[pw-ci] [PATCH v3 1/2] series_db_lib: Prepare to introduce OBS CI

[Michael Santana]

On Fri, Sep 10, 2021 at 2:57 PM Aaron Conole <aconole at redhat.com> wrote:
>
> Hi Michael,
>
> Sorry for the late reply.
>
> Michael Santana <msantana at redhat.com> writes:
>
> > This patch adds a new column, obs_sync, to determine if an entry in the
> > git_builds table was synced in patchwork with OBS build status
> >
> > As part of adding this new table the series_db_lib.sh had to be modified
> > to let CI scripts specify which column is for their CI
> >
> > github_mon was also updated to conform to the new usage
> >
> > Signed-off-by: Michael Santana <msantana at redhat.com>
> > ---
> >  github_mon       |  5 +++--
> >  series_db_lib.sh | 21 +++++++++++++++++----
> >  2 files changed, 20 insertions(+), 6 deletions(-)
> >
>
> snip...
>
> >  function set_synced_for_series() {
> >      local series_id="$1"
> >      local instance="$2"
> > +    local ci_instance="$3"
> >
> >      series_db_exists
> >
> > -    echo "update git_builds set gap_sync=1 where patchwork_instance=\"$instance\" and series_id=$series_id;" | series_db_execute
> > +    echo "update git_builds set gap_sync=1, obs_sync=1 where patchwork_instance=\"$instance\" and series_id=$series_id;" | series_db_execute
> >  }
>
> Reading this - it seems like if we run 'set_synced' it ignores the
> ci_instance variable, and sets all the series types?  That doesn't seem
> right to me.
It doesn't look right at first. set_synced_for_series is only being
called by pw_mon:151 to mark off superseded series. I figured since
the series is superseded it makes sense to mark both github and obs.
Don't confuse set_synced_for_series for set_synced_patch.
set_synced_patch correctly sets the correct column and is what is
being used by github and obs
>
> Did I misread it?
>
> _______________________________________________
> Pwci mailing list
> Pwci at lists.esnacc.org
> http://mail.esnacc.org/mailman/listinfo/pwci
>