Reviewed: https://review.openstack.org/261227 Committed: https://git.openstack.org/cgit/openstack/neutron/commit/?id=18e2117e3df3c119f843aed669d0c6bae9243053 Submitter: Jenkins Branch: master
commit 18e2117e3df3c119f843aed669d0c6bae9243053 Author: Kevin Benton <[email protected]> Date: Thu Dec 24 00:54:06 2015 -0800 Call _allocate_vr_id outside of transaction _allocate_vr_id is called from _set_vr_id, which was starting a transaction before calling it. This caused an error when the retry logic was triggered inside of _allocate_vr_id since it would encounter a DB exception and put the transaction into a bad state that couldn't be used on the retry. This patch just stops _set_vr_id from starting a transaction because it didn't serve a purpose. It also stops _allocate_vr_id from allowing subtransactions when it starts a transaction, since it's retry logic isn't compatible with them. Co-Authored-By: Ann Kamyshnikova <[email protected]> Closes-Bug: #1528201 Change-Id: If7bcae39098f40f5ee9db78d3190bf9fdaf6438b ** Changed in: neutron Status: In Progress => Fix Released -- You received this bug notification because you are a member of Yahoo! Engineering Team, which is subscribed to neutron. https://bugs.launchpad.net/bugs/1528201 Title: [L3 HA] Stacktrace during HA router creation when running rally test Status in Mirantis OpenStack: Confirmed Status in neutron: Fix Released Bug description: Rally test create_and_list_routers fails in some cases. Failure is caused by L3_HA_NAT_db_mixin._set_vr_id/_allocate_vr_id trying to reuse db session after it has been rolled back by DBDuplicateError exception. To manage notifications about this bug go to: https://bugs.launchpad.net/mos/+bug/1528201/+subscriptions -- Mailing list: https://launchpad.net/~yahoo-eng-team Post to : [email protected] Unsubscribe : https://launchpad.net/~yahoo-eng-team More help : https://help.launchpad.net/ListHelp
