how to specify multiple --extra-config? #58
Comments
jgehrcke
added a commit
to conbench/conbench
that referenced
this issue
Feb 2, 2023
still unclear how to pass multiple kv pairs medyagh/setup-minikube#58
jgehrcke
added a commit
to conbench/conbench
that referenced
this issue
Feb 4, 2023
still unclear how to pass multiple kv pairs medyagh/setup-minikube#58
jgehrcke
added a commit
to conbench/conbench
that referenced
this issue
Feb 4, 2023
still unclear how to pass multiple kv pairs medyagh/setup-minikube#58
jgehrcke
added a commit
to conbench/conbench
that referenced
this issue
Feb 7, 2023
* gha: wip for minikube * gha syntax fix * makefile: build-conbench-container-image, deploy-on-minikube * api/index.py: less verbose on alembic version detection * ci: add minikube dir * ci: minikube: deploy-on-minikube, debug * gha: fix * gha: minikube: extra config for kube-prom * ci: in and out * gha: postgres-operator do not stop&start minikube * debug work * gha: fixes, install kube-prometheus * gha: try multiple extra-config related possibly: medyagh/setup-minikube#40 * show minikube status * more kube-prometheus work * gha: fix extra-config GHA syntax still unclear how to pass multiple kv pairs medyagh/setup-minikube#58 * devdev * kube-prometheus after postgres-operator * reload preload * order of things * resource consumption * fix missing comma * try this * medyagh/setup-minikube: too many errors as of outdated release branch * minikube stack: consume less resources on GHA the prometheus operator does not deploy 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod.. that is with minikube: (CPUs=2, Memory=2200MB) * minikube: fix resource request X Exiting due to RSRC_INSUFFICIENT_CORES: Requested cpu count 4 is greater than the available cpus of 2 * medyagh/setup-minikube@master is broken * try 0. there was still too much going on. * dir /tmp/_conbench-promcl-coord-dir must exist in container * more fixes * fixes * try to fix service, servicemonitor port relationships Also see https://stackoverflow.com/a/53013006 * better sleeps, fixes * code tweaks * smarter waiting * gha: fix dir * better comments, tiny fixes * reduce scrape interval * add comments * fix: build image before doing make deploy-on-minikube * db.py: try connecting to db for about one minute * better waiting * do not wait for prometheus-k8s-1 * makefile: add minikube targets (wip) * test-conbench-on-mk.sh: cleanup * reactivate most CI jobs * cleanup * test-conbench-on-mk.sh: cleanup * clean up __init__.py * db.py: catch another variant of 'already exists' Seemingly sqlalchemy.exc.ProgrammingError can also be thrown depending on the exact timing of racers. * gha: fix bash: command not found * makefile: remove info output elsewhere we rely on `make -s ...` to have meaningful output. the info lines mess with that. * one more sleep for less interleaved cmd output * docker-compose.yml: dynamic port on host The containerized stack does not need specific ports on the host system. Postgres and Dex do not really need to be accessed from the host anyway. This fixes errors like these: Bind for 0.0.0.0:5556 failed: port is already allocated Bind for 0.0.0.0:5432 failed: port is already allocated * gunicorn-conf: better comment * init: log info about PROMETHEUS_MULTIPROC_DIR Got an error raised by prometheus_flask_exporter/multiprocess.py in CI and need more info about what's really going on: is the env var set or not set? does the directory exist? * Makefile: better cleanup via --remove-orphans Locally some orphaned resources may pile up, and this tears those down. * init: remove empty comment line * makefile: test-run-app-dev: create /tmp/_conbench-promcl-coord-dir The conbench web app container image has this special directory set up as of: RUN mkdir -p /tmp/_conbench-promcl-coord-dir When mounting /tmp to /tmp we overwrite this. In that case this must exist on the host tmpfs before invoking gunicorn. * Makefile: cleanup * Dockerfile: better comment * test-conbench-on-mk: clean up comments * comment cleanup * Makefile: remove what is not needed in this PR * Makefile: run-app-dev: remove port 8000 notion again * Makefile: set-build-info before build-conbench-container-image * conbench/__init__.py: note that the CLI runs this, too
|
Hi @jgehrcke, thanks for bringing this to our attention, you are correct that passing multiple extra-config fields does not work. In the meantime you can accomplish this using the |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This does not work
Results in
The text was updated successfully, but these errors were encountered: