In the third part of SDDC Design (based on VMware Validated Design Reference Architecture Guide) we will review about one of the major steps on SDDC Design about physical design and availability. Before any steps on data-center physical design, we should consider an important classification based on availability aspect: Regions & Zones. Multiple Availability Zones form a Region, but what is the A-Zone?
Unfortunately many disastrous events like earthquakes, massive floods, and large power fluctuations may cause of interruption of IT communication as the service failures or unavailability of network components. So you need to segregate the DC total infrastructure into the regions & zones. A zone of SDDC, now mentioned as A-Zone is an independent area of infrastructure that is isolated as a physical distinct. A-Zones will improve SLA and redundancy factor and must be highly reliable because controlling of network infrastructure failure boundaries is the main reason of their presence. Interruptions may have internal causes, such as power outage, cooling problems and generators failure so each one of the zones should have their own safety teams (HSE and Fire departments).
There is two main factors to distinguish the differences of A-Zone & Region: distance of two site (Primary/Recovery) and network bandwidth of fiber connections between them. Basically A-Zones have metro distances (less than 50km/30mile) that usually connected with dark fiber to each other and there must be low latency as a single-digit and high network bandwidth. So they can act as Active-Active or Active-Passive sites for each other. For more than that mentioned distance range it’s highly recommended to put each A-Zone to different Regions but related workloads must be spread across multiple A-Zones belongs to same Region.
SDDC Business Continuity can be improved by operating many technologies and replication techniques such as:
- VMware vSphere FT for the VM-level Replication.
- VMware vSphere HA to provide VM availability at Host and Cluster-level.
- VMware vSphere DRS to act as a VM distributer to prevent load/VM aggregation on Host of a cluster.
- VMware vSAN as the Software-Defined Storage solution for better availability on environments without physical storage system.
- VMware Replication as an integrated appliance-based replication solution for inside or outside of the site or zone.
- Storage-vendor Replication solutions as the third-party solutions replication such as DELL EMC RecoverPoint, NetApp SnapMirror and HPE 3PAR.
- Software Replication solution such as Zerto Virtual Replication and Veeam Backup & Replication.
- VMware SRM as one of the best options for site replication & recovery solution.
Link of post on my personal blog: Undercity of Virtualization: VMware SDDC Design Considerations - PART Three: SDDC Availability