From a4fff886a6c77cf414f244f1133ba8f5d1b7aba2 Mon Sep 17 00:00:00 2001
From: Joseph Schorr <joseph.schorr@coreos.com>
Date: Fri, 7 Sep 2018 13:34:22 -0400
Subject: [PATCH] Handle the case where the log metadata JSON cannot be parsed

This can happen if the JSON overflowed the text field in the table, for example

Fixes https://sentry.io/coreos/backend-production/issues/670957592/
---
 workers/logrotateworker.py | 25 ++++++++++++++++++-------
 1 file changed, 18 insertions(+), 7 deletions(-)

diff --git a/workers/logrotateworker.py b/workers/logrotateworker.py
index ddf4f9ad0..e8486403c 100644
--- a/workers/logrotateworker.py
+++ b/workers/logrotateworker.py
@@ -106,13 +106,24 @@ class LogRotateWorker(Worker):
 
 def log_dict(log):
   """ Pretty prints a LogEntry in JSON. """
-  return {'kind_id': log.kind_id,
-          'account_id': log.account_id,
-          'performer_id': log.performer_id,
-          'repository_id': log.repository_id,
-          'datetime': str(log.datetime),
-          'ip': str(log.ip),
-          'metadata_json': json.loads(str(log.metadata_json))}
+  try:
+    metadata_json = json.loads(str(log.metadata_json))
+  except ValueError:
+    logger.exception('Could not parse metadata JSON for log entry %s', log.id)
+    metadata_json = {'__raw': log.metadata_json}
+  except TypeError:
+    logger.exception('Could not parse metadata JSON for log entry %s', log.id)
+    metadata_json = {'__raw': log.metadata_json}
+
+  return {
+    'kind_id': log.kind_id,
+    'account_id': log.account_id,
+    'performer_id': log.performer_id,
+    'repository_id': log.repository_id,
+    'datetime': str(log.datetime),
+    'ip': str(log.ip),
+    'metadata_json': metadata_json,
+  }
 
 
 def main():