AnalyticDB:Purchase, configuration change, and upgrade

How are Enterprise Edition and Basic Edition clusters billed?

When you purchase an Enterprise Edition or Basic Edition cluster, the upfront payment on the buy page covers reserved resources only. Storage and elastic resources are billed on a pay-as-you-go basis as you use them. For the full billing breakdown, see Billable items of Enterprise Edition and Basic Edition.

Example: A pay-as-you-go Enterprise Edition cluster with single-zone deployment in the China (Hangzhou) region, 3 nodes at 8 AnalyticDB Compute Units (ACUs) each, storing 100 GB of hot data and 260 GB of cold data:

Why am I charged pay-as-you-go fees for my subscription cluster?

Subscription fees cover basic resources only. If your cluster consumes additional storage or elastic resources beyond the subscription, those are billed on a pay-as-you-go basis and appear as separate line items in your bill. See Billable items for the complete list.

How do I stop charges for a cluster I no longer use?

The answer depends on your billing method:

• Subscription cluster: You have already paid for reserved resources. To avoid additional pay-as-you-go charges, delete or stop any operations that generate them — such as elastic resources, excess storage, caches, or backups. See Billable items for the full list of pay-as-you-go items.

• Pay-as-you-go cluster: Billed hourly. To stop all charges, delete the cluster.

How do I confirm that billing has stopped?

After deleting a cluster, billing stops immediately. To verify, log on to the AnalyticDB for MySQL console, select the region in the upper-left corner, and confirm the cluster no longer appears in the list.

Can I reduce costs without deleting my cluster?

Cluster shutdown is not supported. To reduce costs, purchase storage plans or ACU-hour plans to offset storage and resource consumption, including reserved computing resources, reserved storage resources, and elastic resources.

What are the differences between Enterprise Edition and Basic Edition?

Both editions support elastic mode (resource group isolation, elastic resource scaling, and tiered storage of hot and cold data) and reserved mode (high throughput, real-time writes, and high-concurrency real-time queries). The key difference is storage architecture:

For the full feature comparison, see Editions.

What are the differences between Data Warehouse Edition and Data Lakehouse Edition?

Data Lakehouse Edition builds on the high-performance real-time analysis of Data Warehouse Edition (Enterprise Edition) and adds cost-effective batch processing and machine learning capabilities. It decouples computing and storage, which eliminates the consistency and timeliness issues that data synchronization can introduce. For details, see Editions.

What are the differences between elastic mode and reserved mode in Data Warehouse Edition?

Separating computing and storage in elastic mode lets you identify whether a performance bottleneck is in compute or storage and scale each independently to control costs.

What are elastic I/O resources?

Elastic I/O resources are measured and sold in elastic I/O units (EIUs). EIUs define the storage performance of a cluster. Real-time data and hot data are stored in EIUs to sustain high-throughput workloads from data import and index building. For more information, see Scale out elastic I/O resources.

Can I shut down an AnalyticDB for MySQL cluster?

No. AnalyticDB for MySQL clusters cannot be shut down.

Can I restart an AnalyticDB for MySQL cluster from the console?

No. To restart a cluster, submit a ticket.

What do I do if my cluster is locked?

If your Data Warehouse Edition cluster in elastic mode is locked because hot data stored in an elastic I/O unit (EIU) exceeds 8 TB, scale up your elastic I/O resources as soon as possible.

In what scenarios may cluster availability be affected?

Cluster availability may be affected when failures occur on the cluster or when the cluster undergoes configuration changes or version upgrades.

How do I check the minor version of my cluster?

1. Connect to the cluster.

2. Run the following statement:

   select adb_version();

3. Read the source_version field in the result. Sample output:

   +----------------+----------------+---------------------+-------------+---------------------+
   | source_version | rpm_version    | build_date          | revision_id | revision_date       |
   +----------------+----------------+---------------------+-------------+---------------------+
   | 3.1.4.10       | 20210809205222 | 2021-08-09 20:52:22 | 97b631a     | 2021-08-09 20:32:04 |
   +----------------+----------------+---------------------+-------------+---------------------+

How long does an upgrade take, and does it affect my business?

Upgrades typically take 30 minutes to 2 hours, depending on cluster specifications and data volume. During the upgrade, your application may experience transient connection drops. Configure your application to automatically reconnect to avoid disruption. For best results, schedule upgrades during nighttime or off-peak hours.

Does a configuration change affect the number of shards?

No. A configuration change does not affect the number of shards.