Business Requirements

ID Description Stakeholder Notes Approval

BR01 Provide IaaS and DRaaS Service Offerings for Enterprises only fully manageable by the customer.
BR02 Eleminate any manual processes of provisioning and operation to reduce management overhead.
BR03 Provide end customers maximum flexibility in VFE Service Offerings without fixed T-Shirt sizes Commercial Team
BR04 Improve current availability and recoverability SLAs of VFE Public Cloud
BR05 Provide scalable platform to facilitate future expansions as needed
 Technical Requirements
Design Quality ID Description Stakeholder Notes Approval

Availability A01 Support customer availability SLA of 99.9% per year

This is to be discussed in details in
Commercial Team next phase.
It can be done with additional sites
and workloads spread against 2 or
A02 Design must be flexible to support future customer availability SLA of 99.99% more sites
Recoverability R01 Support RTO of 3 HRs
Commercial Team, Cloud Team RPO Policy to be determined based on
R02 Support Data Retention Policy for Mgmt Components backups of 30 Days
and Infrastructure Team DataDomain Sizing
R03 Support Data Retention Policy for Customer VMs backups, who is shutting down their contracts with VFE, of 1 year

Performance P01 Control ingress BW on eaach Tenant Edge.

P02 Allow selected customers to access MPLS networks
P03 Automate the access of selected customers to their MPLS Networks
P04 Support 1250 VM based on on Tshirt Size of 1 CPU, 4GB RAM and 50GB Storage and overcommitmet ratio of (3:1 CPU and Memory and 4:1 Storage)
P05 Support VFE proposed Sevrice Offerings Ser01 and Ser02
Manageability M01 Automate required basic security rules (RDP Access Inbound, SSH Access Inbound, HTTP Access Inbound, HTTPS Access Inbound, DNS/NTP Outbound) Security Team
M02 Automate required basic network configuration for each tenant Networking Team
M03 VFE Cloud must support BYOL model and managed-licenses model for the following OSs (Windows/RedHat/Ubunto/CentOS)
M04 Provide self-managed experience for the advanced security features (LB, SSL VPN, L2VPN, IPSEC VPN, etc.)
M05 VFE Cloud should provide customers the ability to upload their own templates and OVFs
Commercial Team and Cloud
M06 Configure threshold based on owned VMs number driven by VFE business case (20VMs) to alert sales team for a selling opportunity
Team
M07 Configure threshold, based on oversubscription ratio 3:1 on CPU Usage, Memory Usage and number of VMs total, of (33%) to alert VFE team for additional capacity requirements
M08 Configure threshold, based on oversubscription ratio 4:1 on Storage Usage, of (25%) to alert VFE team for additional capacity requirements
M09 Provide the customer a show back of resources usage without showing metering policies
Security Sec01 VFE Cloud must be GDPR compliant as per VF Group Policy Security Team to provide the required
Commercial Team
Sec02 VFE Cloud must be ISO27001 compliant list of actions
Security Team
Sec03 Use dedicated AD Domain for users authentication.
Scalability Sca01 9K VM in 3 years from now based on Tshirt Size of 1 CPU, 4 GB RAM and 50 GB Storage overcommitment ratios (3:1 CPU, 3:1 RAM, 4:1 Storage)
Commercial Team
Sca02 Educate Cloud Operation team about how to scale and shrink the environement to be Business-as-usual (BAU) tasks
Service Offerings
ID Description Options Stakeholder Notes Approval
Custom CPU/RAM/Storage Allocation
To be discussed (Static/Dynamic Routing
self-managed capabilities) during the
Different Networking and Security Services (DFW, Edge Firewall, IPSEC VPN, L2VPN, SSL VPN, LB, Static/Dynamic Routing**) design sessions
Storage Tiering** Pending on final design of Dell EMC
No. of VMs (with limit to prevent sprawl)
Ser01 Virtual Private Cloud (VPC) Level of Support Bundle (limited by number of calls, response SLA) Commercial Team
Public IPs Allocation
OS Licensing Model
Not avialable currently, but can be
considered as possible future offereing.
Availability Tiering ** Design will not be able to meet right now
Performance SLA To be discussed in design sessions
Fixed T-shirt sizes
No access but to VM Console
Levels of Support Bundle (limited by number of calls, response SLA) VFE wants this offer to depricate
Ser02 VM Offering Commercial Team
Public IPs Allocation eventually.
Upgrade/ Downgrade VM Size
Manage only power operations on VMs
 Design Constraints
ID Description Notes Approval
C01 Use New Dell Servers for the Compute Clusters
C02 Use new Dell EMC Unity Array for the workload domain storage with single tier (Hybrid Storage).
C03 Billing cycle is in advance with fixed pricing list based on service
C04 Use old Dell EMC Unity Array for the management domain storage with single tier (Hybrid Storage).
C05 Use 2 HP blades and 2 Huwawei Blades for Mgmt Cluster
C06 Use Avamar as Backup solution for management workloads and compute workloads
C07 CPU, Memory, Storage and Backup Destination Capacity is pre-defined
C08 Pre-defined Servers configuration for Mgmt and Compute Clusters
 Design Risks
ID Description Notes Approval
RS01 VFE Team has no knowledge of vCloud Director Platform
RS02 VFE Team has basic knowledge of NSX
RS03 Single rack for compute cluster. Failure of this rack will take whole cluster down and might also take down the storgae array as well
RS04 Single storage array for Mgmt Cluster workloads. Failure in this array will take the whole management layer down
RS05 Single array for cloud workloads. Failure in this array will take the whole cloud down
RS06 Single chassis for Huawei blades. Failure in this chassis will take half capacity of Mgmt Cluster down
RS07 Single site containing the whole environment
RS08 Migration from old HP management blades to new Huawei might be required which is out of scope
RS09 No Edge Cluster available and external networks to be configured on all 8 hosts, which will be a big challenge in case edges will be configured with public IPs directly
RS10 New VCAV architecture does not repliucate customer network configuration from on premises to VFE Cloud outn of the box. They must be created manually
RS11 VCAV Metering for Replication BW is not avialable out of the box.
RS12 Based on available memory resources and at maximum full scale of 1250 VMs, in case of single host failure, memory utilization will reach around 93%.
RS13 Offered CPU physical cores counts do not support the assumed 3:1 CPU overcommitment ratio Calculations need to be revised based on VFE owned
RS14 Offered Storage capacity will be 100% utilized wihtout any safety buffers/slack space at maximum full scale of 1250 VMs harwdare BOM
RS15 Long hardware purchasing cycle of VFE which can easily reach 6 months
Dell EMC to be consulted about assumed RPO and
RS16 Limited size of DataDomain (Backup Destination) of 178 TB Usable Data Retention Policy
Security team promised to provide them by next
RS17 Security team did not provide detailed security guidelines/requirements to achieve the required security compliances week.
This will effect the deadline of the going live. To be
confirmed by the end of this week. BOM of
RS18 Huawei Blades to be racked and stacked by end of March hardware to be delivered this week also.
 Design Assumptions
ID Description Stakeholder Notes Validation Approval

Validated by Commercial and

A01 DC infrastructure (Powering, Cooling, Internet Lines) availability SLA is 99.99% per year Infrastructure teams
A02 Recoverability SLA is assumed based on two failure incidents per year only
A03 Availability SLA is not considering rare cases of whole Rack or Site failure

4:1 CPU, 3:1 Mmeory, 3:1 Storage, not

For hardware calculations, CPU overcommitment, Memory overcommitment and Storage taking into consideration any HA or slack
A04 overcommitment ratios are assumed to be 3:1, 3:1 and 4:1 respectively space as per Hassanien's Statement.
A05 For hardware calculations, assumed VM T-shirt size of 1 CPU, 4 GB Memory and 50 GB Storage