Setup
Using the Machmeter setup command to start creating resources and the execute command to load randomised data into Cloud Spanner with the help of a JMeter template.
Table of contents
Resource Creation using IaaC
Required resources involve executing the setup command of machmeter with spanner instance config, GKE cluster config and path to Schema file containing SQL command describing the database structure. One important thing to remember is that the spanner config will need to remain the same in all these steps.
Machmeter uses Terraform to create a GKE cluster which is used to generate the load. Refer to Kubernetes concepts cheat sheet and kubectl CLI cheat sheet for a quick reference to common concepts and commands.
Workload Identity for Authentication
The steps mentioned in Use Workload Identity are automated in the setup phase. It triggers the creation of a Google Service Account with Spanner Admin role and a Kubernetes Service account with Workload Identity User binding to the Google Service Account. This enables the Kubernetes Service Account to impersonate the Google Service Account when authenticating for requests to Spanner in execute phase.
Configuration File Definition
| Key name | Description |
|---|---|
| infraConfig.spannerInstanceConfig.projectId | Google Cloud Project Id |
| infraConfig.spannerInstanceConfig.instanceId | Name of the instance to be created |
| infraConfig.spannerInstanceConfig.dbName | Name of the database to be created |
| infraConfig.spannerInstanceConfig.configuration | Region in which the spanner instance is to be created |
| infraConfig.spannerInstanceConfig.displayName | Instance display name |
| infraConfig.spannerInstanceConfig.processingUnits | Number of processing units |
| infraConfig.spannerInstanceConfig.environment | Creation environment |
| gkeConfig.clusterName | GKE cluster name |
| gkeConfig.namespace | GKE cluster name space |
| gkeConfig.region | GKE cluster region |
| gkeConfig.network | GKE cluster network |
| gkeConfig.subnetwork | GKE cluster sub network |
| gkeConfig.ipRangePodsName | GKE cluster IP range Pod’s Name |
| gkeConfig.ipRangeServicesName | GKE cluster IP range Services Name |
| gkeConfig.google_service_account | Name of Google Cloud Service Account to be created |
| gkeConfig.kube_service_account | Name of GKE Service Account to be created |
| jMeterParams.spannerInstanceConfig.projectId | Google Cloud Project Id (Should be same as in previoud keys) |
| jMeterParams.spannerInstanceConfig.instanceId | Name of the instance to be created (Should be same as in previoud keys) |
| jMeterParams.spannerInstanceConfig.dbName | Name of the database to be created (Should be same as in previoud keys) |
| jMeterParams.spannerInstanceConfig.configuration | Region in which the spanner instance is to be created |
| jMeterParams.schemaFilePath | Path to the Schema file. |
Please refer to the sample file for example values.
This is a sample file with the above defined configuration.
Execution Command
java -jar target/machmeter/machmeter.jar setup path-to-config-file.json
Data Load
Data load involves executing the execute command of machmeter with data load jmeter script as input. This script, when executed, will load a randomized set of data onto a given mentioned spanner database.
This will also create CSV files containing the keys for created data. These CSV will be used as a source for parameters in queries in the Perf test.
Configuration File Definition
| Key name | Description |
|---|---|
| namespace | Name of the GKE namespace |
| jMeterTemplatePath | Path to Data load Jmeter Script |
| jMeterParams.project | Google Cloud Project Id |
| jMeterParams.instance | Name of the instance to be created |
| jMeterParams.database | Name of the database to be created |
| jMeterParams.connections | Spanner Connection Counts |
Please refer to the sample file for example values.
This is a sample file with the above defined configuration.
Execution Command
Ensure you have the GKE credentials configured via: gcloud container clusters get-credentials CLUSTER_NAME --region=COMPUTE_REGION
java -jar target/machmeter/machmeter.jar execute path-to-config-file.json