File Storage NAS (NAS) is a cloud service that provides a file storage solution for compute nodes. The compute nodes include Elastic Compute Service (ECS) instances, Elastic High-Performance Computing (E-HPC) instances, and Container Service for Kubernetes (ACK) clusters. NAS is a distributed file system that supports the Network File System (NFS) and Server Message Block (SMB) protocols. NAS provides the following features: shared access, elastic scalability, high reliability, and high performance.
NAS provides the following file storage solutions: Extreme NAS, General-purpose Performance NAS, and General-purpose Capacity NAS.
Specification | Description |
Extreme NAS | Extreme NAS is a high-performance file sharing solution that is built based on the latest generation of network architecture and all-flash storage. The maximum capacity is 256 TiB. The bandwidth ranges from 150 MB/s to 1,200 MB/s. The latency is approximately 100 microseconds. Extreme NAS is suitable for latency-sensitive business in which a large number of small files are handled. |
General-purpose Performance NAS | General-purpose Performance NAS uses solid-state drives (SSDs) as storage devices, and provides high throughput, high input/output operations per second (IOPS), and low latency for workloads. General-purpose Performance NAS is a file sharing solution that is suitable if you require file storage of high throughput, high concurrency, business scalability, and low read latency. General-purpose Performance NAS is suitable if you need to perform frequent read/write operations and have high requirements for response latency. |
General-purpose Capacity NAS | General-purpose Capacity NAS uses SATA hard disk drives (SATA HDDs) as storage devices and provides high-performance storage space at low costs. General-purpose Capacity NAS is a file sharing solution that is suitable for cost-sensitive scenarios in which high throughput, high concurrency, and business scalability are required. General-purpose Capacity NAS is more cost-effective if you do not need to perform frequent read/write operations and do not have high requirements on response latency. |
Scenarios
NAS is suitable for the following scenarios:
Container storage
You can use containers to build microservices. Containers support fast pre-configuration, flexible resource allocation, and process isolation. If some containers must access raw data each time the containers start, you must create a shared file system for the containers. This way, the containers can access the file system regardless of the instance on which the containers run. You can use NAS as container storage because NAS provides persistent shared access to files.
Content management and web services
NAS provides high persistence and high throughput. You can use NAS in content management systems and web servers to store and provide data for websites, online publishing applications, and archiving applications. NAS follows the expected file system semantics, file naming conventions, and permissions that are preferred by web developers. You can integrate NAS with web applications and use NAS in websites, online publishing applications, and archiving applications.
Enterprise applications
NAS provides high scalability, elasticity, availability, and persistence. You can use NAS as storage solutions for your enterprise applications and applications delivered as services (ADaaS). NAS provides standard file system interfaces and semantics that allow you to migrate your enterprise applications to Alibaba Cloud or construct new applications.
Media and entertainment workflows
You can use NAS to share and process large files in media workflows, such as video editing, audio and video production, broadcast processing, and audio design and rendering. NAS provides powerful data consistency models, high throughput, and shared access to files. This reduces the time required to complete the preceding workflows and merges multiple on-premises file repositories into a single repository that can be accessed by all users.
Big data analysis
NAS provides high throughput for computing nodes, read and write consistency, and low latency to meet the scale and performance requirements of big data applications. Most analysis workloads call file system API operations to access data or write data to files based on file semantics, such as file locking. In this case, you can use NAS that supports file system semantics such as file locking and provides scalable capacity and performance.
Performance
The peak throughput of a file system is linearly proportional to the used capacity of the file system. A file system with larger capacity has higher peak throughput. NAS can be concurrently accessed and randomly read or written by thousands of ECS instances by using Portable Operating System Interface (POSIX).
Specification | Capacity | Average latency for reading 4 KiB files | IOPS |
Extreme NAS | 256 TiB | Approximately 100 microseconds | 10000~200000 |
General-purpose Performance NAS | 1 PiB | Milliseconds | Up to 30,000 (4K random read/write) |
General-purpose Capacity NAS | 10 PiB | Approximately 10 milliseconds | Up to 15,000 (4K random read/write) |
API operations
You can use POSIX to perform operations on the data of NAS file systems, such as read or write operations. You can migrate local applications to cloud without the need to modify the application code.
You can make API requests over HTTP or HTTPS to manage NAS resources. The GET and POST methods are supported. If you are familiar with network protocols and one or more programming languages, we recommend that you call API operations to manage NAS resources. You can use NAS SDKs, Alibaba Cloud CLI, or OpenAPI Explorer to call NAS API operations to manage NAS resources, such as file systems, mount targets, permission groups, snapshots, and tags. For example, you can create, delete, query, or modify NAS resources. For more information, see List of operations by function. If you prefer a graphical user interface (GUI), you can perform the preceding operations in the NAS console.
Scalability and elasticity
Compared with traditional storage systems, NAS eliminates the need to perform complex operations, such as planning, purchasing, partitioning, and monitoring. The capacity of a NAS file system can automatically scale in or out when you delete files from or add files to the file system. This way, NAS allocates storage resources based on your business requirements without impacts on your applications.
Data durability and service availability
NAS provides multiple replicas for each piece of data that is stored in a file system. These replicas reside on devices that are isolated across different fault domains for geo-redundancy. NAS provides data reliability of 99.999999999% (eleven 9's). This reduces data security risks.
Security
Permission group
NAS uses a permission group as a whitelist to configure the access control for a file system, including the authorized IP addresses, read and write permissions, and user permissions. You can add rules to a permission group to allow access to a file system from specific IP addresses or CIDR blocks. You can also grant different levels of permissions on different IP addresses or CIDR blocks. For more information, see Manage a permission group.
RAM
You can use Resource Access Management (RAM) to manage the users of NAS and control access to NAS resources. RAM implements access control based on users. RAM allows you to create and manage multiple RAM users within an Alibaba Cloud account and grant different permissions to each RAM user. This way, you can authorize different RAM users to access different Alibaba Cloud resources. RAM allows each user to use an exclusive AccessKey pair. You can assign the least permissions to each RAM user to reduce data security risks for your enterprise. For more information, see Perform access control based on RAM policies.
ACL
You can use access control lists (ACLs) to control access to files and directories. ACL implements access control based on resources. Access control and user management are important for enterprise users who need to share files among different users and groups by using a shared file system. NAS provides the ACL feature that allows you to grant users and groups different access permissions on directories and files. For more information, see Overview and Overview of the SMB ACL feature.
Encryption
NAS uses the 256-bit advanced encryption standard (AES-256) to encrypt static data that is stored in file systems and uses Key Management Service (KMS) to manage encryption keys. NAS automatically encrypts data before the data is written to file systems and decrypts data before the data is read by applications. NAS automatically encrypts and decrypts data. Therefore, you do not need to modify your application code when you encrypt or decrypt data. For more information, see Server-side encryption.
Billing method
The capacity of a NAS file system automatically scales in or out based on your business requirements. Therefore, you do not need to partition the file system in advance. You are billed for the usage of NAS based on the pay-as-you-go billing method. You are charged only for the storage usage that is used. If a file is deleted from a file system, you are no longer charged for the storage usage of the file. You can also purchase resource plans to offset your subsequent storage usage fees. In most cases, resource plans are more cost-effective.