ops: trigger base image rebuild after registry blob loss
This commit is contained in:
parent
cfc5f7583f
commit
fa7fb9daf2
1 changed files with 6 additions and 0 deletions
|
|
@ -75,6 +75,12 @@ FROM ${RUNNER_IMAGE} AS base
|
||||||
# pushes leave the value at its `none` default, so day-to-day base
|
# pushes leave the value at its `none` default, so day-to-day base
|
||||||
# rebuilds (Dockerfile.base edits) don't pay the apt-update cost
|
# rebuilds (Dockerfile.base edits) don't pay the apt-update cost
|
||||||
# unless the week has rolled over.
|
# unless the week has rolled over.
|
||||||
|
#
|
||||||
|
# Note: the registry occasionally GCs base-image blobs that the
|
||||||
|
# manifest still references (saw this 2026-05-04 — `final 1/5 FROM
|
||||||
|
# prop-base:latest` failed with "could not fetch content descriptor
|
||||||
|
# … not found"). When that happens, push any change to this file to
|
||||||
|
# re-trigger the build-base workflow and re-upload the layers.
|
||||||
ARG CACHE_BUST=none
|
ARG CACHE_BUST=none
|
||||||
|
|
||||||
# Runtime apt deps. cdo + gdal-bin are split into a second RUN so a
|
# Runtime apt deps. cdo + gdal-bin are split into a second RUN so a
|
||||||
|
|
|
||||||
Loading…
Add table
Reference in a new issue