yahoo-eng-team team mailing list archive

Thread
Date
[Bug 2093433] [NEW] OVS Manager creation is failing at ovs-agent startup causing that values like inactivity_timeout are not set

To: yahoo-eng-team@xxxxxxxxxxxxxxxxxxx
From: Lajos Katona <2093433@xxxxxxxxxxxxxxxxxx>
Date: Fri, 10 Jan 2025 15:10:45 -0000
Reply-to: Bug 2093433 <2093433@xxxxxxxxxxxxxxxxxx>
Sender: noreply@xxxxxxxxxxxxx
Public bug reported:

In some cases neutron-ovs-agent is not able to create the OVS manager due to row collision, example from upstream CI run:
Jan 08 14:52:18.437125 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo_concurrency.processutils [-] Running cmd (subprocess): ovs-vsctl --timeout=10 --id=@manager -- create Manager target="ptcp:6640:127.0.0.1" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000 {{(pid=65448) execute /opt/stack/data/venv/lib/python3.12/site-packages/oslo_concurrency/processutils.py:383}}
Jan 08 14:52:18.449092 np0039512747 ovs-vsctl[65452]: ovs|00001|vsctl|INFO|Called as ovs-vsctl --timeout=10 --id=@manager -- create Manager "target=\"ptcp:6640:127.0.0.1\"" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000
Jan 08 14:52:18.453577 np0039512747 ovs-vsctl[65452]: ovs|00002|db_ctl_base|ERR|multiple rows in Manager match "ptcp:6640:127.0.0.1"
Jan 08 14:52:18.455956 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo_concurrency.processutils [-] CMD "ovs-vsctl --timeout=10 --id=@manager -- create Manager target="ptcp:6640:127.0.0.1" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000" returned: 1 in 0.018s {{(pid=65448) execute /opt/stack/data/venv/lib/python3.12/site-packages/oslo_concurrency/processutils.py:421}}
Jan 08 14:52:18.455956 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo.privsep.daemon [-] privsep: reply[cdb9b623-b580-4bd9-9349-ed6a566f7b16]: (4, ('', 'ovs-vsctl: multiple rows in Manager match "ptcp:6640:127.0.0.1"\n')) {{(pid=65448) _call_back /opt/stack/data/venv/lib/python3.12/site-packages/oslo_privsep/daemon.py:498}}

(https://zuul.opendev.org/t/openstack/build/fc2035fd67f24c87a4ad070fefdcc703/log/controller/logs/screen-
q-agt.txt#561-565 )

I see the same in my local devstack, the issue is that in this stage
values in the tManager table like inactivity_probe are set in normal
case.

** Affects: neutron
     Importance: Undecided
     Assignee: Lajos Katona (lajos-katona)
         Status: New


** Tags: ovs

** Changed in: neutron
     Assignee: (unassigned) => Lajos Katona (lajos-katona)

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to neutron.
https://bugs.launchpad.net/bugs/2093433

Title:
  OVS Manager creation is failing at ovs-agent startup causing that
  values like inactivity_timeout are not set

Status in neutron:
  New

Bug description:
  In some cases neutron-ovs-agent is not able to create the OVS manager due to row collision, example from upstream CI run:
  Jan 08 14:52:18.437125 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo_concurrency.processutils [-] Running cmd (subprocess): ovs-vsctl --timeout=10 --id=@manager -- create Manager target="ptcp:6640:127.0.0.1" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000 {{(pid=65448) execute /opt/stack/data/venv/lib/python3.12/site-packages/oslo_concurrency/processutils.py:383}}
  Jan 08 14:52:18.449092 np0039512747 ovs-vsctl[65452]: ovs|00001|vsctl|INFO|Called as ovs-vsctl --timeout=10 --id=@manager -- create Manager "target=\"ptcp:6640:127.0.0.1\"" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000
  Jan 08 14:52:18.453577 np0039512747 ovs-vsctl[65452]: ovs|00002|db_ctl_base|ERR|multiple rows in Manager match "ptcp:6640:127.0.0.1"
  Jan 08 14:52:18.455956 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo_concurrency.processutils [-] CMD "ovs-vsctl --timeout=10 --id=@manager -- create Manager target="ptcp:6640:127.0.0.1" -- add Open_vSwitch . manager_options @manager -- set Manager ptcp:6640:127.0.0.1 inactivity_probe=10000" returned: 1 in 0.018s {{(pid=65448) execute /opt/stack/data/venv/lib/python3.12/site-packages/oslo_concurrency/processutils.py:421}}
  Jan 08 14:52:18.455956 np0039512747 neutron-openvswitch-agent[63181]: DEBUG oslo.privsep.daemon [-] privsep: reply[cdb9b623-b580-4bd9-9349-ed6a566f7b16]: (4, ('', 'ovs-vsctl: multiple rows in Manager match "ptcp:6640:127.0.0.1"\n')) {{(pid=65448) _call_back /opt/stack/data/venv/lib/python3.12/site-packages/oslo_privsep/daemon.py:498}}

  (https://zuul.opendev.org/t/openstack/build/fc2035fd67f24c87a4ad070fefdcc703/log/controller/logs/screen-
  q-agt.txt#561-565 )

  I see the same in my local devstack, the issue is that in this stage
  values in the tManager table like inactivity_probe are set in normal
  case.

To manage notifications about this bug go to:
https://bugs.launchpad.net/neutron/+bug/2093433/+subscriptions