Fix APM Server managed by Elastic Agent setup #1078
Conversation
💚 Build Succeeded
Expand to view the summary
Build stats
Test stats 🧪
Trends 🧪 |
| @@ -1185,7 +1186,7 @@ def package_registry_url(options): | |||
| return url | |||
| if options.get("enable_package_registry"): | |||
| return "http://package-registry:{}".format(PackageRegistry.SERVICE_PORT) | |||
| elif options.get("snapshot"): | |||
| elif options.get("snapshot") or not options.get("release"): | |||
There was a problem hiding this comment.
@axw this changes the default to snapshot, which will serve snapshot+staging+production packages; just wanted to get your 👀 on this real quick.
There was a problem hiding this comment.
I suppose it's fine. My thinking was that we would rely on the logic in Kibana to select the appropriate registry, though to be fair we were still diverging by selecting the snapshot registry on snapshot stack images. But we can always select the staging repo if needed, so 👍
| @@ -1185,7 +1186,7 @@ def package_registry_url(options): | |||
| return url | |||
| if options.get("enable_package_registry"): | |||
| return "http://package-registry:{}".format(PackageRegistry.SERVICE_PORT) | |||
| elif options.get("snapshot"): | |||
| elif options.get("snapshot") or not options.get("release"): | |||
There was a problem hiding this comment.
I suppose it's fine. My thinking was that we would rely on the logic in Kibana to select the appropriate registry, though to be fair we were still diverging by selecting the snapshot registry on snapshot stack images. But we can always select the staging repo if needed, so 👍
…rt-from-build * upstream/master: (24 commits) Add mergify (elastic#1089) fix: update flake8 and deps to work with Python 3.8+ (elastic#1085) chore: bump elastic stack (elastic#1087) fix: force to activate venv (elastic#1086) fix: --opbeans-node-agent-local-repo=/a/local/dir (elastic#1084) ci: jjbb remove periodic-folder-trigger (elastic#1079) Fix APM Server managed by Elastic Agent setup (elastic#1078) docker/opbeans/go: use -mod=mod (elastic#1077) [fleet] fix elastic-agent setup (elastic#1076) update supported stack versions (elastic#1075) [CI] Increase timeout (elastic#1070) Add --package-registry-url flag (elastic#1069) Create opbeans_user/role with write/read access for the opbeans-python (elastic#1063) reset postgres sequences after database migration (elastic#1062) chore: bump Elastic Stack (elastic#1060) docker/go/nethttp: go build -mod=mod (elastic#1061) check agent liveness via version subcommand (elastic#1059) test: use filebeat step to grab Docker logs (elastic#1029) Update dotnet sdk 5.0 docker image (elastic#1048) docs: using Dyno (elastic#1056) ...
The latest docker builds for Kibana and Elastic Agent for
masterand7.13are working again; the apm-integration testing setup needs to be adjusted to the latest fixes.There seems to be something up with the production package registry - it is not serving an APM package. Snapshot registry works fine.
This PR includes:
--snapshotflag is provided, rather than production one.