Bug 2254547: object: improve the error handling for multisite objs #657

	name: Unit tests
	on:
	push:
	tags:
	- v*
	branches:
	- master
	- release-*
	pull_request:
	branches:
	- master
	- release-*
	workflow_dispatch:
	inputs:
	version:
	type: string
	required: false
	description: "Version of jq to install"
	default: "1.7"

	defaults:
	run:
	# reference: https://docs.github.com/en/actions/reference/workflow-syntax-for-github-actions#using-a-specific-shell
	shell: bash --noprofile --norc -eo pipefail -x {0}

	# cancel the in-progress workflow when PR is refreshed.
	concurrency:
	group: ${{ github.workflow }}-${{ github.event_name == 'pull_request' && github.head_ref \|\| github.sha }}
	cancel-in-progress: true

	permissions:
	contents: read

	jobs:
	unittests:
	runs-on: ubuntu-20.04
	if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')"
	steps:
	- name: checkout
	uses: actions/checkout@v3
	with:
	fetch-depth: 0

	- uses: actions/setup-go@v4
	with:
	go-version: "1.20"

	- name: Setup jq
	uses: dcarbone/install-jq-action@v2.0.2
	with:
	version: "${{ inputs.version }}"

	- name: run unit tests
	run: \|
	export ROOK_UNIT_JQ_PATH="$(which jq)"
	GOPATH=$(go env GOPATH) make -j $(nproc) test \| tee output.txt

	- name: check mds liveness probe script ran successfully
	run: \|
	if grep "skipping mds liveness probe script unit tests because jq binary location is not known" output.txt
	then
	echo "jq not found, fail the test"
	exit 1
	fi

Provide feedback