Cloud cost requirement analysis

Cloud cost planning starts with gathering all cost requirements. Analyze these requirements during the planning phase to identify and track stakeholder business needs, then take targeted actions during implementation to optimize resource utilization and keep cloud costs manageable and sustainable.

Capture cost requirements

Enterprise customers gather cloud cost requirements from the following areas:

• Business geography

• Compliance

• Security

• Business continuity and stability

• Technical team management

• Automation and standardization

• Cost optimization goals

Analyze cost requirements

Business geography

Choosing cloud regions is a critical decision for global enterprises because it affects all aspects of cost. When you create or purchase Alibaba Cloud resources, you must select a deployment region, which cannot be changed later. Consider factors such as proximity to your data centers and end users. For example, if an Asian company has warehouses in its home country and in Shanghai, the Shanghai warehouse management system should be deployed in the Alibaba Cloud Shanghai region. Also consider the costs of leased lines and data transfer between regions.

Compliance

Enterprises face external compliance requirements, such as MLPS 2.0 and 3.0 regulations, for their cloud information systems. As cloud resources grow, enterprises also establish internal compliance baselines for management efficiency and security. These baselines include requirements for recording operation logs and managing resources.

Configuration change logs and the continuous monitoring and automatic alerting capabilities of the cloud platform are essential for self-regulated compliance. For enterprise customers, operation and configuration audits involve services such as Cloud Config, ActionTrail, Function Compute, SLS, and OSS. ActionTrail delivers operation logs to SLS or OSS in the corresponding region, and storage fees vary based on log volume.

Security

Security is a critical infrastructure requirement and a major cost factor. Security measures protect your organization's critical data from leaks and ensure business integrity and availability by defending systems against external attacks and abuse.

Include security requirements in your cloud cost analysis. Security spans data center security, network security, data security, identity security, and defense systems. Identity verification and multi-factor authentication (MFA) increase costs, as do cloud security groups, Cloud Firewall, Anti-DDoS, and application security products. Security monitoring and operations for complex environments are also significant cost considerations. Alibaba Cloud Security Center provides enterprise-level security monitoring and operations features.

Log storage and security audits are necessary to meet security and regulatory requirements.

Data backup is essential for meeting security requirements. To comply with data security regulations, an enterprise might need to increase backup frequency. Backup tool effectiveness, data storage capacity, and bandwidth all affect cost.

For more information about security considerations, see the Security and Compliance Pillar.

Business stability and continuity

Enterprise applications must meet specific service level agreements (SLAs) for both internal and external users. Create high availability and disaster recovery policies based on your business stability and continuity needs.

Your overall SLA, recovery time objective (RTO), and recovery point objective (RPO) shape your cloud infrastructure and product choices. For example, hosting an application across multiple regions costs more than a single region but provides higher availability. You might also choose a database product that supports multi-zone deployment.

Generally, if the cost of high availability exceeds the cost of application downtime, you may have over-engineered your high availability policy. Conversely, if the cost of high availability is less than the cost of acceptable downtime, you may need to invest more.

If the cost of downtime is relatively low, you can save money using backup and restore for disaster recovery. If the cost of downtime per hour is high, you should invest more in high availability and disaster recovery. You need to balance service provisioning, availability requirements, and your organization's risk tolerance.

Choosing the right instances based on application availability, lifecycle, and service characteristics can significantly improve utilization and reduce costs. For example:

1. Instance selection: Understand the key features of different instance types. Choosing a suitable ECS instance type can improve resource utilization and save costs. You need to balance performance, price, and workload to make the best decision for cost-effectiveness and stability. Alibaba Cloud Elastic Compute Service (ECS) provides a variety of instance families categorized by business scenario and configurations, such as vCPU, memory, network performance, and storage throughput. You can start with low-configuration resources and then upgrade the configuration after observing and evaluating the workload. You can also downgrade or release underutilized resources.

2. Billing method selection: Choosing the right billing method for your business is a direct way to optimize costs. For long-term, stable workloads, use savings plans for lower costs. For stateful resources with a clear lifecycle, use the more flexible pay-as-you-go billing method. Pay-as-you-go is suitable for temporary tests and Auto Scaling. For pay-as-you-go resources, you can enable the 'no-charge when stopped' feature to retain data and allow for quick restarts. To run the resource long-term, switch to a subscription billing method. Stateless and fault-tolerant workloads can use very low-cost spot instances. For intermittent or event-driven computing needs, use Alibaba Cloud Function Compute. For example, you can trigger a one-time computing task when a data file in storage changes. This eliminates the need to purchase and manage servers or other infrastructure. Function Compute provides computing resources that run tasks elastically and reliably. It also offers features such as log query, performance monitoring, and alerting.

