Description

There exists a scenario that when you

• Re-ingest cloud logs for a time period that has already been ingested.
• The log files include new events not previous ingested for that time period
• More than one VM with the same instance type but different root volume sizes is started at the same time

This scenario produces an Integrity Constraint Violation error on modw_cloud.instance_type.

The cause of this is multiple rows in modw_cloud.instance_type_staging that have the same resource_id, instance_type, num_cores, memory_mb, start time, and disk_gb. When the CloudInstanceTypeStateIngestor runs it uses the order of the rows to check for changes to an instance type over time. Adding the disk_gb column to the order ensures that rows with resource_id, instance_type, num_cores, memory_mb, start time, and disk_gb the same are collapsed into the same row when CloudInstanceTypeStateIngestor is run and prevents the Integrity Constraint Violation error.

Tests performed

Tested on xdmod-dev where I first ran into the issue.

Checklist:

• The pull request description is suitable for a Changelog entry
• The milestone is set correctly on the pull request
• The appropriate labels have been added to the pull request