3. Service Offering Examples
   
3. Service Offering Examples
Service offerings and their inherent virtual datacenter constructs provide effective means of creating service differentiation within a broader vCloud service landscape. They deliver consistent service levels that invariably align with unique business use case requirements, as presented by individual tenants in either a private or public vCloud setting. The service offerings presented in this section serve as a reference for building a differentiated IaaS service model. They also try to address the full spectrum of enterprise workload requirements observed in the vCloud services market today.
The following is a summary of these service offerings:
*Basic – Based on the Pay As You Go allocation model. This service offering lends itself to quick-start pilot projects or test and development application workloads that typically do not require persistent resource commitments or upfront resource reservations.
*Committed – Based on the Allocation Pool allocation model. This service offering provides consumers with a minimum initial commitment of resource capacity plus the added ability to burst above that minimum if additional infrastructure capacity is available at the time of need. The level of minimum commitment, expressed as a percentage of overall capacity per resource type, provides an extra layer of assurance to consumers who seek deterministic performance levels for their application workloads.
*Dedicated – Based on the Reservation Pool allocation model. This service offering provides consumers reserved resource capacity up front, fully dedicated by individual tenant. The level of resource guarantee (always set to 100%) provides customers with a higher degree of service assurance than the Committed service offering, plus additional layers of security and resource control for their application workloads.
Due to often unpredictable business demands and the elastic nature of vCloud service consumption models, it is not unreasonable for providers of private or public vCloud instances to seed a service environment with a single service offering type and adapt that service over time, given proper business justification. This approach is not only common practice, but also recommended, regardless of the number of service offering examples made available for consideration.
To help decide which service offering makes the most sense for a particular set of business use cases, refer to the key service attributes summarized in the following table. Additional details and reference examples for each service offering can be found in the following sections.
Table 11. Service Offering Matrix Example
 
Basic
Service Offering
Committed
Service Offering
Dedicated
Service Offering
Allocation Model
Pay As You Go
Allocation Pool
Reservation Pool
Control Plane (Management)
Shared, multitenant
Shared, multitenant
Shared, multitenant
Cluster Resources
Shared, multitenant
Shared, multitenant
Dedicated, single-tenant
Unit of Consumption
vApp
Aggregate resource capacity allocated
Aggregate resource capacity reserved
Resource Allocation Settings
(per Organization Virtual Datacenter)
N/A
*CPU (GHz)
*Memory (GB)
*Storage (GB)
*CPU (GHz)
*Memory (GB)
*Storage (GB)
Resource Guarantee Settings
*% of CPU
*vCPU speed
*% of memory
*% of storage
*% of CPU
*vCPU speed
*% of memory
*% of storage
*100% of CPU
*100% of memory
*100% of storage
Limits
(per Organization Virtual Datacenter)
Maximum number of virtual machines
CPU (GHz)
Memory (GB)
Maximum number of virtual machines
CPU (GHz)
Memory (GB)
Maximum number of virtual machines
CPU (GHz)
Memory (GB)
Reporting/Billing Frequency
Per use
Monthly
Monthly or Annual
Metering Frequency
Hourly
Hourly
Hourly
Service Availability
99.95%
99.99%
99.99%
Target Workloads
Test and Development
Tier 2 and Tier 3 Production
Tier 1 Production
Application Workload Examples
*Short-term or bursty workloads.
*QA testing.
*Integration testing.
*New software version testing.
*Short term data analytics.
*Static web content servers.
*Lightly used app servers.
*Active Directory servers.
*Infrastructure servers (DNS, print, file).
*Small/medium database servers.
*Short term content collaboration.
*Staging sites.
*Exchange and SharePoint servers.
*Large database servers (high IOPS).
*PCI-related servers.
*HPC workloads.
*SaaS production applications.
*CRM, EDA, ERP, and SCM applications.
*Financial applications
(high-compliance).