feat(openmetadata): backfill Airflow task detail + dataset lineage

OM 1.13 cannot deserialize Airflow 3.0.6 serialized DAGs, so pipelines had no tasks. af_tasks.py parses serialized_dag and PATCHes task lists; af_lineage.py adds generator->table and mask_to_curated/hadoop_to_trino lineage edges. Note: OM MySQL datadir lives on the 448MB /opt partition which had filled up (103MB binlog). Disabled binary logging + shrank redo log capacity in the OM docker-compose to restore headroom.
2026-06-27 13:30:08 +02:00
parent 79888dc617
commit e8ff760cd1
2 changed files with 177 additions and 0 deletions
@@ -0,0 +1,92 @@
+"""Backfill task-level detail into OpenMetadata Airflow pipelines.
+
+OM 1.13's Airflow connector cannot deserialize Airflow 3.0.6 serialized DAGs, so
+pipelines land with zero tasks. We parse the serialized_dag table ourselves and
+PATCH each pipeline with its tasks + intra-DAG dependencies (downstreamTasks).
+"""
+import json
+import os
+import sqlite3
+import urllib.error
+import urllib.request
+import zlib
+
+DB = "/" + "/".join(["tmp", "airflow.db"])
+OM = "http://openmetadata-server:8585/api"
+TOKEN = os.environ["OM_TOKEN"]
+SERVICE = "atc_airflow"
+
+HDRS = {"Authorization": "Bearer " + TOKEN}
+
+
+def req(method, path, body=None, ctype="application/json"):
+    data = json.dumps(body).encode() if body is not None else None
+    h = dict(HDRS)
+    if data is not None:
+        h["Content-Type"] = ctype
+    r = urllib.request.Request(OM + path, data=data, headers=h, method=method)
+    try:
+        with urllib.request.urlopen(r, timeout=30) as resp:
+            return resp.status, json.loads(resp.read().decode() or "{}")
+    except urllib.error.HTTPError as e:
+        return e.code, e.read().decode()
+
+
+def latest_serialized():
+    c = sqlite3.connect(DB)
+    rows = c.execute(
+        "SELECT s.dag_id, s.data, s.data_compressed FROM serialized_dag s "
+        "JOIN (SELECT dag_id, MAX(id) mid FROM serialized_dag GROUP BY dag_id) m "
+        "ON s.id = m.mid"
+    ).fetchall()
+    out = {}
+    for dag_id, data, comp in rows:
+        if data is None and comp is not None:
+            data = zlib.decompress(comp).decode()
+        d = json.loads(data) if isinstance(data, str) else data
+        out[dag_id] = d.get("dag", d)
+    return out
+
+
+def build_tasks(dag):
+    tasks = []
+    for t in dag.get("tasks", []):
+        tt = t.get("__var", t) if isinstance(t, dict) else t
+        tid = tt.get("task_id")
+        if not tid:
+            continue
+        tasks.append(
+            {
+                "name": tid,
+                "displayName": tid,
+                "taskType": tt.get("task_type") or tt.get("_task_type") or "Task",
+                "downstreamTasks": list(tt.get("downstream_task_ids") or []),
+            }
+        )
+    return tasks
+
+
+def main():
+    dags = latest_serialized()
+    for dag_id, dag in sorted(dags.items()):
+        tasks = build_tasks(dag)
+        fqn = SERVICE + "." + dag_id
+        st, pipe = req("GET", "/v1/pipelines/name/" + fqn)
+        if st != 200:
+            print("SKIP", dag_id, "get-failed", st, pipe)
+            continue
+        pid = pipe["id"]
+        op = "replace" if "tasks" in pipe else "add"
+        patch = [{"op": op, "path": "/tasks", "value": tasks}]
+        st2, res = req(
+            "PATCH", "/v1/pipelines/" + pid, patch, "application/json-patch+json"
+        )
+        edges = sum(len(t["downstreamTasks"]) for t in tasks)
+        print(
+            "%-30s tasks=%-2d edges=%-2d -> %s"
+            % (dag_id, len(tasks), edges, "OK" if st2 == 200 else ("ERR %s %s" % (st2, res)))
+        )
+
+
+if __name__ == "__main__":
+    main()