Manage ECS resource quotas - Elastic Compute Service - Alibaba Cloud Documentation Center

A quota refers to the maximum number of resources that an Alibaba Cloud account can use or the maximum number of operations that an Alibaba Cloud account can perform on a resource. Quota Center allows you to query the quotas and quota usage of Elastic Compute Service (ECS) resources, apply for quotas online, or create alert rules for quota items based on your business requirements. This topic describes how to manage ECS resource quotas.

Resource quotas that can be increased

For information about the limits on ECS resource quotas and the resource quotas that can be increased, see Limits.

View or increase ECS resource quotas

If your business requires a number of ECS resources that exceeds the default quota value and the resource quota is adjustable in Quota Center, you can perform the following steps to view the quota and apply for a quota increase:

Log on to the Quota Center console.
In the left-side navigation pane, choose Products > General Quotas. On the Products with General Quotas page, select Elastic Compute Service, Ecs Quotas by Instance Type, or Elastic Block Storage based on the resource type.
- Ecs Quotas by Instance Type: Manage the quotas of ECS instance types.
- Elastic Compute Service: Manage the quotas of ECS resources, such as images and security groups.
- Elastic Block Storage: Manage the quotas of different disk categories.
On the General Quotas page, you can view or increase resource quotas.
- View resource quotas. On the General Quotas page, the values in the Total Quotas column indicate the maximum number of resources and the values in the In Use column indicate the usage of quotas.
- Apply for a quota increase. On the General Quotas page, find a quota, click Apply in the Actions column, and then follow the on-screen instructions to apply for a quota increase.
  Important
  - In normal cases, the system dynamically adjusts quotas based on your resource usage on Alibaba Cloud to ensure that sufficient quotas are available for you to create resources. However, if a quota remains unused in your account, the system may decrease the quota as appropriate. If you require a larger value, apply for a quota increase.
  - Specific quotas cannot be increased. For more information, see Limits.

Usage notes for quota allocation

The system dynamically adjusts quotas in your account based on your quota usage. If you retain a large number of ECS instances for an extended period of time, you do not need to be concerned about insufficiency of applicable quotas or apply for quota increases. If you expect the usage of a resource quota to surge, you can apply for an increase of the quota.

ECS resource quotas

Note

A quota limits the maximum number of vCPUs or GPUs for ECS resources in the current region, but not in all regions.

