Network deletion fails in certain cases

[Sumit Naiksatam]

If a network is created and deleted without starting a VM, deletion fails since killing the dnsmasq fails (no dnsmasq process was started). Trace below is explanatory:

stack@dd05:~/devstack$ nova-manage --flagfile=/opt/stack/nova/bin/nova.conf network delete --uuid=b581f449-01d0-4411-95f3-1015dd12b842
2012-02-10 19:04:04,698 DEBUG nova.network.quantum.quantum_connection [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Quantum Client Request:
GET /v1.1/tenants/default/networks/b581f449-01d0-4411-95f3-1015dd12b842/ports.json
 from (pid=30231) do_request /opt/stack/nova/nova/network/quantum/client.py:180
2012-02-10 19:04:04,707 DEBUG nova.network.quantum.quantum_connection [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Quantum Client Reply (code = 200) :
 {"ports": []} from (pid=30231) do_request /opt/stack/nova/nova/network/quantum/client.py:191
2012-02-10 19:04:04,707 DEBUG nova.network.quantum.manager [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Ports currently on network: [] from (pid=30231) delete_network /opt/stack/nova/nova/network/quantum/manager.py:247
2012-02-10 19:04:04,708 DEBUG nova.network.quantum.quantum_connection [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Quantum Client Request:
DELETE /v1.1/tenants/default/networks/b581f449-01d0-4411-95f3-1015dd12b842.json
 from (pid=30231) do_request /opt/stack/nova/nova/network/quantum/client.py:180
2012-02-10 19:04:04,722 DEBUG nova.network.quantum.quantum_connection [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Quantum Client Reply (code = 204) :
  from (pid=30231) do_request /opt/stack/nova/nova/network/quantum/client.py:191
2012-02-10 19:04:04,722 DEBUG nova.network.quantum.manager [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Deleting network for tenant: default from (pid=30231) delete_network /opt/stack/nova/nova/network/quantum/manager.py:255
2012-02-10 19:04:04,955 DEBUG nova.utils [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] backend <module 'nova.db.sqlalchemy.api' from '/opt/stack/nova/nova/db/sqlalchemy/api.pyc'> from (pid=30231) __get_backend /opt/stack/nova/nova/utils.py:588
2012-02-10 19:04:05,119 DEBUG nova.utils [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Running cmd (subprocess): sudo kill -9 None from (pid=30231) execute /opt/stack/nova/nova/utils.py:208
2012-02-10 19:04:05,133 DEBUG nova.utils [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Result was 1 from (pid=30231) execute /opt/stack/nova/nova/utils.py:224
Command failed, please check log for more info
2012-02-10 19:04:05,135 CRITICAL nova [req-1ce280e7-68f2-440a-a628-6d40d353ba66 None None] Unexpected error while running command.
Command: sudo kill -9 None
Exit code: 1
Stdout: ''
Stderr: 'ERROR: garbage process ID "None".\nUsage:\n kill pid ... Send SIGTERM to every process listed.\n kill signal pid ... Send a signal to every process listed.\n kill -s signal pid ... Send a signal to every process listed.\n kill -l List all signal names.\n kill -L List all signal names in a nice table.\n kill -l signal Convert between signal numbers and names.\n'
(nova): TRACE: Traceback (most recent call last):
(nova): TRACE: File "/usr/local/bin/nova-manage", line 7, in <module>
(nova): TRACE: execfile(__file__)
(nova): TRACE: File "/opt/stack/nova/bin/nova-manage", line 2382, in <module>
(nova): TRACE: main()
(nova): TRACE: File "/opt/stack/nova/bin/nova-manage", line 2369, in main
(nova): TRACE: fn(*fn_args, **fn_kwargs)
(nova): TRACE: File "/opt/stack/nova/bin/nova-manage", line 909, in delete
(nova): TRACE: fixed_range, uuid)
(nova): TRACE: File "/opt/stack/nova/nova/network/quantum/manager.py", line 261, in delete_network
(nova): TRACE: self.driver.kill_dhcp(dev)
(nova): TRACE: File "/opt/stack/nova/nova/network/linux_net.py", line 680, in kill_dhcp
(nova): TRACE: _execute('kill', '-9', pid, run_as_root=True)
(nova): TRACE: File "/opt/stack/nova/nova/network/linux_net.py", line 837, in _execute
(nova): TRACE: return utils.execute(*cmd, **kwargs)
(nova): TRACE: File "/opt/stack/nova/nova/utils.py", line 231, in execute
(nova): TRACE: cmd=' '.join(cmd))
(nova): TRACE: ProcessExecutionError: Unexpected error while running command.
(nova): TRACE: Command: sudo kill -9 None
(nova): TRACE: Exit code: 1
(nova): TRACE: Stdout: ''
(nova): TRACE: Stderr: 'ERROR: garbage process ID "None".\nUsage:\n kill pid ... Send SIGTERM to every process listed.\n kill signal pid ... Send a signal to every process listed.\n kill -s signal pid ... Send a signal to every process listed.\n kill -l List all signal names.\n kill -L List all signal names in a nice table.\n kill -l signal Convert between signal numbers and names.\n'
(nova): TRACE:

[dan wendlandt (danwent)]

I think this issues may be fixed by the following patch proposed to nova: https://review.openstack.org/#change,4819

Sumit, can you confirm and close this as a duplicate if it is?

[Sumit Naiksatam (snaiksat)]

Dan, the patch you mention does seem to fix this. I see that this bug is already marked as duplicate.