#7541 Code clean-up in push-two-week-atomic script
Merged 5 years ago by mohanboddu. Opened 5 years ago by sinnykumari.
sinnykumari/releng master  into  master

@@ -97,36 +97,6 @@ 

  ATOMIC_HOST_FEDMSG_MODNAME = "releng"

  ATOMIC_HOST_FEDMSG_CERT_PREFIX = "releng"

  

- MARK_ATOMIC_HOST_BAD_COMPOSES = None

- MARK_ATOMIC_HOST_GOOD_COMPOSES = None

- BLOCK_ATOMIC_HOST_RELEASE = None

- 

- try:

-     MARK_ATOMIC_HOST_BAD_JSON_URL = \

-         'https://pagure.io/mark-atomic-bad/raw/master/f/bad-composes.json'

-     MARK_ATOMIC_HOST_BAD_JSON = requests.get(MARK_ATOMIC_HOST_BAD_JSON_URL).text

-     MARK_ATOMIC_HOST_BAD_COMPOSES = json.loads(MARK_ATOMIC_HOST_BAD_JSON)[u'bad-composes']

- 

-     BLOCK_ATOMIC_HOST_RELEASE_JSON_URL = \

-         'https://pagure.io/mark-atomic-bad/raw/master/f/block-release.json'

-     BLOCK_ATOMIC_HOST_RELEASE_JSON = \

-         requests.get(BLOCK_ATOMIC_HOST_RELEASE_JSON_URL).text

-     BLOCK_ATOMIC_HOST_RELEASE = \

-         json.loads(BLOCK_ATOMIC_HOST_RELEASE_JSON)[u'block-release']

- 

-     MARK_ATOMIC_HOST_GOOD_URL = \

-         'https://pagure.io/mark-atomic-bad/raw/master/f/good-composes.json'

-     MARK_ATOMIC_HOST_GOOD_JSON = \

-         requests.get(MARK_ATOMIC_HOST_GOOD_URL).text

-     MARK_ATOMIC_HOST_GOOD_COMPOSES = \

-         json.loads(MARK_ATOMIC_HOST_GOOD_JSON)[u'good-composes']

- except Exception as e:

-     log.exception(

-         "!!!!{0}!!!!\n{1}".format("Failed to fetch or parse json", e)

-     )

-     sys.exit(1)

- 

- 

  DATAGREPPER_URL = "https://apps.fedoraproject.org/datagrepper/raw"

  # delta = 2 weeks in seconds

  DATAGREPPER_DELTA = 1209600
@@ -284,23 +254,6 @@ 

  

      return release_artifacts_info

  

- def compose_manually_marked_bad(compose_id, bad_composes=MARK_ATOMIC_HOST_BAD_COMPOSES):

-     """

-     compose_manually_marked_bad

- 

-         Check for a compose that has been marked bad manually

- 

-         compose_id

-             Compose id of most recently found auto-tested good compose build

- 

-     return -> bool

-         True if the build was marked bad, else False

-     """

- 

-     bad = [c for c in bad_composes if c == compose_id]

- 

-     return len(bad) > 0

- 

  def send_atomic_announce_email(

          email_filelist,

          ostree_commit_data,
@@ -712,11 +665,6 @@ 

      if not pargs.ostree_pungi_compose_id:

          pargs.ostree_pungi_compose_id = pargs.pungi_compose_id

  

-     log.info("Checking to make sure release is not currently blocked")

-     if BLOCK_ATOMIC_HOST_RELEASE:

-         log.info("Release Blocked: Exiting.")

-         sys.exit(0)

- 

      log.info("Fetching images information for Compose ID %s", pargs.pungi_compose_id )

      # Get image artifacts information for given Pungi Compose ID

      release_artifacts_info = get_release_artifacts_info_from_compose(
@@ -871,7 +819,7 @@ 

                                 ostree_commit_data,

                                 mail_receivers=mail_receivers)

  

-     # FIXME - The logic in this functioni is broken, leave it disabled for now

+     # FIXME - The logic in this function is broken, leave it disabled for now

      #log.info("Pruning old Atomic Host test composes")

      #prune_old_composes(ATOMIC_HOST_STABLE_BASEDIR, 2)

  

Remove logic for manually marking Atomic Host release
as bad, good or release blocking. They are no longer used
since Fedora 27 Atomic Two Week Release

Signed-off-by: Sinny Kumari sinny@redhat.com

While doing code clean-up, I came across prune_old_compose() function, which was used in past for pruning old composes during atomic twoweek release process. We disabled pruning 2 years back in commit.

Do we have plan to enable this again during FAH twoweek release?

the code LGTM.

Do we have plan to enable this again during FAH twoweek release?

No plans exactly. Although I would prefer to keep more than just two (something like 10 would be cool so that we can get 20 weeks worth of releases). Here is what is in that directory now: https://kojipkgs.fedoraproject.org/pub/alt/atomic/stable/

the code LGTM.

Do we have plan to enable this again during FAH twoweek release?

No plans exactly. Although I would prefer to keep more than just two (something like 10 would be cool so that we can get 20 weeks worth of releases). Here is what is in that directory now: https://kojipkgs.fedoraproject.org/pub/alt/atomic/stable/

+1

Pull-Request has been merged by mohanboddu

5 years ago
Metadata