Disks are block-level data storage products provided by Alibaba Cloud for ECS and feature low latency, high performance, high durability, and high reliability. Disks use a distributed triplicate mechanism to ensure 99.9999999% data durability for ECS instances. If service disruptions occur (for example, due to hardware failures) within a zone, data within that zone is copied to an available disk in another zone to help ensure data availability.
Disk categories
- Enhanced SSDs (ESSDs): ESSDs are based on the next-generation distributed block storage
architecture and the 25 Gigabit Ethernet (25 GE) and remote direct memory access (RDMA)
technologies. Each ESSD can deliver up to one million input/output operations per
second (IOPS) and has low latency. For more information, see Enhanced SSDs.
We recommend that you use ESSDs for scenarios such as online transactional processing (OLTP) databases, NoSQL databases, and Elasticsearch, Logstash, and Kibana (ELK) log analysis.
- Standard SSDs: Standard SSDs are high-performance disks that feature consistent and
high random IOPS and high data reliability.
We recommend that you use standard SSDs for scenarios such as I/O-intensive applications, small and medium-sized relational databases, and NoSQL databases.
- Ultra disks: Ultra disks are cost-effective and have medium random IOPS and high data
reliability.
We recommend that you use ultra disks for scenarios such as development and testing business and system disks.
- Basic disks: Basic disks are the previous generation of disks and are unavailable for purchase.
For more information about disk performance, see EBS performance.
- System disks contain operating systems and can only be created along with instances. The lifecycle of a system disk is the same as that of the instance to which it is attached.
- Data disks are used to store application data and can be created separately or along with instances.
Limits
A disk can be attached to only one ECS instance in the same zone.
The following table describes other limits.
| Item | Limit | How to apply for an extension on the limit |
|---|---|---|
| Permissions to create a pay-as-you-go disk | You must complete real-name verification before you can create a disk within a region in mainland China. | None. |
| Quota for pay-as-you-go disks in all regions for an account | This quota is calculated by using the following formula: Number of ECS instances across all regions × 5. A minimum of ten pay-as-you-go disks can be created in each account. If only one instance exists in your account, you can create a maximum of ten pay-as-you-go disks in all regions, instead of five pay-as-you-go disks. | Submit a ticket. |
| Quota for the capacity of pay-as-you-go disks that are used as data disks in an account | This limit is subject to ECS resource usage, region, and disk category. You can go to the Privileges & Quotas page in the ECS console to view this information. For more information, see View and increase quotas (new version). | Submit a ticket. |
| Quota for system disks on an instance | 1 | None. |
| Quota for data disks on an instance | 16. | None. |
| Capacity of a basic disk | 5 GiB to 2,000 GiB. | None. |
| Capacity of a standard SSD | 20 GiB to 32,768 GiB. | None. |
| Capacity of an ultra disk | 20 GiB to 32,768 GiB. | None. |
| Capacity of an enhanced SSD (ESSD) | 20 GiB to 32,768 GiB. | None. |
| Capacity of a local SSD | 5 GiB to 800 GiB. | None. |
| Total capacity of all local SSDs on an instance | 1,024 GiB. | None. |
| Capacity of a system disk |
|
None. |
| Permissions to attach new local disks to instances that are equipped with local disks | Not allowed. | None. |
| Permissions to change configurations of instances that are equipped with local disks | Only bandwidth configurations of instances that are equipped with local disks can be changed. | None. |
| Mount points of system disks | /dev/vda | None. |
| Mount points of data disks | /dev/vd[b-z] | None. |
Billing methods
For more information, see the Pricing tab on the Elastic Compute Service page.
Disks support the subscription and pay-as-you-go billing methods. For more information, see Subscription and Pay-as-you-go.
- Disks that are created along with subscription instances or created separately for subscription instances are billed on a subscription basis.
- Disks that are created along with pay-as-you-go instances or created separately for pay-as-you-go instances are billed on a pay-as-you-go basis. Storage capacity units (SCUs) can be used to offset bills of pay-as-you-go disks.
After a disk is created, you can change its billing method. For more information, see Change billing methods of disks.
Related operations
The following table describes the operations that you can perform on disks.
| Operation | Reference |
|---|---|
| Attach an idle pay-as-you-go disk to an ECS instance |
|
| Create and use a data disk |
|
| Encrypt data stored on a disk | For more information, see Encryption overview.
For more information, see the following topics:
|
| Resize a system disk or data disk |
For more information about resizing, see Overview. For more information, see the following topics:
|
| Replace a system disk | Replace the system disk (public images) |
| Back up disk data | |
| Restore a disk to its initial status | Reinitialize a system disk |
| Restore a disk to its status at a specific point of time | Roll back a disk |
| Detach the damaged system disk of an instance and attach the disk back to the original instance after the disk is repaired | Detach or attach a system disk |
| Release an instance and retain data stored on its system disk | |
| Release a subscription disk that is no longer needed | |
| Release a pay-as-you-go disk that is no longer needed |