STASH add platform to metadata and metrics

Author: lobsterkatie

src/sentry/grouping/ingest/grouphash_metadata.py

@@ -119,6 +119,7 @@ def create_or_update_grouphash_metadata_if_needed(
             latest_grouping_config=grouping_config,
             hash_basis=hash_basis,
             hashing_metadata=hashing_metadata,
+            platform=event.platform,
         )
     elif grouphash.metadata and grouphash.metadata.latest_grouping_config != grouping_config:
         # Keep track of the most recent config which computed this hash, so that once a
@@ -203,7 +204,9 @@ def get_hash_basis_and_metadata(
     return hash_basis, hashing_metadata
 
 
-def record_grouphash_metadata_metrics(grouphash_metadata: GroupHashMetadata) -> None:
+def record_grouphash_metadata_metrics(
+    grouphash_metadata: GroupHashMetadata, platform: str | None
+) -> None:
     # TODO: Once https://peps.python.org/pep-0728 is a thing (still in draft but theoretically on
     # track for 3.14), we can mark the various hashing metadata types as closed and that should
     # narrow the types for the tag values such that we can stop stringifying everything
@@ -222,7 +225,7 @@ def is_stacktrace_hashing(
     hashing_metadata = grouphash_metadata.hashing_metadata
 
     if hash_basis:
-        hash_basis_tags: dict[str, str] = {"hash_basis": hash_basis}
+        hash_basis_tags: dict[str, str] = {"hash_basis": hash_basis, "platform": platform}
         if hashing_metadata:
             hash_basis_tags["is_hybrid_fingerprint"] = str(
                 hashing_metadata.get("is_hybrid_fingerprint", False)
@@ -244,7 +247,11 @@ def is_stacktrace_hashing(
                 metrics.incr(
                     f"grouping.grouphashmetadata.event_hashing_metadata.{hash_basis}",
                     sample_rate=1.0,
-                    tags=hashing_metadata_tags,
+                    tags={
+                        **hashing_metadata_tags,
+                        # Add this in at the end so it's not the reason we log the metric
+                        "platform": platform,
+                    },
                 )
 
 

src/sentry/grouping/ingest/hashing.py

@@ -241,7 +241,7 @@ def get_or_create_grouphashes(
                 sentry_sdk.capture_exception(exc)
 
         if grouphash.metadata:
-            record_grouphash_metadata_metrics(grouphash.metadata)
+            record_grouphash_metadata_metrics(grouphash.metadata, event.platform)
         else:
             # Collect a temporary metric to get a sense of how often we would be adding metadata to an
             # existing hash. (Yes, this is an overestimate, because this will fire every time we see a given

src/sentry/migrations/0827_add_platform_to_grouphash_metadata.py

@@ -0,0 +1,33 @@
+# Generated by Django 5.1.5 on 2025-02-07 05:09
+
+from django.db import migrations, models
+
+from sentry.new_migrations.migrations import CheckedMigration
+
+
+class Migration(CheckedMigration):
+    # This flag is used to mark that a migration shouldn't be automatically run in production.
+    # This should only be used for operations where it's safe to run the migration after your
+    # code has deployed. So this should not be used for most operations that alter the schema
+    # of a table.
+    # Here are some things that make sense to mark as post deployment:
+    # - Large data migrations. Typically we want these to be run manually so that they can be
+    #   monitored and not block the deploy for a long period of time while they run.
+    # - Adding indexes to large tables. Since this can take a long time, we'd generally prefer to
+    #   run this outside deployments so that we don't block them. Note that while adding an index
+    #   is a schema change, it's completely safe to run the operation after the code has deployed.
+    # Once deployed, run these manually via: https://develop.sentry.dev/database-migrations/#migration-deployment
+
+    is_post_deployment = False
+
+    dependencies = [
+        ("sentry", "0826_make_sentryapp_uuid_unique"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="grouphashmetadata",
+            name="platform",
+            field=models.CharField(null=True),
+        ),
+    ]

src/sentry/models/grouphashmetadata.py

@@ -51,6 +51,10 @@ class GroupHashMetadata(Model):
         "sentry.GroupHash", related_name="_metadata", on_delete=models.CASCADE
     )
     date_added = models.DateTimeField(default=timezone.now)
+    # The platform of the event when generated the metadata. Likely different than the project
+    # platform, as event platforms are normalized to a handful of known values, whereas project
+    # platforms are all over the place.
+    platform = models.CharField(null=True)
 
     # HASHING
 

tests/sentry/grouping/test_grouphash_metadata.py

@@ -143,11 +143,15 @@ def _assert_and_snapshot_results(
 
     with patch("sentry.grouping.ingest.grouphash_metadata.metrics.incr") as mock_metrics_incr:
         record_grouphash_metadata_metrics(
-            GroupHashMetadata(hash_basis=hash_basis, hashing_metadata=hashing_metadata)
+            GroupHashMetadata(hash_basis=hash_basis, hashing_metadata=hashing_metadata),
+            event.platform,
         )
 
         metric_names = [call.args[0] for call in mock_metrics_incr.mock_calls]
         tags = [call.kwargs["tags"] for call in mock_metrics_incr.mock_calls]
+        # Filter out all the `platform` tags, because many of the inputs don't have a `platform`
+        # value, so we're going to get a lot of Nones
+        [t.pop("platform", None) for t in tags]
         metrics_data = dict(zip(metric_names, tags))
 
         expected_metric_names = ["grouping.grouphashmetadata.event_hash_basis"]