operator: adjust state transition rules to ensure 'running' state alw…

…ays accounted for in db (#1989) don't rely on current status, always set state to running when running to ensure idempotency in case of multiple calls
webrecorder · Aug 5, 2024 · 4a2725a · 4a2725a
1 parent 1c153df
commit 4a2725a
Showing 1 changed file with 17 additions and 19 deletions.
diff --git a/backend/btrixcloud/operator/crawls.py b/backend/btrixcloud/operator/crawls.py
@@ -807,27 +807,25 @@ async def sync_crawl_state(
                 status.resync_after = self.fast_retry_secs
                 return status
 
-            # set state to running (if not already)
-            if status.state not in RUNNING_STATES:
-                # if true (state is set), also run webhook
-                if await self.set_state(
-                    "running",
-                    status,
-                    crawl,
-                    allowed_from=["starting", "waiting_capacity"],
-                ):
-                    if not crawl.qa_source_crawl_id:
-                        self.run_task(
-                            self.event_webhook_ops.create_crawl_started_notification(
-                                crawl.id, crawl.oid, scheduled=crawl.scheduled
-                            )
+            # if true (state is set), also run webhook
+            if await self.set_state(
+                "running",
+                status,
+                crawl,
+                allowed_from=["starting", "waiting_capacity"],
+            ):
+                if not crawl.qa_source_crawl_id:
+                    self.run_task(
+                        self.event_webhook_ops.create_crawl_started_notification(
+                            crawl.id, crawl.oid, scheduled=crawl.scheduled
                         )
-                    else:
-                        self.run_task(
-                            self.event_webhook_ops.create_qa_analysis_started_notification(
-                                crawl.id, crawl.oid, crawl.qa_source_crawl_id
-                            )
+                    )
+                else:
+                    self.run_task(
+                        self.event_webhook_ops.create_qa_analysis_started_notification(
+                            crawl.id, crawl.oid, crawl.qa_source_crawl_id
                         )
+                    )
 
             # update lastActiveTime if crawler is running
             if crawler_running: