[Monitoring] Collecting a metric for the age of untriaged testcases (#…

…4381)

### Motivation


Once a testcase is generated (or manually uploaded), followup tasks
(analyze/progression) are started. This happens by publishing to a
pubsub queue, both for the manually uploaded case, and for the fuzzer
generated case.

If for any reason the messages are not processed, the testcase gets
stuck. To get better visibility into these stuck testcases, the
UNTRIAGED_TESTCASE_AGE metric is introduced, to pinpoint how old these
testcases that have not yet been triaged are(more precisely, gone
through analyze/regression/impact/progression tasks).


### Attention points

Testcase.timestamp mutates in analyze task:


https://github.com/google/clusterfuzz/blob/6ed80851ad0f6f624c5b232b0460c405f0a018b5/src/clusterfuzz/_internal/bot/tasks/utasks/analyze_task.py#L589

This makes it unreliable as a source of truth for testcase creation
time. To circumvent that, a new ```created``` field is added to the
Testcase entity, from which we can derive the correct creation time.

Since this new field will only apply for testcases created after this PR
is merged, Testcase.timestamp will be used instead to calculate the
testcase age when the new field is missing.

### Testing strategy

Ran the triage cron locally, and verified the codepath for the metric is
hit and produces sane output (reference testcase: 4505741036158976).

![image](https://github.com/user-attachments/assets/6281b44f-768a-417e-8ec1-763f132c8181)


Part of #4271

src/clusterfuzz/_internal/cron/triage.py

@@ -298,6 +298,22 @@ def _file_issue(testcase, issue_tracker, throttler):
   return filed
 
 
+def _emit_untriaged_testcase_age_metric(critical_tasks_completed: bool,
+                                        testcase: data_types.Testcase):
+  """Emmits a metric to track age of untriaged testcases."""
+  if critical_tasks_completed:
+    return
+  if not testcase.timestamp:
+    return
+
+  monitoring_metrics.UNTRIAGED_TESTCASE_AGE.add(
+      testcase.get_age_in_seconds(),
+      labels={
+          'job': testcase.job_type,
+          'platform': testcase.platform,
+      })
+
+
 def main():
   """Files bugs."""
   try:
@@ -328,6 +344,8 @@ def main():
       # Already deleted.
       continue
 
+    critical_tasks_completed = data_handler.critical_tasks_completed(testcase)
+
     # Skip if testcase's job is removed.
     if testcase.job_type not in all_jobs:
       continue
@@ -336,6 +354,9 @@ def main():
     if testcase.job_type in excluded_jobs:
       continue
 
+    # Emmit the metric for testcases that should be triaged.
+    _emit_untriaged_testcase_age_metric(critical_tasks_completed, testcase)
+
     # Skip if we are running progression task at this time.
     if testcase.get_metadata('progression_pending'):
       continue
@@ -351,7 +372,7 @@ def main():
 
     # Require that all tasks like minimizaton, regression testing, etc have
     # finished.
-    if not data_handler.critical_tasks_completed(testcase):
+    if not critical_tasks_completed:
       continue
 
     # For testcases that are not part of a group, wait an additional time to

src/clusterfuzz/_internal/datastore/data_handler.py

@@ -788,6 +788,7 @@ def store_testcase(crash, fuzzed_keys, minimized_keys, regression, fixed,
   testcase.archive_filename = archive_filename
   testcase.http_flag = http_flag
   testcase.timestamp = datetime.datetime.utcnow()
+  testcase.created = testcase.timestamp
   testcase.gestures = gestures
   testcase.redzone = redzone
   testcase.disable_ubsan = disable_ubsan
@@ -1377,6 +1378,7 @@ def create_user_uploaded_testcase(key,
       testcase.set_metadata(metadata_key, metadata_value, update_testcase=False)
 
   testcase.timestamp = utils.utcnow()
+  testcase.created = testcase.timestamp
   testcase.uploader_email = uploader_email
   testcase.put()
 

src/clusterfuzz/_internal/datastore/data_types.py

@@ -421,9 +421,16 @@ class Testcase(Model):
   # File name of the original uploaded archive.
   archive_filename = ndb.TextProperty()
 
-  # Timestamp.
+  # The time when a testcase is considered valid. This is the same as the
+  # creation time, except for analyze task, in which this field is a
+  # placeholder and will be refreshed.
   timestamp = ndb.DateTimeProperty()
 
+  # Source of truth for creation time. This is missing for testcases
+  # created before it was introduced, in which case the timestamp
+  # field will be a proxy for creation time.
+  created = ndb.DateTimeProperty(indexed=False)
+
   # Does the testcase crash stack vary b/w crashes ?
   flaky_stack = ndb.BooleanProperty(default=False, indexed=False)
 
@@ -671,6 +678,17 @@ def _ensure_metadata_is_cached(self):
 
     setattr(self, 'metadata_cache', cache)
 
+  # Returns testcase.created in case it is present, as it is
+  # the source of truth for creation time. If missing, returns
+  # testcase.timestamp as a proxy for creation time.
+  def get_created_time(self) -> ndb.DateTimeProperty:
+    return self.created if self.created else self.timestamp
+
+  def get_age_in_seconds(self):
+    current_time = datetime.datetime.utcnow()
+    testcase_age = current_time - self.get_created_time()
+    return testcase_age.total_seconds()
+
   def get_metadata(self, key=None, default=None):
     """Get metadata for a test case. Slow on first access."""
     self._ensure_metadata_is_cached()

src/clusterfuzz/_internal/metrics/monitoring_metrics.py

@@ -273,6 +273,17 @@
         monitor.StringField('status'),
     ])
 
+UNTRIAGED_TESTCASE_AGE = monitor.CumulativeDistributionMetric(
+    'issues/untriaged_testcase_age',
+    description='Age of testcases that were not yet triaged '
+    '(have not yet completed analyze, regression,'
+    ' minimization, impact task), in seconds.',
+    bucketer=monitor.GeometricBucketer(),
+    field_spec=[
+        monitor.StringField('job'),
+        monitor.StringField('platform'),
+    ])
+
 ANALYZE_TASK_REPRODUCIBILITY = monitor.CounterMetric(
     'task/analyze/reproducibility',
     description='Outcome count for analyze task.',