Trying to split testing in GHA and use docker buildx to have cache of…

… builds (#108) * First stuff -trying to get bake to work in gha and local testing * Set uid,gid in gha * Bake yml separate from run yml * Test echo output * Test echo output 2 * Test if using buildx bake in other job leads to caching from prev job where it is also used * Try registry instead of gha for cache * Login to ghcr first * Login to ghcr first2 * Also use cache in unit tests, write docker cmds in gha yml * Also use cache in unit tests, write docker cmds in gha yml - typo * Fix unittest bake thing * Clean a bit, try to pull images also in the bake step for caching * Increase sleep time to see if that makes jobs complete on gha * Put timeouts in place to see what is wrong on gha * More sleep * Try to get stdout for erroring jobs on specific gha * Show logs of tasks to find out if task runs at all * stupid typo * Put rsync key in docker image so file permissions are correct - for testing * Fix key into right, readable location * Decrease sleep times a bit, catch errors to show and then fail, and explicitly check last line so it wont be a wait in case of slow operation or something
lehtiolab · Oct 30, 2024 · a4122f6 · a4122f6
1 parent 3598dc2
commit a4122f6
Show file tree

Hide file tree

Showing 11 changed files with 226 additions and 108 deletions.
diff --git a/.github/run_tests_integration.sh b/.github/run_tests_integration.sh
diff --git a/.github/run_tests_unit.sh b/.github/run_tests_unit.sh
diff --git a/.github/workflows/integration-tests.yml b/.github/workflows/integration-tests.yml
@@ -1,5 +1,5 @@
-name: integration-tests
-run-name: Run integration tests
+name: lint-and-test
+run-name: Run linting and tests
 on:
     push:
         branches:
@@ -12,21 +12,76 @@ jobs:
         runs-on: ubuntu-latest
         steps:
             - uses: actions/checkout@v4
+
             - name: Setup buildx
               uses: docker/setup-buildx-action@v3
-            - run: docker buildx bake --file src/docker/docker-compose-testing.yml --file src/docker/docker-compose-gha-cache.json
-            - run: bash run_lint.sh
+
+            - name: Get USER_ID to env
+              run: echo "USER_ID=$(id -u)" >> $GITHUB_ENV
+
+            - name: Get GROUP_ID to env
+              run: echo "GROUP_ID=$(id -g)" >> $GITHUB_ENV
+
+            - name: Login to Docker Registry, bake/cache
+              uses: docker/login-action@v3
+              with:
+                registry: ghcr.io
+                username: ${{ github.actor }}
+                password: ${{ secrets.GITHUB_TOKEN }}
+
+            - run: docker buildx bake --file src/docker/testing-bake.yml --file src/docker/docker-compose-gha-cache.json
+            - run: |
+                docker compose --env-file src/docker/.compose.testing.env \
+                -f src/docker/docker-compose-testing.yml run web pylint \
+                -E --disable E1101,E0307 --ignore-paths \
+                '.*\/migrations\/[0-9]+.*.py' analysis datasets dashboard home jobs kantele rawstatus
 
     integration-tests:
         runs-on: ubuntu-latest
         needs: linting
         steps:
             - uses: actions/checkout@v4
-            - run: bash .github/run_tests_integration.sh
+            - name: Setup buildx
+              uses: docker/setup-buildx-action@v3
+            - name: Get USER_ID to env
+              run: echo "USER_ID=$(id -u)" >> $GITHUB_ENV
+            - name: Get GROUP_ID to env
+              run: echo "GROUP_ID=$(id -g)" >> $GITHUB_ENV
+            - name: Login to Docker Registry to get bake cache
+              uses: docker/login-action@v3
+              with:
+                registry: ghcr.io
+                username: ${{ github.actor }}
+                password: ${{ secrets.GITHUB_TOKEN }}
+            - run: docker buildx bake --file src/docker/testing-bake.yml --file src/docker/docker-compose-gha-cache.json
+            - run: |
+                docker compose --env-file  src/docker/.compose.testing.env -f src/docker/docker-compose-testing.yml \
+                  up --detach db mq
+                sleep 5 # DB needs to be up or app crashes trying to connect
+                docker compose --env-file  src/docker/.compose.testing.env -f src/docker/docker-compose-testing.yml \
+                  run --use-aliases web python manage.py test --tag slow --exclude-tag mstulos || docker compose --env-file src/docker/.compose.testing.env -f src/docker/docker-compose-testing.yml logs storage_mvfiles storage_downloads
 
     unit-tests:
         runs-on: ubuntu-latest
         needs: linting
         steps:
             - uses: actions/checkout@v4
-            - run: bash .github/run_tests_unit.sh
+            - name: Setup buildx
+              uses: docker/setup-buildx-action@v3
+            - name: Get USER_ID to env
+              run: echo "USER_ID=$(id -u)" >> $GITHUB_ENV
+            - name: Get GROUP_ID to env
+              run: echo "GROUP_ID=$(id -g)" >> $GITHUB_ENV
+            - name: Login to Docker Registry to get bake cache
+              uses: docker/login-action@v3
+              with:
+                registry: ghcr.io
+                username: ${{ github.actor }}
+                password: ${{ secrets.GITHUB_TOKEN }}
+            - run: docker buildx bake --file src/docker/testing-bake.yml --file src/docker/docker-compose-gha-cache.json
+            - run: |
+                docker compose --env-file  src/docker/.compose.testing.env -f src/docker/docker-compose-testing.yml \
+                  up --detach db mq
+                sleep 5 # DB needs to be up or app crashes trying to connect
+                docker compose --env-file  src/docker/.compose.testing.env -f src/docker/docker-compose-testing.yml \
+                  run --use-aliases web python manage.py test --exclude-tag slow
diff --git a/src/backend/rawstatus/tests.py b/src/backend/rawstatus/tests.py
@@ -107,7 +107,7 @@ def test_new_file(self):
         old_raw = rm.RawFile.objects.last()
         sp = self.run_script(fullp)
         # Give time for running script, so job is created before running it etc
-        sleep(1)
+        sleep(2)
         new_raw = rm.RawFile.objects.last()
         self.assertEqual(new_raw.pk, old_raw.pk + 1)
         sf = rm.StoredFile.objects.last()
@@ -117,20 +117,31 @@ def test_new_file(self):
         self.assertFalse(sf.checked)
         self.run_job()
         sf.refresh_from_db()
+        try:
+            spout, sperr = sp.communicate(timeout=10)
+        except subprocess.TimeoutExpired:
+            sp.terminate()
+            # Properly kill children since upload.py uses multiprocessing
+            os.killpg(os.getpgid(sp.pid), signal.SIGTERM)
+            spout, sperr = sp.communicate()
+            print(sperr.decode('utf-8'))
+            raise
         self.assertTrue(sf.checked)
-        spout, sperr = sp.communicate()
         explines = ['Registering 1 new file(s)', 
                 f'File {new_raw.name} matches remote file {new_raw.name} with ID {new_raw.pk}',
                 f'Checking remote state for file {new_raw.name} with ID {new_raw.pk}',
                 f'State for file {new_raw.name} with ID {new_raw.pk} was: transfer',
                 f'Uploading {fullp} to {self.live_server_url}',
                 f'Succesful transfer of file {fullp}',
                 f'Checking remote state for file {new_raw.name} with ID {new_raw.pk}',
-                f'State for file with ID {new_raw.pk} was "done"']
-        for out, exp in zip(sperr.decode('utf-8').strip().split('\n'), explines):
+                ]
+        outlines = sperr.decode('utf-8').strip().split('\n')
+        for out, exp in zip(outlines, explines):
             out = re.sub('.* - INFO - .producer.worker - ', '', out)
             out = re.sub('.* - INFO - root - ', '', out)
             self.assertEqual(out, exp)
+        lastexp = f'State for file with ID {new_raw.pk} was "done"'
+        self.assertEqual(re.sub('.* - INFO - .producer.worker - ', '', outlines[-1]), lastexp)
 
     def test_transfer_again(self):
         '''Transfer already existing file, e.g. overwrites of previously
@@ -154,7 +165,15 @@ def test_transfer_again(self):
         self.f3sf.refresh_from_db()
         self.assertFalse(self.f3sf.checked)
         self.run_job()
-        spout, sperr = sp.communicate()
+        try:
+            spout, sperr = sp.communicate(timeout=10)
+        except subprocess.TimeoutExpired:
+            sp.terminate()
+            # Properly kill children since upload.py uses multiprocessing
+            os.killpg(os.getpgid(sp.pid), signal.SIGTERM)
+            spout, sperr = sp.communicate()
+            print(sperr.decode('utf-8'))
+            raise
         self.f3sf.refresh_from_db()
         self.assertTrue(self.f3sf.checked)
         self.assertEqual(self.f3sf.md5, self.f3raw.source_md5)
@@ -166,11 +185,14 @@ def test_transfer_again(self):
                 f'Uploading {fullp} to {self.live_server_url}',
                 f'Succesful transfer of file {fullp}',
                 f'Checking remote state for file {self.f3raw.name} with ID {self.f3raw.pk}',
-                f'State for file with ID {self.f3raw.pk} was "done"']
-        for out, exp in zip(sperr.decode('utf-8').strip().split('\n'), explines):
+                ]
+        outlines = sperr.decode('utf-8').strip().split('\n')
+        for out, exp in zip(outlines, explines):
             out = re.sub('.* - INFO - .producer.worker - ', '', out)
             out = re.sub('.* - INFO - root - ', '', out)
             self.assertEqual(out, exp)
+        lastexp = f'State for file with ID {self.f3raw.pk} was "done"'
+        self.assertEqual(re.sub('.* - INFO - .producer.worker - ', '', outlines[-1]), lastexp)
 
     def test_transfer_same_name(self):
         # Test trying to upload file with same name/path but diff MD5
@@ -232,7 +254,15 @@ def test_transfer_file_namechanged(self):
         sp = self.run_script(fullp)
         sleep(1)
         self.run_job()
-        spout, sperr = sp.communicate()
+        try:
+            spout, sperr = sp.communicate(timeout=10)
+        except subprocess.TimeoutExpired:
+            sp.terminate()
+            # Properly kill children since upload.py uses multiprocessing
+            os.killpg(os.getpgid(sp.pid), signal.SIGTERM)
+            spout, sperr = sp.communicate()
+            print(sperr.decode('utf-8'))
+            raise
         newsf = rm.StoredFile.objects.last()
         self.assertEqual(newsf.pk, lastsf.pk + 1)
         self.assertEqual(rawfn.pk, newsf.rawfile_id)
@@ -272,7 +302,15 @@ def test_rsync_not_finished_yet(self):
         job.save()
         self.f3sf.checked = True
         self.f3sf.save()
-        spout, sperr = sp.communicate()
+        try:
+            spout, sperr = sp.communicate(timeout=10)
+        except subprocess.TimeoutExpired:
+            sp.terminate()
+            # Properly kill children since upload.py uses multiprocessing
+            os.killpg(os.getpgid(sp.pid), signal.SIGTERM)
+            spout, sperr = sp.communicate()
+            print(sperr.decode('utf-8'))
+            raise
         explines = [f'Checking remote state for file {self.f3raw.name} with ID {self.f3raw.pk}',
                 f'State for file {self.f3raw.name} with ID {self.f3raw.pk} was: wait',
                 f'Checking remote state for file {self.f3raw.name} with ID {self.f3raw.pk}',

diff --git a/src/docker/Dockerfile b/src/docker/Dockerfile
@@ -26,8 +26,9 @@ FROM django AS django_test
 USER root
 RUN apt-get install -y rsync
 RUN pip install pylint
+RUN mkdir /kantelessh && chmod 700 /kantelessh && chown kantele: /kantelessh
 USER kantele
-RUN mkdir /home/kantele/.ssh && chmod 700 /home/kantele/.ssh
+COPY docker/test_rsync_sshkey /kantelessh/rsync_key
 
 # Build django static for the nginx prod container
 FROM django AS django_static
@@ -36,7 +37,7 @@ RUN python manage.py collectstatic --no-input
 
