Transparent data encryption (TDE) performance - ApsaraDB RDS

This topic analyzes how enabling Transparent Data Encryption (TDE) on ApsaraDB RDS for MySQL instances affects performance, including CPU and memory usage. Use this report to evaluate if TDE is suitable for your workload.

Summary of results

Enabling TDE introduces measurable performance overhead and increases CPU utilization on ApsaraDB RDS for MySQL instances.

At low concurrency, the performance overhead is greater, reaching up to 20%. At high concurrency, the overhead decreases to less than 5% due to factors such as I/O consolidation.
On a 4-core, 16 GB instance, enabling TDE increases CPU utilization by 0.06% to 4.22% when accessing encrypted tables. On an 8-core, 32 GB instance, the increase is 0.12% to 2.86%.
Overall, the performance impact is most noticeable in write-intensive workloads, such as the oltp_write_only and oltp_update_index test scenarios.

Test environment

Parameter	Test scenario 1	Test scenario 2
Region and zone	China (Hangzhou)
Network type	Virtual Private Cloud (VPC)
Hardware architecture	x86-64
Storage type	local SSD
CPU	4-core	8-core
Memory	16 GB	32 GB
buffer pool size	12 GB	24 GB
Maximum IOPS	7000	12000
Database engine version	mysql80_8.0.28_20230610
Test dataset size	41 GB

Benchmark tool

sysbench is an open-source, modular, cross-platform, and multi-threaded benchmark tool for evaluating system performance under high database loads. For more information, see the sysbench documentation.

Key metrics

transactions per second (TPS): The number of transactions a database commits per second.
average latency: The average time to execute a single transaction, measured in milliseconds (ms).

Test table schema

The default table schema used by sysbench is as follows:

CREATE TABLE `sbtest8` (
  `id` int NOT NULL AUTO_INCREMENT,
  `k` int NOT NULL DEFAULT '0',
  `c` char(120) NOT NULL DEFAULT '',
  `pad` char(60) NOT NULL DEFAULT '',
  PRIMARY KEY (`id`),
  KEY `k_8` (`k`)
) ENGINE=InnoDB AUTO_INCREMENT=100001 DEFAULT CHARSET=utf8mb3

Procedure

Create an ApsaraDB RDS for MySQL instance. For more information, see Create an ApsaraDB RDS for MySQL instance.
Create a database and an account. For more information, see Create a database and an account.
Install a MySQL client on your Linux-based ECS instance.
- For CentOS, run sudo yum install mysql.
- For Ubuntu, run sudo apt-get update and then sudo apt install mysql-server.
Connect to your ApsaraDB RDS for MySQL instance from the command line:
```
mysql -h<endpoint> -u<username> -P<port> -p<password>
```
Note
To find the endpoint and port for your instance, see View and manage instance endpoints and ports.
Exit the database and use sysbench to prepare and load the test data.
Note
- In this test, sysbench loads 128 tables, each with 100,000 rows. The following script uses the oltp_read_write test model.
- The sample code sets the thread count to 16. Adjust the --threads parameter to your desired value, such as 1, 8, 16, 32, 64, 128, or 256.
```
sysbench --db-driver=mysql --mysql-host=[database server host] --mysql-port=[database server port] --mysql-user=[database user name] --mysql-password=[database user password] --mysql-db=testdb --table_size=1000000 --tables=128 --threads=16 --time=60 oltp_read_write prepare
```

Run the benchmark test.

sysbench --db-driver=mysql --mysql-host=[database server host] --mysql-port=[database server port] --mysql-user=[database user name] --mysql-password=[database user password] --mysql-db=testdb --table_size=1000000 --tables=128 --threads=16 --time=60 oltp_read_write run

Clean up the test data.

sysbench --db-driver=mysql --mysql-host=[database server host] --mysql-port=[database server port] --mysql-user=[database user name] --mysql-password=[database user password] --mysql-db=testdb --table_size=1000000 --tables=128 --threads=16 --time=60  oltp_read_write cleanup