After data migration is complete, you can perform a performance assessment before switching the database to OceanBase Database. This topic describes how to create a full performance assessment task.
Background information
Performance evaluation tasks include three types: Full Assessment, SQL File Parsing, and SQL File Replay. A complete performance evaluation task can automatically schedule parsing and playback tasks; the playback will start automatically after the parsing is completed. You can choose the full process, or perform file parsing and load playback independently. The prerequisite for independently creating an SQL file playback task is that the file parsing is already complete.
Currently, performance evaluation supports playback from AWS Aurora MySQL databases and AWS MySQL databases, Huawei Cloud RDS MySQL databases and DWS databases, Alibaba Cloud RDS MySQL databases and PolarDB MySQL databases, and GCP MySQL databases to OceanBase databases.
Prerequisites
Before creating a complete performance evaluation task, you need to obtain the audit logs of the corresponding source database.
If you select Source Database Type as Aurora MySQL, AWS_MYSQL, RDS_MYSQL, POLAR_MYSQL, or GCP_MYSQL, see Obtain traffic files for a database instance for details.
If you select Source Database Type as HUAWEI_RDS_MySQL, see Enable SQL audit logging for details.
If you select Source Database Type as HUAWEI_DWS, see Export database audit logs for details.
Before performing an SQL file playback task, you need to enable the SQL audit feature in the target OceanBase database. For details, see Enable SQL audit.
Limitations
- Only users with the project roles of Project Administrator or Data Services Administrator can create a complete performance evaluation task.
Procedure
Create a full performance assessment task.
Log in to the OceanBase Cloud console.
In the left-side navigation pane, click Data Services > Data Migration.
On the Data Migration page, click the Assess Performance tab.
On the Assess Performance tab, click Create in the upper-right corner and select Full Assessment.
On the Set Up Source Database page, configure the parameters.
ParameterDescriptionTask Name The name of the custom assessment task, which must not exceed 30 characters. Instance Source Currently supports AWS, HUAWEI, ALIYUN, and Google Cloud. Source Database Type - If Cloud Vendor is set to AWS, the source database types are Aurora MySQL and AWS_MYSQL.
- If Cloud Vendor is set to HUAWEI, the source database types are HUAWEI_RDS_MySQL and HUAWEI_DWS.
- If Cloud Vendor is set to ALIYUN, the source database types are RDS_MYSQL and POLAR_MYSQL.
- If Cloud Vendor is set to Google Cloud, the source database type is GCP_MYSQL.
Region Select the region where the source database is located. Source Database Instance When Source Database Type is set to Aurora MySQL, AWS_MYSQL, HUAWEI_DWS, RDS_MYSQL, POLAR_MYSQL, or GCP_MySQL, you need to enter the source database instance. For details about how to view the database instance name, see Query database instances.
Separate multiple database instances with commas (,).Source Database This parameter is displayed only when Cloud Vendor is set to Google Cloud. Click Next. On the Configure Data Collection page, configure the parameters.
If the instance source is Google Cloud, Data Collected From supports only Import Object Storage (Google Cloud Object Storage).
If the instance source is AWS, Data Collected From supports S3 Import and Upload File.
If the instance source is ALIYUN, Data Collected From supports Import Object Storage Service (OSS) and Upload File.
If you have saved the audit logs of an Aurora MySQL instance or AWS MySQL instance to S3, or the audit logs of an Alibaba Cloud RDS_MySQL instance or Polar_MySQL instance to OSS, select S3 Import or Import Object Storage Service (OSS) to import the audit logs for assessment. For details about how to obtain the traffic files of GCP MySQL, Aurora MySQL, AWS_MySQL, RDS_MySQL, and Polar_MySQL instances, see Query database instances.
Note
We strongly recommend that you create a dedicated user for the migration assessment service and grant permissions only for the relevant bucket. For information about the Alibaba Cloud account system, see Access control.
ParameterDescriptionAccessKey Account The AccessKey ID of the S3, GCS, or Alibaba Cloud account. For details about the AccessKey for S3 or GCS, see Create an AccessKey. For details about the AccessKey for Alibaba Cloud, see Create an AccessKey. AccessKey Secret The AccessKey Secret of the S3, GCS, or Alibaba Cloud account. S3/OSS Bucket for Log Import Enter the name of the S3 or OSS bucket for log import. GCP Bucket for File Import Enter the name of the GCP bucket for log import.
This field appears when the instance source is Google Cloud.File Delivery Directory Enter the directory for delivering audit logs. Traffic Collection Time Range Select the time range to be evaluated. The maximum selectable time range is 24 hours.
If the obtained data is empty, no log files were generated within the selected time range.After configuring the parameters above, click Verify. After verification passes, Valid File List displays the files that can be parsed.
If you have saved the audit logs of an Aurora MySQL instance or AWS MySQL instance, or the traffic files from the Log Service SLS, locally, or if the instance source is HUAWEI, select Upload File to manually import ZIP or GZ format files for evaluation. Note that the maximum supported file size for upload is 5 GB, and only one file can be uploaded.
Click Next. On the Set Up Target Database page, configure the parameters.
ParameterDescriptionRegion The target region must match the source region. Target Database Instance Ensure the table structure and data of the target instance are consistent with the original instance to guarantee normal execution of the stress test traffic. Tenant (Optional) Select the ID of the tenant where the target database instance resides from the drop-down list. Target High-Privilege Account The username for the target database. For user permission details, see Replay User Permission Instructions. Password The password for the target database. After configuring the parameters above, click Verify. After verification passes, the Database List of Target End section displays the list of databases available for replay. Select the name of the database you want to replay, click >, and add it to the Target list on the right.
Click Next. After confirming that the selected target database matches the one in the traffic files, click OK in the dialog box that appears.
On the Configure Workload Replay page, configure the parameters.
ParameterDescriptionReplay Type Includes Read Traffic (DQL) and Write Traffic (DML). Replay Speed (x) Select the target replay speed from the drop-down list. A maximum of 10x is supported.
Replay speed indicates compressing the QPS or TPS of traffic forward to a shorter time range. For example, a speed of 2 means completing what would originally take 4 hours of replay in 2 hours.Sampling Rate Select the target sampling rate from the drop-down list. A maximum of 100% is supported.
The sampling rate indicates the probability that each SQL statement is independently selected for replay in an SQL replay scenario. For example, a sampling rate of 30% means each SQL has a 30% chance of being selected for replay, and the selection of each SQL is independent.Simulated Concurrency Simulated concurrency is enabled by default. Click Next to perform a precheck.
After the precheck passes, click Start Now to begin the performance assessment task.
After starting the task, you can view the assessment report on the Task List page. For details, see View the Performance Assessment Report. The parsed files are saved in the Assess Performance > Parse files page.
Notice
When starting a replay, performance assessment automatically adds an allowlist group named omaxxx to the OceanBase database instance you selected, ensuring the machine running the performance assessment replay task can connect to the OceanBase database instance normally.