3. Design an elastic architecture: When designing your architecture, create an elastic structure based on the application lifecycle. Use the elastic scaling of computing resources to add deployment nodes during peak business hours and scale down cloud resources during off-peak hours. This approach significantly improves resource utilization and maximizes cost-effectiveness.

4. Survey the near-future business plans of various departments. Then, list the appropriate resource specifications and usage amounts, leaving some room for redundancy.

Technical team management

Enable your technical staff and partners to learn and use Alibaba Cloud effectively. Consider Alibaba Cloud's enterprise ticket service, enterprise support plans, expert services, and Alibaba Cloud University learning resources.

1. Ticket service: Provides 24/7 online support from experts who can answer questions from your technical staff.

2. Enterprise support plans: In addition to basic after-sales support, customers with complex business systems or higher service requirements can choose from various enterprise support plans. These plans provide benefits such as rapid ticket response, dedicated technical support channels, and a Technical Account Manager (TAM).

3. Enterprise value-added services: Provides professional services for the entire lifecycle to meet various enterprise needs. These services include solution consulting before cloud adoption, migration and deployment during cloud adoption, and operations, management, optimization, and improvement after cloud adoption.

4. Professional training solutions for customers: This solution addresses the digital transformation needs of enterprises and the demand for application-oriented talent from universities by providing professional cooperation plans and training systems. Enterprise training relies on Alibaba Cloud's industry digitalization practices and its exploration of cutting-edge technologies, such as cloud computing, big data, artificial intelligence, and cloud-native. It builds a professional talent development system and service. This training empowers digital transformation, connects with the Alibaba ecosystem, overcomes development bottlenecks, and helps enterprises achieve rapid growth.

Automation and standardization

Define provisioning and deployment processes using standardized procedures and tools, covering account management, cloud resource management, access settings, event response, and disaster recovery. Standardized processes can reduce cloud costs, though implementation may require additional development investment.

For example, you can use Cloud Shell and Terraform to define your IT infrastructure:

• Automated infrastructure management: Terraform can create templates of configuration files. These templates define, provision, and configure ECS resources in a repeatable and predictable way, which reduces deployment and management errors caused by human factors. You can use the same template to create identical development, testing, and production environments.

• Infrastructure as Code (IaC): You can use code to manage and maintain resources. This approach lets you save the state of your infrastructure. You can then track changes to different components in the system and share these configurations with others.

• Reduced development costs: You can reduce costs by quickly creating portable development and deployment environments on demand.

You can also use the visual application resource creation tool, Cloud Architect Design Tools (CADT), to quickly generate your cloud architecture. CADT provides a rich set of pre-built application architecture templates and supports a drag-and-drop interface to define your application's cloud architecture. This significantly reduces the difficulty and time required to manage applications on the cloud.

Cost optimization goals

At different stages of development, the cost accounting department sets different cloud usage requirements, such as annual cost constraints. Cloud resource usage may grow significantly, but costs should grow more slowly, requiring nonlinear growth relative to usage. This means the cloud vendor must reduce costs through technological iteration and increased computing power, while the enterprise must also evolve its IT systems. A cost optimization goal might be to reduce cloud resource usage by 10% while maintaining existing business operations. These goals must be identified during the planning phase and implemented during execution.

During planning, follow the latest cloud computing trends to take advantage of cost optimizations from new technology. As cloud computing evolves, new technologies, products, and tools bring performance and efficiency improvements that enhance cost-effectiveness. For example, server and database upgrades or new R&D efficiency platforms can help enterprises adopt new architectures and improve cloud efficiency.

Best practices

The following are recommended practices for requirement analysis:

1. To align with business goals, stakeholders in the enterprise must define the requirements.

   • Good practice: The enterprise should define its IT requirements internally. It should establish a high-level design for requirement collection and tracking and provide feedback on implementation results. This creates a complete and effective long-term process for gathering requirements.

   • Bad practice: Do not completely rely on vendors or allow them to define and manage cloud resources without providing guidance and constraints. This approach will make subsequent cloud governance work extremely complex.

2. The enterprise must collect and implement security requirements during IT implementation.

   • Good practice: Analyze the security needs of the business and the enterprise when planning your business requirements.

   • Bad practice: Do not ignore security requirements or deploy security products only after the IT system is deployed. This approach fails to manage security costs from the beginning. Before security products are deployed, business security and stability can be compromised, which can cause greater losses to the enterprise.

3. The enterprise should create high availability and disaster recovery policies based on service levels.

4. The enterprise should invest in automation and standardization. However, it should also consider the actual situation and business needs to avoid over-implementation.