vCPU-based quotas

Quota code	Description	Instance family and references
q_ecs_restrict_prepay_c	Maximum number of vCPUs for subscription instances for limited purchase	xn4, n4, mn4, and e4, previous-generation shared instance families sccg5, general-purpose SCC instance family Retired instance families
q_ecs_local_storage_prepay_c	Maximum number of vCPUs for subscription instances of instance families with local SSDs, such as d and i series instance families	Big data instance families Instance families with local SSDs (i series)
q_ecs_share_prepay_c	Maximum number of vCPUs for subscription instances of shared instance families, such as e and t series instance families	e, economy instance family t6, burstable instance family t5, burstable instance family
q_ecs_ebm_prepay_c	Maximum number of vCPUs for subscription instances of ECS Bare Metal Instance families	Overview (excluding GPU-accelerated compute-optimized ECS Bare Metal Instance families)
q_ecs_hpc_prepay_c	Maximum number of vCPUs for subscription instances of compute-optimized high-performance instance families, such as High Performance Computing (HPC) and Super Computing Cluster (SCC) instance families	scchfc6, compute-optimized SCC instance family with high clock speeds scchfg6, general-purpose SCC instance family with high clock speeds scchfr6, memory-optimized SCC instance family with high clock speeds scch5, SCC instance family with high clock speeds
q_ecs_high_mem_prepay_c	Maximum number of vCPUs for subscription instances of memory-optimized instance families, such as re and se series instance families	Memory-optimized instance families Memory-optimized instance families Memory-optimized instance families se1ne, network-enhanced memory-optimized instance family se1, memory-optimized instance family
q_ecs_enterprise_prepay_c	Maximum number of vCPUs for subscription instances of enterprise-level instance families, such as g, c, r, u, hf, and sn series instance families	General-purpose instance families (g series) Compute-optimized instance families (c series) Memory-optimized instance families (excluding re6p, re6, re4, se1ne, and se1 instance families) Universal instance families Instance families with high clock speeds (hf series)
q_ecs_restrict_postpay_c	Maximum number of vCPUs for pay-as-you-go instances for limited purchase	xn4, n4, mn4, and e4, previous-generation shared instance families sccg5, general-purpose SCC instance family Retired instance families
q_ecs_local_storage_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of instance families with local SSDs, such as d and i series instance families	Big data instance families Instance families with local SSDs (i series)
q_ecs_share_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of shared instance families, such as e and t series instance families	e, economy instance family t6, burstable instance family t5, burstable instance family
q_ecs_ebm_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of ECS Bare Metal Instance families	Overview (excluding GPU-accelerated compute-optimized ECS Bare Metal Instance families)
q_ecs_hpc_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of compute-optimized high-performance instance families, such as HPC and SCC instance families	scchfc6, compute-optimized SCC instance family with high clock speeds scchfg6, general-purpose SCC instance family with high clock speeds scchfr6, memory-optimized SCC instance family with high clock speeds scch5, SCC instance family with high clock speeds
q_ecs_high_mem_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of memory-optimized instance families, such as re and se series instance families	Memory-optimized instance families Memory-optimized instance families Memory-optimized instance families se1ne, network-enhanced memory-optimized instance family se1, memory-optimized instance family
q_ecs_enterprise_postpay_c	Maximum number of vCPUs for pay-as-you-go instances of enterprise-level instance families, such as g, c, r, u, hf, and sn series instance families	General-purpose instance families (g series) Compute-optimized instance families (c series) Memory-optimized instance families (excluding re6p, re6, re4, se1ne, and se1 instance families) Universal instance families Instance families with high clock speeds (hf series)
q_ecs_restrict_spot_c	Maximum number of vCPUs for preemptible instances for limited purchase	xn4, n4, mn4, and e4, previous-generation shared instance families sccg5, general-purpose SCC instance family Retired instance families
q_ecs_local_storage_spot_c	Maximum number of vCPUs for preemptible instances of instance families with local SSDs, such as d and i series instance families	Big data instance families Instance families with local SSDs (i series)
q_ecs_share_spot_c	Maximum number of vCPUs for preemptible instances of shared instance families, such as e and t series instance families	e, economy instance family t6, burstable instance family t5, burstable instance family
q_ecs_ebm_spot_c	Maximum number of vCPUs for preemptible instances of ECS Bare Metal Instance families	Overview (excluding GPU-accelerated compute-optimized ECS Bare Metal Instance families)
q_ecs_hpc_spot_c	Maximum number of vCPUs for preemptible instances of compute-optimized high-performance instance families, such as HPC and SCC instance families	scchfc6, compute-optimized SCC instance family with high clock speeds scchfg6, general-purpose SCC instance family with high clock speeds scchfr6, memory-optimized SCC instance family with high clock speeds scch5, SCC instance family with high clock speeds
q_ecs_high_mem_spot_c	Maximum number of vCPUs for preemptible instances of memory-optimized instance families, such as re and se series instance families	Memory-optimized instance families Memory-optimized instance families Memory-optimized instance families se1ne, network-enhanced memory-optimized instance family se1, memory-optimized instance family
q_ecs_enterprise_spot_c	Maximum number of vCPUs for preemptible instances of enterprise-level instance families, such as g, c, r, u, hf, and sn series instance families	General-purpose instance families (g series) Compute-optimized instance families (c series) Memory-optimized instance families (excluding re6p, re6, re4, se1ne, and se1 instance families) Universal instance families Instance families with high clock speeds (hf series)

