ApsaraMQ for RabbitMQ:RabbitMQ serverless instance billing

Billing formula

Hourly messaging fee = Total messaging requests × Unit price per request (USD)

Unit prices

Billing example

Consider a serverless instance in the China (Hangzhou) region. In one hour, the usage is as follows:

• The broker receives 7 million normal messages from producers and 2 million delayed messages. Each message is 8 KB.

• The broker delivers 8 million normal messages to consumers and 3.2 million delayed messages. Each message is 8 KB.

Calculate the hourly messaging fee as follows: (7 + 8) × ⌈8/4⌉ × 0.31 + 2 × ⌈8/4⌉ × 1.55 + 3.2 × ⌈8/4⌉ × 0.31 = USD 17.484

Billing formula

Hourly provisioned capacity fee = Provisioned capacity (TPS) × Unit price of provisioned capacity (USD per TPS-hour)

Unit prices

Billing example

For example, consider a serverless ApsaraMQ for RabbitMQ instance in the China (Hangzhou) region. This instance is deployed in shared cluster mode and uses the Provisioned + Elastic billing method. If you provision a capacity of 3,000 TPS, your hourly provisioned capacity charge is 3,000 × 0.00013 = USD 0.39. The total charge for a 30-day month is 0.39 × 24 × 30 = USD 280.8.

Billing formula

$$Hourly\; elastic\; traffic\; fee=\sum _{1\le i\le 60}[Ti\times Elastic\; TPS]\times Elastic\; traffic\; unit\; price\; (USD/TPS/minute)$$

In this formula, T1, T2, ..., Ti, ..., and T60 represent the 1st, 2nd, ..., ith, ..., and 60th minutes of the hour. $$\sum$$ represents the sum of the elastic traffic used during that hour.

Unit prices

Billing example

Assume you have a shared Serverless instance in the China (Hangzhou) region with a provisioned capacity of 5,000 TPS. The instance uses the "Provisioned Capacity + Elastic Traffic" billing method. The following table shows the instance's usage over a one-hour period.

The instance's elastic traffic fee for this hour is (1 × 200 + 1 × 300 + 1 × 100) × 0.000019 = USD 0.0114.

Billing formula

Hourly queue fee = Number of queues on an instance × Unit price of a queue (USD/queue)

The billed quantity is the peak number of queues on an instance within a specific hour.

For example, if you create 300 queues and then delete 100 queues within one hour, you are charged for 300 queues for that hour.

Unit price

The unit price per queue is tiered based on the total number of queues.

Billing example

The following example calculates queue fees for an instance over a 2-hour period.

Billing formula

Hourly storage fee = Storage space (GB) × Unit price (USD per GB-hour)

Unit prices

Billing formula

Hourly internet fee = Total outbound internet traffic (GB) × Unit price (USD per GB)

Unit prices

Billing formula

Hourly message trace fee = Number of message trace logs × Unit price per trace (USD)

Log counting rules

The following API operations generate a billable message trace log. Each generated log incurs a charge, even if the operation fails.

• BasicAck

• BasicConsume

• BasicGet

• BasicNAck

• BasicReject

• SendMessage

• PushMessage

Unit prices

Billing formula

Unit prices

Q: My Serverless instance has low traffic but high monthly charges, or the provisioned capacity charge seems too high. What should I do?

Serverless instance charges consist of two components:

• Actual resource usage fees: Charges based on messaging requests, storage usage, queue count, outbound internet traffic, and message traces (if enabled).

• Provisioned capacity fees: Charges based on the provisioned capacity specification (TPS) and usage duration, independent of actual traffic volume. This applies only to instances using the Provisioned Capacity + Elastic billing method.

If your monthly charges are higher than expected, follow these steps to identify the cause:

1. Log on to Alibaba Cloud Billing Center to view the itemized billing details and identify the billing item with the highest charges.

2. If the provisioned capacity fee is high, check whether the provisioned capacity specification (TPS) is appropriately sized for your workload. Refer to the Provisioned Capacity and Elastic Traffic sections of this document for the billing rules and use the information to adjust your provisioned capacity specification if needed.

Q: Does the bill support viewing cost details at the individual queue level?

Currently, the bill and the console do not support viewing cost details broken down to the individual queue level.

As an alternative, you can estimate per-queue costs using the following method:

1. Enable and query Simple Log Service (SLS) to retrieve the sendCnt (message receptions) and pushCnt (message deliveries) values for each queue.

2. Apply the counting rules described in the Counting Rules for Messaging section — including the 4 KB payload size multiplier (⌈message size / 4 KB⌉) and the applicable unit price — to calculate the estimated cost for each queue.

Q: How can I estimate or compare the testing cost of different Serverless billing modes (pay-by-messaging-request vs. pay-by-provisioned-capacity-and-elastic-traffic)?

On the instance creation page, the estimated price updates in real time in the lower-right corner as you configure the instance. You can enter the purchase page, switch between different billing methods — pay-by-messaging-request, pay-by-provisioned-capacity-and-elastic-traffic, and subscription — and directly compare the estimated costs without confirming the purchase. This lets you evaluate which option is most cost-effective for your testing or production scenario.