openstack
/
neutron
Code Issues Proposed changes

Wait longer before deleting DPDK vhu trunk bridges 

In [1] we added a delay before deleting DPDK vhu trunk bridges to
mitigate a race condition when instances are rebooted. As explained in
[1], with DPDK rhu, a reboot is esentially a super fast bridge
delete-then-create that is prone to race conditions. We have recently
encountered in customer deployments that the wait added in [1] is not
long enough. As a consequence, this change increases the wait.

[1] https://review.opendev.org/c/openstack/neutron/+/717394

Change-Id: I5c1474b405d436d3b1e5db745d77999f1723b660
Partial-Bug: #1869244
(cherry picked from commit 140bb63665)
This commit is contained in:
Miguel Lavalle 2022-02-02 19:25:40 -06:00 committed by Miguel Lavalle
parent 02af009025
commit 7338441c41
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
neutron/services/trunk/drivers/openvswitch/agent/ovsdb_handler.py
View File

@ -44,7 +44,7 @@ from neutron.services.trunk.rpc import agent
LOG = logging.getLogger(__name__)
DEFAULT_WAIT_FOR_PORT_TIMEOUT = 60
WAIT_BEFORE_TRUNK_DELETE = 3
WAIT_BEFORE_TRUNK_DELETE = 6
def lock_on_bridge_name(required_parameter):