using neutron CLI to set a bad gateway result in dnsmasq continuously restart
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
neutron |
Invalid
|
Medium
|
ugvddm |
Bug Description
I'm using devstack stable/icehouse, and my neutron version is 1409da709594963
I created an internal network not connected to the router using default gateway and dhcp enabled.
I then created a VM which activated dnsmasq, which offered an IP to this VM.
If I mis-update the gateway, by going to Horizon, "Networks", "test-net", "subnet-2" "Edit Subnet", and set
Gateway IP (optional): 10.100.100.100
or mis-update the gateway using neutron CLI,
it results in neutron-dhcp-agent continously restarting the dnsmasq for the "private" network.
Horizon will not allow the user to mis-update the gateway, if I set "force_
But this parameter does not affect the behavior of the neutron CLI.
> neutron net-create test-net
Created a new network:
+------
| Field | Value |
+------
| admin_state_up | True |
| id | b7bb10bb-
| name | test-net |
| shared | False |
| status | ACTIVE |
| subnets | |
| tenant_id | 8092813be8fd412
+------
> neutron net-list
+------
| id | name | subnets |
+------
| 0dd5722d-
| 27c1649d-
| b7bb10bb-
+------
> neutron subnet-create --name subnet-2 test-net --enable_dhcp=True 10.10.150.0/24
Created a new subnet:
+------
| Field | Value |
+------
| allocation_pools | {"start": "10.10.150.2", "end": "10.10.150.254"} |
| cidr | 10.10.150.0/24 |
| dns_nameservers | |
| enable_dhcp | True |
| gateway_ip | 10.10.150.1 |
| host_routes | |
| id | 4e462af6-
| ip_version | 4 |
| name | subnet-2 |
| network_id | b7bb10bb-
| tenant_id | 8092813be8fd412
+------
> neutron subnet-list
+------
| id | name | cidr | allocation_pools |
+------
| 4e462af6-
| 81859ee5-
+------
> ps -ef |grep dnsmasq | grep -v grep
nobody 995 635 0 16:09 ? 00:00:00 /usr/sbin/dnsmasq --no-resolv --keep-
nobody 12991 1 0 18:21 ? 00:00:00 dnsmasq --no-hosts --no-resolv --strict-order --bind-interfaces --interface=
> nova boot --image cirros-
:
:
> nova list
+------
| ID | Name | Status | Task State | Power State | Networks |
+------
| d98511f7-
| 5934a236-
+------
> ps -ef | grep dnsmasq | grep -v grep
nobody 995 635 0 16:09 ? 00:00:00 /usr/sbin/dnsmasq --no-resolv --keep-
nobody 12991 1 0 18:21 ? 00:00:00 dnsmasq --no-hosts --no-resolv --strict-order --bind-interfaces --interface=
nobody 25284 1 0 19:02 ? 00:00:00 dnsmasq --no-hosts --no-resolv --strict-order --bind-interfaces --interface=
19:17:42 > ps -ef | grep dnsmasq | grep -v grep | awk '{print $2}'
995
12991
25284
> neutron subnet-update 4e462af6-
Updated subnet: 4e462af6-
19:19:36 > ps -ef | grep dnsmasq | grep -v grep | awk '{print $2}'
995
25284
30892
19:19:49 > ps -ef | grep dnsmasq | grep -v grep | awk '{print $2}'
995
25284
30999
19:19:59 > ps -ef | grep dnsmasq | grep -v grep | awk '{print $2}'
995
25284
31137
I get the following output from screen:
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
2014-07-11 19:23:26.212 TRACE neutron.
Changed in neutron: | |
assignee: | nobody → ugvddm (271025598-9) |
Changed in neutron: | |
importance: | Undecided → Medium |
Changed in neutron: | |
status: | New → Confirmed |
Apparently the failure to setup a default gateway leads to a resync for a given subnet, which in turn leads to dnsmasq respawn