chore: add weekly ci (#4176)

.github/workflows/weekly.yml

@@ -0,0 +1,54 @@
+name: Weekly tests
+
+permissions:
+  contents: read
+
+concurrency:
+  group: weekly-${{ github.ref }}
+  cancel-in-progress: true
+
+on:
+  schedule:
+    - cron: "0 0 * * 1"
+  workflow_dispatch:
+
+jobs:
+  weekly_longevity:
+    name: Weekly Longevity test
+    runs-on: ${{ matrix.os }}
+    timeout-minutes: 120
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        rust-target: [x86_64-unknown-linux-musl]
+    steps:
+      - uses: actions/checkout@v4
+      - name: Setup Fluvio
+        uses: infinyon/fluvio/.github/actions/setup-fluvio@master
+        with:
+          version: latest
+      - name: Run Resume test
+        timeout-minutes: 60
+        run: |
+           FLUVIO_BIN=fluvio INTERACTIONS=100 make resume-test
+      - name: Run diagnostics
+        if: failure()
+        timeout-minutes: 5
+        run: fluvio cluster diagnostics
+      - name: Upload diagnostics
+        uses: actions/upload-artifact@v4
+        timeout-minutes: 5
+        if: failure()
+        with:
+          name: weekly-longevity.diag
+          path: diagnostics*.gz
+      - name: Slack Notification
+        uses: 8398a7/action-slack@v3
+        if: ${{ !success() }}
+        with:
+          status: ${{ job.status }}
+          fields: repo,message,commit,author,action,eventName,ref,workflow,job
+        env:
+          SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}
+          MATRIX_CONTEXT: ${{ toJson(matrix) }}

makefiles/test.mk

@@ -187,6 +187,16 @@ upgrade-test: build-cli build_k8_image
 	./tests/upgrade-test.sh
 endif
 
+ifeq (${CI},true)
+# In CI, we expect all artifacts to already be built and loaded for the script
+resume-test:
+	./tests/local-resume-test.sh
+else
+# When not in CI (i.e. development), load the dev k8 image before running test
+resume-test: build-cli
+	./tests/local-resume-test.sh
+endif
+
 # When running in development, might need to run `cargo clean` to ensure correct fluvio binary is used
 validate-release-stable:
 	./tests/fluvio-validate-release.sh $(VERSION) $(GIT_COMMIT)

tests/local-resume-test.sh

@@ -0,0 +1,55 @@
+#!/bin/bash
+
+readonly INTERACTIONS=${INTERACTIONS:-5}
+
+function run_test() {
+	${FLUVIO_BIN} cluster delete --force
+	${FLUVIO_BIN} cluster start --local
+
+	seq 1 10 | parallel -j 10 ${FLUVIO_BIN} topic create test-topic-{}
+	seq 1 10 | parallel -j 10 ${FLUVIO_BIN} remote register test-remote-{}
+
+	${FLUVIO_BIN} cluster shutdown
+	${FLUVIO_BIN} cluster resume
+
+	# Create topic
+	${FLUVIO_BIN} topic create test-topic-11 # THIS WAS HANGING
+
+	TOPIC_LIST=$(${FLUVIO_BIN} topic list 2>/dev/null)
+	PARTITION_LIST=$(${FLUVIO_BIN} partition list 2>/dev/null)
+	REMOTE_LIST=$(${FLUVIO_BIN} remote list 2>/dev/null)
+
+	# Check if the topic list has 11+1 lines
+	if [ $(echo "$TOPIC_LIST" | wc -l) -eq 12 ]; then
+	    echo "PASS"
+	else
+	    echo "FAIL"
+	    exit 1
+	fi
+
+	# Check if the partition list has 11+1 lines
+	if [ $(echo "$PARTITION_LIST" | wc -l) -eq 12 ]; then
+	    echo "PASS"
+	else
+	    echo "FAIL"
+	    exit 1
+	fi
+
+	# Check if the remote list has 10+1 lines
+	if [ $(echo "$REMOTE_LIST" | wc -l) -eq 11 ]; then
+	    echo "PASS"
+	else
+	    echo "FAIL"
+	    exit 1
+	fi
+}
+
+function main() {
+	for i in $(seq 1 $INTERACTIONS);
+	do
+		echo "INTERATION: $i"
+		run_test
+	done
+}
+
+main;