For the enterprise, storage is a critical feature of cloud computing. The ability to remotely store data, and then access that data anywhere, provides greater flexibility and helps reduce or eliminate on-premises storage costs. Top cloud providers, such as Amazon Web Services, Google and Microsoft Azure, offer their own storage options to meet specific IT needs.
The Azure public cloud offers five main storage options: blob, queue, file, table and disk or premium. Each option offers high availability and elasticity. Azure storage types continue to expand, including its recent addition of Azure Cool Blob Storage to its cool access tier. To meet the varying needs of different enterprises, Azure storage supports a variety of operation systems, including Windows and Linux, as well as different programming languages, such as .NET, Java, Node.js and Python.
To get better acquainted with the various Azure storage types, know these important terms:
Azure Blob storage: Also known as object storage, Azure Blob storage stores unstructured data in the cloud as blobs. The service can house large amounts of data that can be accessed from anywhere using HTTP or HTTPS. There are three Azure storage types within this category: page blobs, which are meant for random read and write operations; block blobs, which are designed to support large blobs; and append blobs, which are similar to block blobs, but designed for append operations, such as logging. Azure blob storage comes in two tiers: the hot storage tier is frequently accessed data and cool storage is for infrequently accessed data.
Azure Queue storage: Azure Queue storage is a service for storing high volumes of messages between application components, whether on the cloud, on premises or on mobile devices. In the event of a component failure, the service buffers requests using queue storage until the component can be processed again. The Queue option helps prevent peak traffic bursts from negatively affecting application performance. Azure Queue storage can support messages up to 64 KB.
Azure Table storage: Azure Table storage is a NoSQL data store for semi-structured data. The term table is defined as a group of entities and, the schemaless design of Azure Table storage allows for flexibility as application needs evolve. Each entity has key pairs for identification purposes and an automatic timestamp. As demand increases, tables scale and maintain high availability.
Azure File storage: Azure File storage is a distributed file sharing service that uses the standard SMB 2.1 or SMB 3.0 protocol. Organizations can use the service to migrate on-premises file- or file share-based applications to Azure without having to modify the code. Other uses for File storage include storing shared application settings, as well as diagnostic data, such as logs.
Azure Premium storage: Azure Premium storage is an SSD-based service for I/O-heavy applications. The service provides high throughput and low-latency disk support. It features 64 terabytes of storage and 80,000 IOPS per VM with 2,000 megabytes per second disk throughput per VM. There are three Premium disk sizes to choose from: P10, P20 and P30.
Analyst Marc Staimer explains why the best object storage use cases are those that require high capacities and low costs.
Azure Data Lake Store: Azure Data Lake Store is a hyperscale storage service, based on Apache Hadoop, for large amounts of data. The service can store structured and unstructured data in its native format. There are no fixed limits for file or account size. For big data analysis, the storage service is integrated with Azure Data Analytics and Azure HDInsight.
Azure StorSimple: Azure StorSimple is a hybrid data management service that facilitates storage tasks between on-premises and Azure cloud storage. Data is organized into three storage tiers -- SSD, HDD or the cloud -- based on its usage, age and relationship to other data. For example, less active and inactive data is automatically archived to the cloud to free up on-premises storage capacity for active data.
Azure Site Recovery: Site Recovery is unlike other Azure storage types in that it's focused on replication and failover. The service automates data and application replication from on-premises environments to the Azure cloud, offering an alternative to replicating workloads to a secondary data center. To ensure workloads are replicating correctly, failover testing is available. Users can customize their failover plan to support multi-tier applications over multiple VMs. Azure Site Recovery can replicate any workloads supported on Hyper-V, VMware VMs and Windows or Linux servers.
Get acquainted with Azure big data services
Test your Azure cloud knowledge
Improve your Azure management skills