Cloud spanner instance auto-scaling

Question

I have created a Cloud Spanner instance, with a number of nodes that will provide compute resources for the instance. As the instance's workload changes, Will  Cloud Spanner  adjust the number of nodes in the instance? If not can I set some alerts to notify the utilization of resources so that it doesn't go beyond my budget limit?

Answer 1

As the instance's workload changes, Cloud Spanner does not automatically adjust the number of nodes in the instance. 

As a result, you need to set up several alerts to ensure that the instance stays within the recommended maximums for CPU utilization and the recommended limit for storage per node.

To create the recommended alerts:

1. In the Google Cloud Platform Console, select Monitoring, or use the following button:

2. Select Alerting > Create a Policy.

3. In the Find resource type and metric box, enter the values that may be required, for:
   High priority CPU - spanner.googleapis.com/instance/cpu/utilization_by_priority.
   24 hour rolling average CPU - spanner.googleapis.com/instance/cpu/smoothed_utilization.
   storage -  spanner.googleapis.com/instance/storage/used_bytes.

4. Click Show advanced options, then enter the  recommended values for the alerting policy's target and configuration (refer to https://cloud.google.com/spanner/docs/monitoring-stackdriver)

5. Click Save.

6. Optional: To configure your notification settings, click Add Notification Channel.

   You can elect to receive notifications by email, SMS, and several other options.

7. Optional: Enter a notification message in the Documentation section.

8. Name your policy and click Save.

9. Repeat these steps for each of the metrics.

Hope it helps!!

If you need to know more about Google Cloud, We recommend joining GCP Certification course today.

Thanks!

Comments

Although Spanner does not natively support Autoscaling, you can look at using the Cloud Spanner ecosystem's autoscaler project.