[FAI-14747] Support ingesting deployments data from bigquery #1892
Conversation
meldiner
requested review from
cjwooo,
ypc-faros and
willmarks
as code owners
|
| build: { | ||
| uid: buildUid, | ||
| }, | ||
| changeset: { |
There was a problem hiding this comment.
shouldn't those line be removed since we add the DeploymentChangeset object below?
| if (uid) { | ||
| res.push({ | ||
| model: 'cicd_Deployment', | ||
| record: { |
There was a problem hiding this comment.
typically, for the changeset feed to work, we need the application; i understand we may not have it in the example, but maybe we should have support for it?
| const res: DestinationRecord[] = []; | ||
|
|
||
| //cicd_Deployment | ||
| const uid = record.record.data.uid; |
There was a problem hiding this comment.
are we documenting anywhere what is the structure of the expected record?
| }, | ||
| changeset: { | ||
| commit: { | ||
| sha: commit, |
There was a problem hiding this comment.
Typically, the commit sha of the deployment goes in a different object: vcs_Commit through a cicd_ArtifactCommitAssociation object through a dummy cicd_Artifact object through a dummy cicd_ArtifactDeployment object
See what the changeset feed queries to find it here: https://github.com/faros-ai/feeds/blob/master/feeds/cicd/changeset-feed/resources/graphql/deployments.gql#L28
| }); | ||
|
|
||
| res.push({ | ||
| model: 'cicd_DeploymentChangeset', |
There was a problem hiding this comment.
assuming you write the commit sha through a vcs_Commit object as stated above, I would remove this and let the changeset feed set it
No description provided.