CloudQuery issue with missing read-replica in MySQL database

CloudQuery Plugins
1

Hey everyone,

I’m running into an issue with CloudQuery when taking an inventory of Amazon RDS instances using AWS (v19.0.0). I notice a single instance (read-replica) is not showing up in my MySQL (v2.2.1) database. All other RDS instances on other AWS accounts seem to populate just fine, even other read replicas.

By manually populating an AWS RDS describe-db-instances, I was able to see the instance in question, so I don’t think it’s a user permissions issue.

In my MySQL plugin, I am using:

migrate_mode: 'forced'
write_mode: 'overwrite'

Any help would be appreciated! Thanks.

2

Hi @model-mutt :wave:

Thanks, that’s strange, we’ll need to look into this. Could you raise an issue about this on GitHub with any further details you’re able to provide? The full config you’re using for the v19.0.0 AWS plugin and MySQL plugin would be very useful.

I see the primary key on the aws_rds_instances table is ARN; so I’m curious if maybe somehow this read replica has the same ARN as another instance? That’s one possibility, although it sounds unlikely. Is there anything else different about this read replica worth mentioning?

You can open a new issue here: GitHub Issue

3

The only other thing when using debug logging is that the table sync does show the expected count of resources. Now that you mention it, the ARM is the same except at the end where we added -reader. I’ll see if I can add an issue.

4

Thanks, if you could open an issue that would be great!