Jonathan Sick, Joshua Hoblitt

Related to the JSON generated by post_qa, right now it looks like the example below

{
"ci_id": "177",
"ci_name": "demo",
"ci_url": "https://ci.lsst.codes/job/validate_drp/dataset=cfht,label=centos-7,python=py2/177/",
"date": "2016-07-27T16:32:08.300213+00:00",
"measurements": [

,

ci_name should assume the value validate_drp right?

and I am planning to add

"ci_dataset": "cfht"

and

"ci_label": "centos-7"

to implement the ability of filtering the data by dataset and platform in the dashboard (related to ticket DM-6992).

do you agree with this change?

Here's the line that sets ci_name:

https://github.com/lsst-sqre/post-qa/blob/master/postqa/jenkinsenv.py#L26

In other words, we just need a $DATASET environment variable in the Jenkins environment that indicates the current dataset. I think "ci_name" and your "ci_dataset" are pretty much the same thing?

"ci_label" seems like a good idea. I'll have to find out how I can get that from the Jenkins environment.

Update: Ah, it might make sense for "ci_name" to be the name of the Jenkins project. Again, I'll have to find out how to get that from the environment.

I think so, we could have:

"ci_name" : "validate_drp",
"ci_dataset": "cfht"
"ci_label": "centos-7"

I would say you could use the Jenkins API given the ci_id

https://ci.lsst.codes/job/validate_drp/177/api/json?pretty=true

but "dataset" and "label" only appear enconded in the url...

Can I call the API from inside the Jenkins job though? post-qa runs on Jenkins itself.

oh, this link provides all the enviroment variables that should be available for post_qa

https://ci.lsst.codes/job/validate_drp/177/dataset=cfht,label=centos-7,python=py2/injectedEnvVars/

there's already PRODUCT, dataset and label (but these two should be uppercase)

The changes discussed above will be implemented in a separate ticket DM-7098.