feat(manual-jobs): add job to rerun completed idempotent migrations

snuba/manual_jobs/rerun_idempotent_migration.py

@@ -0,0 +1,75 @@
+from typing import Any, Mapping, Optional
+
+from snuba.clusters.storage_sets import StorageSetKey
+from snuba.manual_jobs import Job, JobLogger, JobSpec
+from snuba.migrations.groups import get_group_loader
+from snuba.migrations.migration import ClickhouseNodeMigration
+from snuba.migrations.runner import MigrationKey, Runner
+from snuba.migrations.status import Status
+
+
+class RerunIdempotentMigration(Job):
+    """
+    Reruns the forward SQL operations of a completed idempotent migration.
+
+    Parameters:
+        - storage_set: The StorageSetKey value (e.g. "events_analytics_platform")
+        - migration_id: The migration ID (e.g. "0001_events_initial")
+    """
+
+    def __init__(self, job_spec: JobSpec) -> None:
+        self.__validate_job_params(job_spec.params)
+        super().__init__(job_spec)
+
+    def __validate_job_params(self, params: Optional[Mapping[Any, Any]]) -> None:
+        assert params is not None, "storage_set and migration_id parameters required"
+        assert params.get("storage_set"), "storage_set required"
+        assert params.get("migration_id"), "migration_id required"
+        self._storage_set = params["storage_set"]
+        self._migration_id = params["migration_id"]
+        assert isinstance(self._storage_set, str)
+        assert isinstance(self._migration_id, str)
+
+    def execute(self, logger: JobLogger) -> None:
+        storage_set_key = StorageSetKey(self._storage_set)
+
+        # Find the migration group for this storage set
+        from snuba.migrations.groups import _STORAGE_SET_TO_MIGRATION_GROUP_MAPPING
+
+        group = _STORAGE_SET_TO_MIGRATION_GROUP_MAPPING.get(storage_set_key)
+        assert group is not None, f"No migration group found for storage set {self._storage_set}"
+
+        # Check that the migration is in COMPLETED state
+        migration_key = MigrationKey(group=group, migration_id=self._migration_id)
+        runner = Runner()
+        status, _ = runner.get_status(migration_key)
+        assert status == Status.COMPLETED, (
+            f"Migration {self._migration_id} in group {group.value} has status "
+            f"{status.value}, expected completed"
+        )
+
+        # Load the migration
+        group_loader = get_group_loader(group)
+        migration = group_loader.load_migration(self._migration_id)
+        assert isinstance(migration, ClickhouseNodeMigration), (
+            f"Migration {self._migration_id} is not a ClickhouseNodeMigration, "
+            f"only SQL migrations are supported"
+        )
+
+        # Get and execute forward operations
+        ops = migration.forwards_ops()
+        logger.info(
+            f"{self.job_spec.job_id}: rerunning {len(ops)} operations from "
+            f"migration {self._migration_id} (group={group.value})"
+        )
+
+        for i, op in enumerate(ops):
+            sql = op.format_sql()
+            logger.info(f"{self.job_spec.job_id}: executing op {i + 1}/{len(ops)}: {sql}")
+            op.execute()
+            logger.info(f"{self.job_spec.job_id}: op {i + 1}/{len(ops)} completed")
+
+        logger.info(
+            f"{self.job_spec.job_id}: successfully reran all operations for "
+            f"migration {self._migration_id}"
+        )