View Data Transmission Service quotas in MaxCompute and configure their concurrency settings. Split purchased quotas into smaller units, assign reserved and elastic concurrency to each, and schedule time-based scaling plans.
For more information about exclusive Data Transmission Service resource groups, see Purchase and use exclusive Data Transmission Service resource groups and Billing of exclusive Data Transmission Service resources (subscription).
Quota hierarchy
Each purchased subscription creates a level-1 quota with a fixed reserved concurrency equal to the total concurrency purchased. Divide a level-1 quota into up to 20 level-2 quotas to isolate resources across teams or workloads.
Two types of concurrency apply at each level:
-
Reserved concurrency -- Guaranteed capacity allocated monthly. For a level-1 quota, this equals the total concurrency purchased. To change it, upgrade or downgrade the subscription.
-
Elastic reserved concurrency -- Additional flexible capacity. For a level-1 quota, this must be less than or equal to the reserved concurrency. The default is 0. Configuring elastic reserved concurrency incurs additional fees.
The following rules govern level-2 quotas:
-
The sum of reserved concurrency across all level-2 quotas equals the level-1 reserved concurrency.
-
The sum of elastic reserved concurrency across all level-2 quotas equals the level-1 elastic reserved concurrency.
-
Every level-1 quota includes a default level-2 quota whose values are auto-calculated. The default level-2 quota absorbs whatever concurrency is not assigned to custom level-2 quotas.
Example:
A level-1 quota has 100 reserved concurrency and 40 elastic reserved concurrency. Two custom level-2 quotas are created:
|
Level-2 quota |
Reserved concurrency |
Elastic reserved concurrency |
|
team-analytics |
60 |
20 |
|
team-etl |
25 |
15 |
|
default (auto-calculated) |
15 |
5 |
|
Level-1 total |
100 |
40 |
Formula: Default level-2 reserved = Level-1 reserved - Sum(Custom level-2 reserved)
Prerequisites
Before you begin, make sure that you have:
-
A MaxCompute project with a subscription billing method
-
An exclusive Data Transmission Service (subscription) resource group
-
Permissions to configure quotas (Alibaba Cloud accounts have this permission by default; grant it to Resource Access Management (RAM) users as needed -- see RAM permissions)
Configure basic settings
-
Log on to the MaxCompute console and select a region in the top-left corner.
-
In the left navigation pane, choose Manage Configurations > Quotas.
-
On the Quotas page, find the target quota and click Quota Configuration in the Actions column.
Only quotas with the Subscription Billing Method are supported.
-
On the Quota Configuration page, select the Basic Configurations tab and click Edit Basic Configurations.
-
Click Add Level-2 Quota to create a new sub-quota, or configure Reserved Concurrency and Elastic Reserved Concurrency for an existing level-2 quota.
Parameter
Description
Quota Name
A custom name. Must start with a letter and can contain letters, digits, and underscores (_). When referencing this quota in an SDK or connector, use the system-generated name shown in the console quota management list, for example
"ot_42854300324****_169821756****_p#ot_42854300324****_169821756****".Reserved Concurrency
Monthly reserved concurrency for the level-2 quota. The level-1 reserved concurrency cannot be changed here — upgrade or downgrade the subscription instead. The sum of all level-2 reserved concurrency equals the level-1 reserved concurrency. The default level-2 quota absorbs the remainder.
Elastic Reserved Concurrency
Monthly elastic reserved concurrency for the level-2 quota. The level-1 elastic reserved concurrency must be less than or equal to the level-1 reserved concurrency (default: 0). The sum of all level-2 elastic reserved concurrency equals the level-1 elastic reserved concurrency. The default level-2 quota absorbs the remainder.
Delete a level-2 quota
On the Quota Configuration page, select the Basic Configurations tab, find the target quota, and click Delete in the Actions column.
After deletion, the Reserved Concurrency and Elastic Reserved Concurrency of the deleted level-2 quota are added back to the default level-2 quota. To redistribute these values, edit the configuration plans on the Scaling Configuration tab.
Configure scaling plans
After purchasing a subscription MaxCompute project, each level-1 quota has a default resource configuration plan named Default.
Each configuration plan includes the reserved concurrency and elastic reserved concurrency for the level-1 quota and all its level-2 quotas. Use multiple plans with time-based management when concurrency requirements change throughout the day -- for example, different computing unit (CU) requirements between the 00:00--08:00 and 08:00--24:00 periods.
