the cross-site visibility and configuration consistency layer that individual controller instances cannot offer. Rather than requiring an operator to log into each controller separately to verify configuration compliance, push a global NTP or syslog policy, or check the state of a specific device across the estate, MCD presents all instances through one interface with real-time data pulled from each controller.
MCD deploys as a virtual machine appliance requiring only 4 vCPU, 6 GB RAM, and 400 GB of thinly provisioned storage, making it a lightweight addition to an existing data centre virtualisation estate. Supported hypervisors include VMware ESXi 6.5 or higher, KVM on Red Hat Enterprise Linux 7.5 or CentOS 7.6, and Nutanix AHV 5.18.1.2. The platform scales to 50 DMF instances, 50 CCF instances, and 25 Analytics Node (AN) clusters, with a combined ceiling of 65 total managed instances, sufficient for large enterprise and service provider deployments that operate dozens of monitoring fabric pods across geographically distributed sites. Template-based configuration workflows allow global policies (AAA authentication, NTP servers, syslog destinations, sFlow collectors) to be defined once in MCD and deployed consistently to all instances, with compliance checks that identify any instance deviating from the defined global standard.
MCD aggregates live data from up to 65 total DMF, CCF, and Analytics Node instances simultaneously, presenting each instance's status, configuration, and alert state in a single interface without requiring the operator to authenticate to each controller separately. For a service provider or large enterprise operating monitoring fabric pods across 10 or 20 data centre sites, this federation means that a cross-site event, such as an NTP drift affecting controllers at multiple locations, is visible as a correlated finding in MCD rather than as separate alerts on 20 independent controller dashboards that must each be checked individually.
Global configuration templates defined in MCD (covering AAA authentication, NTP server lists, syslog destinations, and sFlow collectors) can be deployed to all managed instances simultaneously and subsequently verified for compliance on a scheduled or on-demand basis. Compliance checks compare each instance's running configuration against the MCD-defined global standard and report any deviations: an instance that has drifted from the approved NTP configuration due to a local change, or one that has a syslog destination pointing to a decommissioned collector, is flagged without requiring the operator to manually audit each controller. This shift from manual auditing to automated compliance monitoring is the operational value that makes MCD justified for deployments beyond a handful of controller instances.
Global search across all managed instances allows an operator to locate a specific device, flow policy, or alarm across the entire multi-site DMF or CCF estate from the MCD interface, without knowing which specific controller instance the device is registered to. For incident response scenarios, such as a security analyst searching for all monitoring fabric policies that match a specific source IP or VLAN, global search retrieves matching results from all 65 managed instances simultaneously and presents them in a unified result set, rather than requiring sequential searches on each controller. Real-time data display means the results reflect the current state of each instance at the time of the query rather than a cached snapshot.
At 4 vCPU, 6 GB RAM, and 400 GB thin-provisioned storage, MCD consumes fewer resources than almost any production application VM, making it practical to deploy on an existing server estate without dedicating new hardware. VMware ESXi, KVM on Red Hat and CentOS, and Nutanix AHV are all supported without additional licensing or hypervisor-specific configuration, meaning MCD can be placed on the same virtualisation platform already used for DMF and CCF controller VMs without requiring a separate hypervisor deployment. Thin provisioning on the 400 GB vHDD means MCD's actual disk consumption at deployment is minimal and grows only as configuration and audit history accumulates over time.
Full specifications for Arista Multi-Cloud Director
Download product documentation and resources
Explore other configurations and models that might suit your needs.
Recommended#Arista-CLOUD-TEST
Recommended#Arista-CLOUDVISION-CUE
Recommended#Arista-DANZ-MONITORING-FABRIC
Our team of experts is ready to help you find the perfect solution for your business needs. Get personalized advice and competitive quotes.
We're here to help with any questions