Activity log for bug #1451520

Date Who What changed Old value New value Message
2015-05-04 18:00:18 Mike Barton bug added bug
2015-05-31 23:20:40 Mike Barton description There's a race condition that isn't handled well in diskfile.py - when a file gets removed between the object hash listdir and reading the metadata, _failsafe_read_metadata will quarantine the whole dir even though nothing particularly wrong happened. Such as: DELETE request does a listdir, sees a .data file. PUT operation completes, removing the old .data and adding a new .data. DELETE request attempts to read metadata from old .data, gets an exception, and quarantines the shiny new .data file. There's a race condition that isn't handled well in diskfile.py - in some cases when a file gets removed between the object hash listdir and reading the metadata, _failsafe_read_metadata will quarantine the whole dir even though nothing particularly wrong happened. Such as: DELETE request does a listdir, sees an existing .ts file. A separate PUT operation completes, removing the .ts and adding a new .data. DELETE request attempts to read metadata from old .ts, gets an exception, and quarantines the shiny new .data file.
2015-06-03 05:58:35 OpenStack Infra swift: status New Fix Committed
2015-07-15 21:17:49 OpenStack Infra tags in-feature-hummingbird
2015-07-15 21:17:50 OpenStack Infra bug watch added http://bugs.python.org/issue16037
2015-07-24 23:50:04 OpenStack Infra tags in-feature-hummingbird in-feature-crypto in-feature-hummingbird
2015-09-01 12:26:48 Thierry Carrez swift: status Fix Committed Fix Released
2015-09-01 12:26:48 Thierry Carrez swift: milestone 2.4.0