Create skip_nodes_if_on_run_start_fails behavior change flag

.changes/unreleased/Features-20240918-162959.yaml

@@ -0,0 +1,6 @@
+kind: Features
+body: Create 'skip_nodes_if_on_run_start_fails' behavior change flag
+time: 2024-09-18T16:29:59.268422+01:00
+custom:
+    Author: aranke
+    Issue: "7387"

core/dbt/contracts/graph/nodes.py

@@ -1654,6 +1654,7 @@ class ParsedMacroPatch(ParsedPatch):
 ResultNode = Union[
     ManifestNode,
     SourceDefinition,
+    HookNode,
 ]
 
 # All nodes that can be in the DAG

core/dbt/contracts/project.py

@@ -341,13 +341,15 @@ class ProjectFlags(ExtensibleDbtClassMixin):
     require_explicit_package_overrides_for_builtin_materializations: bool = True
     require_resource_names_without_spaces: bool = False
     source_freshness_run_project_hooks: bool = False
+    skip_nodes_if_on_run_start_fails: bool = False
 
     @property
     def project_only_flags(self) -> Dict[str, Any]:
         return {
             "require_explicit_package_overrides_for_builtin_materializations": self.require_explicit_package_overrides_for_builtin_materializations,
             "require_resource_names_without_spaces": self.require_resource_names_without_spaces,
             "source_freshness_run_project_hooks": self.source_freshness_run_project_hooks,
+            "skip_nodes_if_on_run_start_fails": self.skip_nodes_if_on_run_start_fails,
         }
 
 

core/dbt/events/types.py

@@ -1513,10 +1513,20 @@ def code(self) -> str:
         return "Q033"
 
     def message(self) -> str:
-        msg = f"OK hook: {self.statement}"
+        if self.status == "success":
+            info = "OK"
+            status = green(info)
+        elif self.status == "skipped":
+            info = "SKIP"
+            status = yellow(info)
+        else:
+            info = "ERROR"
+            status = red(info)
+        msg = f"{info} hook: {self.statement}"
+
         return format_fancy_output_line(
             msg=msg,
-            status=green(self.status),
+            status=status,
             index=self.index,
             total=self.total,
             execution_time=self.execution_time,

core/dbt/task/clone.py

@@ -1,7 +1,7 @@
 import threading
 from typing import AbstractSet, Any, Collection, Iterable, List, Optional, Set, Type
 
-from dbt.adapters.base import BaseRelation
+from dbt.adapters.base import BaseAdapter, BaseRelation
 from dbt.artifacts.resources.types import NodeType
 from dbt.artifacts.schemas.run import RunResult, RunStatus
 from dbt.clients.jinja import MacroGenerator
@@ -124,14 +124,15 @@ def get_model_schemas(self, adapter, selected_uids: Iterable[str]) -> Set[BaseRe
 
         return result
 
-    def before_run(self, adapter, selected_uids: AbstractSet[str]):
+    def before_run(self, adapter: BaseAdapter, selected_uids: AbstractSet[str]) -> RunStatus:
         with adapter.connection_named("master"):
             self.defer_to_manifest()
             # only create target schemas, but also cache defer_relation schemas
             schemas_to_create = super().get_model_schemas(adapter, selected_uids)
             self.create_schemas(adapter, schemas_to_create)
             schemas_to_cache = self.get_model_schemas(adapter, selected_uids)
             self.populate_adapter_cache(adapter, schemas_to_cache)
+            return RunStatus.Success
 
     @property
     def resource_types(self) -> List[NodeType]:

core/dbt/task/freshness.py

@@ -4,6 +4,7 @@
 from typing import AbstractSet, Dict, List, Optional, Type
 
 from dbt import deprecations
+from dbt.adapters.base import BaseAdapter
 from dbt.adapters.base.impl import FreshnessResponse
 from dbt.adapters.base.relation import BaseRelation
 from dbt.adapters.capability import Capability
@@ -204,10 +205,25 @@
             resource_types=[NodeType.Source],
         )
 
-    def before_run(self, adapter, selected_uids: AbstractSet[str]) -> None:
-        super().before_run(adapter, selected_uids)
-        if adapter.supports(Capability.TableLastModifiedMetadataBatch):
-            self.populate_metadata_freshness_cache(adapter, selected_uids)
+    def before_run(self, adapter: BaseAdapter, selected_uids: AbstractSet[str]) -> RunStatus:
+        populate_metadata_freshness_cache_status = RunStatus.Success
+
+        before_run_status = super().before_run(adapter, selected_uids)
+
+        if before_run_status == RunStatus.Success and adapter.supports(
+            Capability.TableLastModifiedMetadataBatch
+        ):
+            populate_metadata_freshness_cache_status = self.populate_metadata_freshness_cache(
+                adapter, selected_uids
+            )
+
+        if (
+            before_run_status == RunStatus.Success
+            and populate_metadata_freshness_cache_status == RunStatus.Success
+        ):
+            return RunStatus.Success
+        else:
+            return RunStatus.Error
 
     def get_runner(self, node) -> BaseRunner:
         freshness_runner = super().get_runner(node)
@@ -243,7 +259,9 @@
                 deprecations.warn("source-freshness-project-hooks")
             return []
 
-    def populate_metadata_freshness_cache(self, adapter, selected_uids: AbstractSet[str]) -> None:
+    def populate_metadata_freshness_cache(
+        self, adapter, selected_uids: AbstractSet[str]
+    ) -> RunStatus:
         if self.manifest is None:
             raise DbtInternalError("Manifest must be set to populate metadata freshness cache")
 
@@ -266,6 +284,7 @@
                 batch_metadata_sources
             )
             self._metadata_freshness_cache.update(metadata_freshness_results)
+            return RunStatus.Success
         except Exception as e:
             # This error handling is intentionally very coarse.
             # If anything goes wrong during batch metadata calculation, we can safely
@@ -276,6 +295,7 @@
                 Note(msg=f"Metadata freshness could not be computed in batch: {e}"),
                 EventLevel.WARN,
             )
+            return RunStatus.Error
 
     def get_freshness_metadata_cache(self) -> Dict[BaseRelation, FreshnessResult]:
         return self._metadata_freshness_cache

core/dbt/task/printer.py

@@ -33,7 +33,8 @@ def get_counts(flat_nodes) -> str:
 
         counts[t] = counts.get(t, 0) + 1
 
-    stat_line = ", ".join([pluralize(v, k).replace("_", " ") for k, v in counts.items()])
+    sorted_items = sorted(counts.items(), key=lambda x: x[0])
+    stat_line = ", ".join([pluralize(v, k).replace("_", " ") for k, v in sorted_items])
 
     return stat_line