14.2. SLAs for Core Azure Services
Understanding SLAs for Core Azure Services
Azure’s SLAs describe Microsoft’s commitment to providing uptime and connectivity. The specifics of an SLA can vary widely depending on the Azure service in question. Some common elements include the percentage of uptime guaranteed (typically represented in nines – for example, “three nines” represents 99.9% uptime), the performance targets for service responsiveness, and the network connectivity expectations.
SLAs for Azure Compute Services
Azure’s compute services are the backbone of its offering, and the SLAs here are particularly critical. They cover Virtual Machines (VMs), Azure Kubernetes Service (AKS), Azure Functions, and more.
- Virtual Machines (VMs): For single-instance VMs, Azure guarantees a connectivity rate of 99.9%, provided that the instance is deployed within an approved configuration. For multiple instances running in availability sets, the SLA improves to 99.95%, reflecting the reduced risk of simultaneous downtime across multiple instances.
- Azure Kubernetes Service (AKS): The SLA for AKS guarantees 99.95% uptime for the Kubernetes API server for clusters that use Azure Availability Zones. This does not cover the individual pods or containers running within the clusters, as Kubernetes’ design is meant to handle container failures.
- Azure Functions: For applications running on Azure Functions, the SLA guarantees 99.95% availability when running in premium plan configurations, reflecting the higher level of resilience against failures provided by premium features such as pre-warmed instances.
SLAs for Azure Storage Services
Storage is another critical area with a well-defined SLA, as data is a valuable asset for any organization.
- Azure Blob Storage: With Azure Blob Storage, the SLA guarantees at least 99.9% availability for the read accesses of GRS (Geo-Redundant Storage) and RA-GRS (Read Access-Geo-Redundant Storage) accounts, and at least 99.99% availability for all other storage accounts, which is critical for applications that require constant access to data.
- Azure File Storage: The Azure File Service SLA provides a 99.9% availability guarantee for SMB and REST protocols, ensuring that shared files are highly available for operations that require consistent file shares.
- Azure Disk Storage: For Azure Disk Storage, the SLA commits to connectivity of at least 99.9% for Locally Redundant Storage (LRS) and Zone Redundant Storage (ZRS) options, providing assurances for users who need persistent and high-performance disk storage.
SLAs for Azure Database Services
Databases are at the heart of many applications, making their SLAs critical to understand.
- Azure SQL Database: Microsoft offers a 99.99% SLA for Azure SQL Database, covering both single databases and elastic pools, which is indicative of the high level of confidence in the database service’s reliability.
2. Azure Cosmos DB: As a globally distributed database service, Azure Cosmos DB offers a comprehensive SLA that includes throughput, consistency, availability, and latency guarantees, with an impressive 99.999% availability for read and write operations.