compute_optimizer_ebs_volume_lines Tag Column is referencing the wrong column [55] should be column 56.

[Doublecountersunk]

In the compute_optimizer_ebs_volume_lines table the schema is set so that the tags is referencing column 55, it is infact column 56. Consequently tags always return true/false on if it is a rootvolume not the expected tag values.

This is it:

cloud-intelligence-dashboards-framework/cid/builtin/core/data/queries/co/ebs_volume.json

Lines 224 to 227 in 7a1c6ba

	{
	"Name": "tags",
	"Type": "string"
	}

It should be [note the following is NOT a duplicate of currentconfiguration_rootvolume, there is a column called rootvolume directly after it in the CSV's and before the tags column]:

                {
                    "Name": "rootVolume",
                    "Type": "string"
                },
                {
                    "Name": "tags",
                    "Type": "string"
                }

This bumps the tags along to column 56 and they then display correctly in the views.

[iakov-aws]

@VoicuAWS I guess we must set fixed set of columns. Can you look in this direction?

[Doublecountersunk]

Yes its one of the few things (in the deployment) that doesn't have a crawlers and relies on a fixed schemas. I think from having a look there are a couple of things which aren't quite right in the compute optimiser space because of this. Lots of CSVs have more columns now than glue reveals (at least now... I suspect more columns have been added at source). Also things like the lambda optimiser CSVs have tags, but the glue entry doesn't have that column mapped, and the query explicitly varchar's it to '' to null it out to allow the union to succeed.