Keep earlier builds recoverable across region image upgrades

Complete the maintainer's server image bump to edb39f by also setting
the region build identity, and add REGION_RETAINED so the host keeps
the builds that wrote its history on its recovery allowlist; without it
the upgraded host would quarantine its own head as unsupported. Record
the M8.4 drill (rollout 3.0 s, forced deletion 19.3 s to Ready, epochs
3 to 5) and the maintainer's M8.5 sync.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
dingo
2026-09-30 10:34:03 -04:00
co-authored by Claude Opus 5.5
parent bcf2eacb5e
commit 4ef0fe8167
8 changed files with 51 additions and 7 deletions
+8
View File
@@ -226,6 +226,14 @@ for env in ('dev', 'stage'):
if v.get('name') == 'REGION_BUILD']
check(len(server) == 1 and len(builds) == 2 and all('sha256:' + b == server[0] for b in builds),
f'{overlay}: the build identity must equal the server image digest')
retained = [v.get('value', '') for d in manifests.get(overlay + 'region-values.yaml', [])
for c in d['spec']['template']['spec']['containers'] for v in c.get('env', [])
if v.get('name') == 'REGION_RETAINED']
for value in retained:
entries = value.split()
check(all(__import__('re').fullmatch(r'[0-9a-f]{64}', e) for e in entries)
and not set(entries) & set(builds),
f'{overlay}: REGION_RETAINED must list other builds as 64 hex digits')
# M8.5 monitoring (ADR-0072): third-party images are pinned by digest, and Prometheus
# needs no Kubernetes API access.