GPU-based quotas

Quota code	Description	Instance family and references
q_ecs_ag_prepay_g	Maximum number of GPUs for subscription Arm instances of GPU-accelerated instance families	ebmgn6ia, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_vgpu_prepay_g	Maximum number of subscription instances of GPU- and vGPU-accelerated instance families	sgn7i-vws, vGPU-accelerated instance family with shared CPUs vgn7i-vws, vGPU-accelerated instance family vgn6i-vws, vGPU-accelerated instance family vgn6i, vGPU-accelerated instance family vgn5i, vGPU-accelerated instance family
q_ecs_gn7i_prepay_g	Maximum number of GPUs for subscription instances of GPU-accelerated instance families, such as gn7i, gn7s, and ebmgn7i instance families	gn7i, GPU-accelerated compute-optimized instance family gn7s, GPU-accelerated compute-optimized instance family ebmgn7i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn5_prepay_g	Maximum number of GPUs (excluding vGPUs) for subscription instances of GPU-accelerated instance families, such as gn5 and gn5i instance families	gn5, GPU-accelerated compute-optimized instance family gn5i, GPU-accelerated compute-optimized instance family
q_ecs_gn6i_prepay_g	Maximum number of GPUs (excluding vGPUs) for subscription instances of GPU-accelerated instance families, such as gn6i and ebmgn6i instance families	gn6i, GPU-accelerated compute-optimized instance family ebmgn6i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6v_prepay_g	Maximum number of GPUs for subscription instances of GPU-accelerated instance families, such as gn6v and ebmgn6v instance families	gn6v, GPU-accelerated compute-optimized instance family ebmgn6v, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6e_prepay_g	Maximum number of GPUs for subscription instances of GPU-accelerated instance families, such as gn6e and ebmgn6e instance families	gn6e, GPU-accelerated compute-optimized instance family ebmgn6e, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7v_prepay_g	Maximum number of GPUs for subscription instances of GPU-accelerated instance families, such as gn7, ebmgn7, and gn7v instance families	gn7, GPU-accelerated compute-optimized instance family ebmgn7, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7e_prepay_g	Maximum number of GPUs for subscription instances of GPU-accelerated instance families, such as gn7e and ebmgn7e instance families	gn7e, GPU-accelerated compute-optimized instance family ebmgn7e, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_ag_postpay_g	Maximum number of GPUs for pay-as-you-go Arm instances of GPU-accelerated instance families	ebmgn6ia, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_vgpu_postpay_g	Maximum number of pay-as-you-go instances of GPU- and vGPU-accelerated instance families	sgn7i-vws, vGPU-accelerated instance family with shared CPUs vgn7i-vws, vGPU-accelerated instance family vgn6i-vws, vGPU-accelerated instance family vgn6i, vGPU-accelerated instance family vgn5i, vGPU-accelerated instance family
q_ecs_gn7i_postpay_g	Maximum number of GPUs for pay-as-you-go instances of GPU-accelerated instance families, such as gn7i, gn7s, and ebmgn7i instance families	gn7i, GPU-accelerated compute-optimized instance family gn7s, GPU-accelerated compute-optimized instance family ebmgn7i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn5_postpay_g	Maximum number of GPUs (excluding vGPUs) for pay-as-you-go GPU-accelerated instances, such as gn5 and gn5i instance families	gn5, GPU-accelerated compute-optimized instance family gn5i, GPU-accelerated compute-optimized instance family
q_ecs_gn6i_postpay_g	Maximum number of GPUs (excluding vGPUs) for pay-as-you-go instances of GPU-accelerated instance families, such as gn6i and ebmgn6i instance families	gn6i, GPU-accelerated compute-optimized instance family ebmgn6i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6v_postpay_g	Maximum number of GPUs for pay-as-you-go instances of GPU-accelerated instance families, such as gn6v and ebmgn6v instance families	gn6v, GPU-accelerated compute-optimized instance family ebmgn6v, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6e_postpay_g	Maximum number of GPUs for pay-as-you-go instances of GPU-accelerated instance families, such as gn6e and ebmgn6e instance families	gn6e, GPU-accelerated compute-optimized instance family ebmgn6e, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7v_postpay_g	Maximum number of GPUs for pay-as-you-go instances of GPU-accelerated instance families, such as gn7, ebmgn7, and gn7v instance families	gn7, GPU-accelerated compute-optimized instance family ebmgn7, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7e_postpay_g	Maximum number of GPUs for pay-as-you-go instances of GPU-accelerated instance families, such as gn7e and ebmgn7e instance families	gn7e, GPU-accelerated compute-optimized instance family ebmgn7e, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_ag_spot_g	Maximum number of GPUs for preemptible Arm instances of GPU-accelerated instance families	ebmgn6ia, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_vgpu_spot_g	Maximum number of preemptible instances of GPU- and vGPU-accelerated instance families	sgn7i-vws, vGPU-accelerated instance family with shared CPUs vgn7i-vws, vGPU-accelerated instance family vgn6i-vws, vGPU-accelerated instance family vgn6i, vGPU-accelerated instance family vgn5i, vGPU-accelerated instance family
q_ecs_gn7i_spot_g	Maximum number of GPUs for preemptible instances of GPU-accelerated instance families, such as gn7i, gn7s, and ebmgn7i instance families	gn7i, GPU-accelerated compute-optimized instance family gn7s, GPU-accelerated compute-optimized instance family ebmgn7i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn5_spot_g	Maximum number of GPUs (excluding vGPUs) for preemptible instances of GPU-accelerated instance families, such as gn5 and gn5i instance families	gn5, GPU-accelerated compute-optimized instance family gn5i, GPU-accelerated compute-optimized instance family
q_ecs_gn6i_spot_g	Maximum number of GPUs (excluding vGPUs) for preemptible instances of GPU-accelerated instance families, such as gn6i and ebmgn6i instance families	gn6i, GPU-accelerated compute-optimized instance family ebmgn6i, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6v_spot_g	Maximum number of GPUs for preemptible instances of GPU-accelerated instance families, such as gn6v and ebmgn6v instance families	gn6v, GPU-accelerated compute-optimized instance family ebmgn6v, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn6e_spot_g	Maximum number of GPUs for preemptible instances of GPU-accelerated instance families, such as gn6e and ebmgn6e instance families	gn6e, GPU-accelerated compute-optimized instance family ebmgn6e, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7v_spot_g	Maximum number of GPUs for preemptible instances of GPU-accelerated instance families, such as gn7, ebmgn7, and gn7v instance families	gn7, GPU-accelerated compute-optimized instance family ebmgn7, GPU-accelerated compute-optimized ECS Bare Metal Instance family
q_ecs_gn7e_spot_g	Maximum number of GPUs for preemptible instances of GPU-accelerated instance families, such as gn7e and ebmgn7e instance families	gn7e, GPU-accelerated compute-optimized instance family ebmgn7e, GPU-accelerated compute-optimized ECS Bare Metal Instance family

References

You can call the DescribeAccountAttributes operation to query the quotas of ECS resources that you can create in an Alibaba Cloud region. For a specific region, you can query the maximum numbers of security groups, elastic network interfaces (ENIs), vCPUs for pay-as-you-go instances, vCPUs for preemptible instances, the capacity quotas for pay-as-you-go disks that you can create and the number of dedicated hosts that you can create. You can also query information such as network types or whether an account has completed real-name verification.
Quota Alerts: You can create an alert rule for a quota item. If the usage of a quota reaches the specified threshold, Quota Center sends an alert notification to you. You can apply for a quota increase at the earliest opportunity.
Query applications and view application details: After you apply for quota increases, you can query the applications and view the application details. You can also view the results of the applications.
For information about more quota operations, see Quota Center documentation.