[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: RFR: email about regressions [was: Dealing with ci.d.n for package regressions]



Ian Jackson:
> AFAICT we had consensus that by default both the delayer and the
> delayee should get mails about test failures.  But I don't think that
> is implemented yet.
> 
> I recently found out rather late that a test had failed which was
> important to me.  I want to set up a thing to email me.  I think I can
> do this by screenscraping pages like this one
>   https://ci.debian.net/packages/d/dgit/testing/amd64/
> 
> I have some questions:
> 
> 1. Has anyone implemented this for themselves and if so can I
>    crib yours ?
> 
> 2. Is there a way to get this information that does not involed
>    either (i) screenscraping or (ii) pratting about with API keys
>    for an operation which reads a small amount of public data ?
> 
> Thanks,
> Ian.
> 

Britney generates a machine-readable format that should be useful for
solving this issue. The data file is updated hourly and available from:
https://release.debian.org/britney/excuses.yaml

(Pull between HH:45 - HH:59 if you want the least latency)

You are looking for:

"""
  item-name: gdcm
  [...]
  policy_info:
    age:
      age-requirement: 5
      current-age: 0
      verdict: REJECTED_TEMPORARILY
    autopkgtest:
      camitk/4.0.4-2:
        amd64:
        - ALWAYSFAIL
        - https://ci.debian.net/data/autopkgtest/[...]
        - https://ci.debian.net/packages/c/camitk/testing/amd64
        - null
        - null
      [...]
"""

Where:

 * item-name is the updated source that (potentially) introduces an
   issue

 * policy_info.autopkgtest.dgit/<version> (if present) is the test
   result for dgit.  In this structure ($RESULT), you are looking for
   $RESULT.${ARCH}[0] which contains the test result for dgit on that
   architecture.
   - Please see [the source] for known status labels

Thanks,
~Niels

[the source]:
https://salsa.debian.org/release-team/britney2/blob/master/britney2/policies/autopkgtest.py#L35




Reply to: