tests: fixes for crawl cancel + crawl stopped (#864)

* tests:
- fix cancel crawl test by ensuring state is not running or waiting
- fix stop crawl test by ensuring stop is only initiated after at least one page has been crawled,
otherwise result may be failed, as no crawl data has been crawled yet (separate fix in crawler to avoid loop if stopped
before any data written webrecorder/browsertrix-crawler#314)
- bump page limit to 4 for tests to ensure crawl is partially complete, not fully complete when stopping
- allow canceled or partial_complete due to race condition

* chart: bump frontend limits in default, not just for tests (addresses #780)

* crawl stop before starting:
- if crawl stopped before it started, mark as canceled
- add test for stopping immediately, which should result in 'canceled' crawl
- attempt to increase resync interval for immediate failure
- nightly tests: increase page limit to test timeout

* backend:
- detect stopped-before-start crawl as 'failed' instead of 'done'
- stats: return stats counters as int instead of string

.github/workflows/k3d-nightly-ci.yaml

@@ -62,7 +62,7 @@ jobs:
 
       - name: Start Cluster with Helm
         run: |
-          helm upgrade --install -f ./chart/values.yaml -f ./chart/test/test.yaml --set invite_expire_seconds=10 btrix ./chart/
+          helm upgrade --install -f ./chart/values.yaml -f ./chart/test/test.yaml --set invite_expire_seconds=10 --set max_pages_per_crawl=10 btrix ./chart/
 
       - name: Install Python
         uses: actions/setup-python@v3

backend/btrixcloud/crawls.py

@@ -99,7 +99,7 @@ class Crawl(CrawlConfigCore):
 
     state: str
 
-    stats: Optional[Dict[str, str]]
+    stats: Optional[Dict[str, int]]
 
     files: Optional[List[CrawlFile]] = []
 
@@ -146,7 +146,7 @@ class ListCrawlOut(BaseMongoModel):
 
     state: str
 
-    stats: Optional[Dict[str, str]]
+    stats: Optional[Dict[str, int]]
 
     fileSize: int = 0
     fileCount: int = 0

backend/btrixcloud/operator.py

@@ -409,6 +409,7 @@ class BtrixOperator(K8sAPI):
 
         return True
 
+    # pylint: disable=too-many-branches
     async def update_crawl_state(self, redis, crawl, status, pods):
         """update crawl state and check if crawl is now done"""
         results = await redis.hvals(f"{crawl.id}:status")
@@ -478,8 +479,14 @@ class BtrixOperator(K8sAPI):
 
         # check if all crawlers failed
         if failed >= crawl.scale:
+            # if stopping, and no pages finished, mark as canceled
+            if crawl.stopping and not status.pagesDone:
+                state = "canceled"
+            else:
+                state = "failed"
+
             status = await self.mark_finished(
-                redis, crawl.id, crawl.cid, status, state="failed"
+                redis, crawl.id, crawl.cid, status, state=state
             )
 
         return status

backend/test/conftest.py

@@ -183,7 +183,7 @@ def _crawler_create_config_only(crawler_auth_headers, default_org_id):
         "description": "crawler test crawl",
         "config": {
             "seeds": [{"url": "https://webrecorder.net/"}],
-            "pageExtraDelay": 5,
+            "pageExtraDelay": 20,
             "limit": 4,
         },
     }

backend/test/test_run_crawl.py

@@ -145,8 +145,8 @@ def test_verify_wacz():
     pages = z.open("pages/extraPages.jsonl").read().decode("utf-8")
     assert '"https://webrecorder.net/blog"' in pages
 
-    # 1 other page + header line
-    assert len(pages.strip().split("\n")) == 2
+    # 3 other page + header line
+    assert len(pages.strip().split("\n")) == 4
 
 
 def test_update_crawl(admin_auth_headers, default_org_id, admin_crawl_id):

backend/test/test_settings.py

@@ -13,6 +13,6 @@ def test_settings():
         "registrationEnabled": False,
         "jwtTokenLifetime": 86400,
         "defaultBehaviorTimeSeconds": 300,
-        "maxPagesPerCrawl": 2,
+        "maxPagesPerCrawl": 4,
         "defaultPageLoadTimeSeconds": 120,
     }

backend/test/test_stop_cancel_crawl.py

@@ -48,7 +48,7 @@ def test_cancel_crawl(default_org_id, crawler_auth_headers):
 
     data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
 
-    while data["state"] == "running":
+    while data["state"] in ("running", "waiting"):
         data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
 
     assert data["state"] == "canceled"
@@ -57,8 +57,44 @@ def test_cancel_crawl(default_org_id, crawler_auth_headers):
     assert len(data["resources"]) == 0
 
 
-@pytest.mark.skipif(os.environ.get("CI") is not None, reason="Skip Test on CI")
-def test_start_crawl_to_stop(
+def test_start_crawl_and_stop_immediately(
+    default_org_id, crawler_config_id_only, crawler_auth_headers
+):
+    r = requests.post(
+        f"{API_PREFIX}/orgs/{default_org_id}/crawlconfigs/{crawler_config_id_only}/run",
+        headers=crawler_auth_headers,
+    )
+    assert r.status_code == 200
+    data = r.json()
+
+    crawl_id = data["started"]
+
+    r = requests.post(
+        f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawl_id}/stop",
+        headers=crawler_auth_headers,
+    )
+    data = r.json()
+    assert data["success"] == True
+
+    # test crawl
+    data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
+    assert data["stopping"] == True
+
+    # test workflow
+    r = requests.get(
+        f"{API_PREFIX}/orgs/{default_org_id}/crawlconfigs/{crawler_config_id_only}",
+        headers=crawler_auth_headers,
+    )
+    assert r.json()["currCrawlStopping"] == True
+
+    while data["state"] in ("starting", "running", "waiting"):
+        data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
+
+    assert data["state"] in ("canceled", "partial_complete")
+    assert data["stopping"] == True
+
+
+def test_start_crawl_to_stop_partial(
     default_org_id, crawler_config_id_only, crawler_auth_headers
 ):
     r = requests.post(
@@ -74,12 +110,15 @@ def test_start_crawl_to_stop(
     crawl_id = data["started"]
 
 
-@pytest.mark.skipif(os.environ.get("CI") is not None, reason="Skip Test on CI")
-def test_stop_crawl(default_org_id, crawler_config_id_only, crawler_auth_headers):
+def test_stop_crawl_partial(
+    default_org_id, crawler_config_id_only, crawler_auth_headers
+):
     data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
-    while data["state"] == "starting":
-        time.sleep(5)
+    done = False
+    while not done:
+        time.sleep(2)
         data = get_crawl(default_org_id, crawler_auth_headers, crawl_id)
+        done = data.get("stats") and data.get("stats").get("done") > 0
 
     r = requests.post(
         f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawl_id}/stop",

backend/test_nightly/conftest.py

@@ -232,8 +232,11 @@ def error_crawl_id(admin_auth_headers, default_org_id):
         "runNow": True,
         "name": "Youtube crawl with errors",
         "config": {
-            "seeds": [{"url": "https://www.youtube.com/watch?v=Sh-x3QmbRZc"}],
-            "limit": 10,
+            "seeds": [
+                {"url": "https://invalid.webrecorder.net/"},
+                {"url": "https://invalid-2.webrecorder.net/"},
+            ],
+            "limit": 1,
         },
     }
     r = requests.post(

chart/test/test.yaml

@@ -6,7 +6,7 @@ frontend_pull_policy: "Never"
 
 default_crawl_filename_template: "@ts-testing-@hostsuffix.wacz"
 
-operator_resync_seconds: 5
+operator_resync_seconds: 3
 
 mongo_auth:
   # specify either username + password (for local mongo)
@@ -26,7 +26,7 @@ superuser:
 local_service_port: 30870
 
 # test max pages per crawl global limit
-max_pages_per_crawl: 2
+max_pages_per_crawl: 4
 
 registration_enabled: "0"
 

chart/values.yaml

@@ -94,10 +94,10 @@ frontend_image: "docker.io/webrecorder/browsertrix-frontend:latest"
 frontend_pull_policy: "Always"
 
 frontend_requests_cpu: "3m"
-frontend_limits_cpu: "10m"
+frontend_limits_cpu: "30m"
 
 frontend_requests_memory: "12Mi"
-frontend_limits_memory: "20Mi"
+frontend_limits_memory: "40Mi"
 
 # if set, maps nginx to a fixed port on host machine
 # must be between 30000 - 32767