-
A quota resource group supports a maximum of 48 time-based intervals. The minimum duration of a single interval is 30 minutes.
-
To downgrade the Data Transmission Service (subscription), first reduce the time-based concurrency.
-
The time-based configuration takes 0 to 5 minutes to take effect. Billing starts from the actual effective time.
Add a configuration plan:
-
On the Quota Configuration page, select the Scaling Configuration tab.
-
Click Add Configuration Plan. In the Add Configuration Plan dialog box, enter a Configuration Plan Name and configure Reserved Concurrency and Elastic Reserved Concurrency for each level-2 quota.
-
Click OK to add the plan.
Apply a configuration plan immediately:
On the Scaling Configuration tab, find the target plan and click Apply Immediately in the Actions column.
If time-based management is enabled, the manually applied plan is overwritten by the next scheduled plan. To keep a manually activated plan active, clear all time-based settings first.
More operations on configuration plans
|
Operation |
Steps |
|
View |
On the Scaling Configuration tab, click the expand icon to the left of Configuration Plan Name to view details. |
|
Edit |
Find the target plan and click Edit in the Actions column. If the plan is currently active, click Apply Immediately after editing to apply the update. |
|
Delete |
Find the target plan and click Delete in the Actions column. The |
|
Clone |
Find the target plan and click Clone in the Actions column. |
Set up time-based management
Schedule different configuration plans at different times of the day to match concurrency to workload patterns.
-
A maximum of 48 configuration plans can be added to a time-based schedule.
-
Time-based scheduling may have a delay of up to 5 minutes.
-
On the Scaling Configuration tab, click Edit Time Plan in the Scheduled Scaling Management area.
-
Click Add Effective Period, then select an Effective Start Time and a Configuration Plan.
Note-
The default start time is 00:00 and cannot be modified or deleted. Its configuration plan is Default, which can be changed. If no other start times are added, the same plan applies all day.
-
Start times are available on the hour and half-hour. A configuration plan can be reused across multiple time slots.
-
-
Click Save to save the schedule.
View quota usage
Monitor resource usage of the Data Transmission Service within each level-2 quota. Both Alibaba Cloud accounts and RAM users can view this information by default.
-
Log on to the MaxCompute console and select a region in the top-left corner.
-
In the left navigation pane, choose Manage Configurations > Quotas.
-
On the Quotas page, find the target quota and click Configure Resources in the Actions column.
The following metrics are available:
|
Metric |
Description |
|
Request Parallelism |
Slot usage curve based on the filter conditions. Displays current usage and quota limits. Unit: Slot. |
|
Throughput |
Throughput curve based on the filter conditions. Unit shown on the vertical axis (for example, B/min or MB/min). |
|
Table-level Request Parallelism |
Concurrency curve for data uploaded to a specific table using Tunnel Batch. Set Usage Type (for example, Tunnel Batch upload) and Table Name as filter conditions. Unit: Slot. |
|
Table-level IP Throughput |
Throughput curve from each source IP for data uploaded to a specific table using Tunnel Batch. Set Usage Type and Table Name as filter conditions. |
|
Total Requests and Error Requests |
Total and failed request counts. Failed requests include all requests with a 4xx or 5xx status code. For more information, see Data Transmission Service overview. |
|
Total Throughput |
Data volume summary for different usage types within the specified time range. A pie chart shows the proportion of each usage type. |
|
Slot Average Transfer Rate |
Average transfer rate per slot for Tunnel Batch upload requests. Set Usage Type as a filter condition. |
Supported regions
|
Region name |
Region ID |
|
China (Shenzhen) |
cn-shenzhen |
|
China (Shanghai) |
cn-shanghai |
|
China (Beijing) |
cn-beijing |
|
China (Hangzhou) |
cn-hangzhou |
|
China (Hong Kong) |
cn-hongkong |
|
Singapore |
ap-southeast-1 |
|
Indonesia (Jakarta) |
ap-southeast-5 |
|
US (Virginia) |
us-east-1 |
Limits
|
Constraint |
Value |
|
Maximum level-2 quotas per level-1 quota |
20 |
|
Maximum time-based intervals per schedule |
48 |
|
Minimum interval duration |
30 minutes |
|
Scheduling granularity |
On the hour and half-hour |
|
Time-based configuration activation delay |
0 to 5 minutes |