Unit 3: SLA management and securing cloud services

A service level agreement (SLA) is a contract between provider and customer stating the services, quality levels, measurements and penalties. A service level objective (SLO) is one measurable target inside it, such as 99.9 percent uptime. Inspiration: customers need guarantees about availability and performance, and providers need a clear, measurable promise. In the cloud, many customers share one platform, so SLAs must be defined, monitored and enforced automatically. Traditional approaches to SLO management were manual and fixed: SLOs were set once for a known system, monitored by people with periodic reports, and violations were found after the fact. They do not suit the cloud, where resources and demand change every minute.

SLA management in the cloud means defining, monitoring, enforcing and reporting SLAs for many tenants at scale. The provider measures metrics such as availability and response time continuously, compares them with the SLOs, raises alerts before breaches and calculates penalties. Automated policy-based management expresses rules as policies that the system enforces by itself. For example: if average response time exceeds 300 ms for five minutes, add two servers; if CPU stays below 20 percent for an hour, remove one. This keeps the SLOs met without human action and keeps cost under control.

Security in the cloud is a shared responsibility: the provider secures the facilities, hardware and virtualization layer; the customer secures their data, accounts, access settings and applications. Main concerns: unauthorised access and data breaches, weak identity and access management, insecure interfaces, data loss, multi-tenant isolation failures, insider threats and compliance violations.

• Brokered cloud storage access: a broker sits between the user and the storage. The client asks the broker for access, the broker checks the request against policy and hands out a short-lived credential, and the data is fetched through a proxy so the customer never reaches the storage directly. This limits damage if a client is compromised.
• Storage location and tenancy: know where data is stored, because laws differ by country and some require data to stay in India. Tenancy means many customers on shared storage; choose providers with strong isolation, or dedicated storage for very sensitive data.
• Encryption: convert data into unreadable form using a key. Encrypt data at rest (in storage) and data in transit (over the network, using TLS). Keep the keys separate from the data, ideally under the customer's control.
• Auditing and compliance: keep logs of who accessed what and when, review them regularly, and follow standards and laws such as ISO 27001, SOC 2 and local data protection rules. Audits give evidence that the service is secure.