 # Compile node frontend stuff for prod container
 FROM node:18 AS build_frontend
-COPY ./frontend /src/frontend
+COPY frontend /src/frontend
 RUN cd /src/frontend/analysis && npm install && npm run build
 RUN cd /src/frontend/dashboard && npm install && npm run build
 RUN cd /src/frontend/datasets && npm install && npm run build

diff --git a/src/docker/docker-compose-dev.yml b/src/docker/docker-compose-dev.yml
@@ -2,29 +2,30 @@
 #
 
 services:
-    storage:
-        image: django_test
-        build:
-            context: "${BUILD_CONTEXT_DEV:-../}"
-            dockerfile: ./docker/Dockerfile
-            target: django_test
-            args:
-                USER_ID: "${USER_ID:-You must run export USER_ID}"
-                GROUP_ID: "${GROUP_ID:-You must run export GROUP_ID}"
-        volumes:
-            - ../../data/storage:/storage
-            - ../../data/newstorage:/newstorage
-            - ../../data/analysisfiles:/analysisfiles
-            - ../backend:/kantele
-        environment:
-            APIKEY: "${APIKEY_STORAGE}"
-            KANTELEHOST: nginx
-            PROTOCOL: http
-            TMPSHARE: '/s3storage/tmp'
-            ANALYSISSHARE: /analysisfiles
-            RABBITHOST: mq
-            RABBIT_VHOST: "${RABBIT_VHOST}"
-            RABBITUSER: "${RABBITUSER:-guest}"
-            RABBITPASS: "${RABBITPASS:-guest}"
-            STORAGESHARES: '/storage,/newstorage'
-            PRIMARY_STORAGE: s3storage
+  storage:
+    image: django_test
+    build:
+      context: "${BUILD_CONTEXT_DEV:-../}"
+        #context: ../../
+      dockerfile: ./docker/Dockerfile
+      target: django_test
+      args:
+        USER_ID: "${USER_ID:-You must run export USER_ID}"
+        GROUP_ID: "${GROUP_ID:-You must run export GROUP_ID}"
+    volumes:
+      - ../../data/storage:/storage
+      - ../../data/newstorage:/newstorage
+      - ../../data/analysisfiles:/analysisfiles
+      - ../backend:/kantele
+    environment:
+      APIKEY: "${APIKEY_STORAGE}"
+      KANTELEHOST: nginx
+      PROTOCOL: http
+      TMPSHARE: '/s3storage/tmp'
+      ANALYSISSHARE: /analysisfiles
+      RABBITHOST: mq
+      RABBIT_VHOST: "${RABBIT_VHOST}"
+      RABBITUSER: "${RABBITUSER:-guest}"
+      RABBITPASS: "${RABBITPASS:-guest}"
+      STORAGESHARES: '/storage,/newstorage'
+      PRIMARY_STORAGE: s3storage
diff --git a/src/docker/docker-compose-gha-cache.json b/src/docker/docker-compose-gha-cache.json
@@ -1,37 +1,37 @@
 {"target": {
 	"web": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-web:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-web:cache"],
 		"output": ["type=docker"]
 	},
 	"db": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-db:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-db:cache"],
 		"output": ["type=docker"]
 	},
 	"mq": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-mq:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-mq:cache"],
 		"output": ["type=docker"]
 	},
 	"storage_mvfiles": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
 		"output": ["type=docker"]
 	},
 	"tulos_ingester": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
 		"output": ["type=docker"]
 	},
 	"storage_downloads": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-stor:cache"],
 		"output": ["type=docker"]
 	},
 	"upload_bay_rsync": {
-		"cache-from": ["type=gha"],
-		"cache-to": ["type=gha"],
+		"cache-from": ["type=registry,ref=ghcr.io/glormph/kantele-rsync:cache"],
+		"cache-to": ["type=registry,ref=ghcr.io/glormph/kantele-rsync:cache"],
 		"output": ["type=docker"]
 	}
 }

diff --git a/src/docker/docker-compose-prod.yml b/src/docker/docker-compose-prod.yml
@@ -5,7 +5,7 @@ services:
             - db
             - mq
         env_file:
-            - ./prod-container.env
+            - ./src/docker/prod-container.env
         logging:
             driver: syslog
             options:
@@ -24,8 +24,8 @@ services:
             - web
         image: kantele_nginx
         build:
-            context: ../
-            dockerfile: ./docker/Dockerfile
+            context: ./
+            dockerfile: ./src/docker/Dockerfile
             target: nginx_prod
             args:
                 USER_ID: "${USER_ID:-You must run export USER_ID}"
@@ -66,7 +66,7 @@ services:
             - mq
         network_mode: host
         env_file:
-            - ./prod-container.env
+            - ./src/docker/prod-container.env
         logging:
             driver: syslog
             options: