openstack
/
neutron
Code Issues Proposed changes

Stop spilling tracebacks during normal operation 

Yes, they are innocent; yes, we log them with DEBUG level. But they
scare people, they make people judge the software, they are bad PR, and
they are probably redundant, string representation of the exception
should be enough. If there is an issue that will result in retry
attempts used without remedy, only then we should spill tracebacks and
scare users.

Guidelines: https://specs.openstack.org/openstack/openstack-specs/specs/log-guidelines.html#stacktraces-in-logs

Partial-Bug: #1673249
Change-Id: I791a52161635da64b1840c01ee6cf56a51bc7400
This commit is contained in:
Ihar Hrachyshka 2017-03-08 04:05:18 +00:00
parent a457949bf7
commit 4f7fc8435f
1 changed files with 1 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
neutron/db/api.py
View File

@ -32,7 +32,6 @@ import sqlalchemy
from sqlalchemy import event # noqa
from sqlalchemy import exc as sql_exc
from sqlalchemy.orm import exc
import traceback
from neutron._i18n import _LE
from neutron.objects import exceptions as obj_exc
@ -125,8 +124,7 @@ def retry_db_errors(f):
except Exception as e:
with excutils.save_and_reraise_exception():
if is_retriable(e):
LOG.debug("Retry wrapper got retriable exception: %s",
traceback.format_exc())
LOG.debug("Retry wrapper got retriable exception: %s", e)
return wrapped