VIRL - Routers ABSENT To ACTIVE - Delay

@Jason_Sowers wrote:

All,

I have seen a lot of people having issues with the nodes not passing from [ABSENT] to [ACTIVE] for various reasons - normally it has to do with salt issues.

however, I haven't see anyone answer this one.

When I restart my VM, as far as I can see, all services are up and running -

virl@virl:/var/local/virl/logs$ neutron agent-list
+--------------------------------------+--------------------+------+-------+----------------+
| id | agent_type | host | alive | admin_state_up |
+--------------------------------------+--------------------+------+-------+----------------+
| 3b2e73ce-f765-44cb-b342-82baeb0828c0 | Linux bridge agent | virl | | True |
| 8c490ddf-9de8-4b87-b78b-b10bf543206d | L3 agent | virl | | True |
| a1f56643-d127-4b9e-84a8-65a014e517d5 | DHCP agent | virl | | True |
| ea6bf2f0-3328-4c51-890c-70d09815d1b1 | Metadata agent | virl | | True |
+--------------------------------------+--------------------+------+-------+----------------+
virl@virl:/var/local/virl/logs$ nova service-list
+----+------------------+------+----------+---------+-------+----------------------------+-----------------+
| Id | Binary | Host | Zone | Status | State | Updated_at | Disabled Reason |
+----+------------------+------+----------+---------+-------+----------------------------+-----------------+
| 6 | nova-cert | virl | internal | enabled | up | 2015-11-13T13:27:15.000000 | - |
| 7 | nova-consoleauth | virl | internal | enabled | up | 2015-11-13T13:27:15.000000 | - |
| 8 | nova-scheduler | virl | internal | enabled | up | 2015-11-13T13:27:19.000000 | - |
| 9 | nova-conductor | virl | internal | enabled | up | 2015-11-13T13:27:17.000000 | - |
| 10 | nova-compute | virl | nova | enabled | up | 2015-11-13T13:27:15.000000 | - |
+----+------------------+------+----------+---------+-------+----------------------------+-----------------+

However whenever I restart my VM and then start my simulation, it takes almost exactly 8 hours for my nodes to move from ABSENT to ACTIVE without me touching anything. It doesn't matter if I have 2 nodes or 8, or more. However, once it has started, I am able to stop and restart simulations virtually instantaneously, until my VM reboots again.

See below the console:

(INFO) [Nov/12/2015 19:23:14] Session "My_Topologies@basic-mpls-91qsOH": nodes can start
(INFO) [Nov/12/2015 19:23:21] Starting node "iosv-8"
(INFO) [Nov/12/2015 19:23:27] Starting node "~mgmt-lxc"
(INFO) [Nov/12/2015 19:23:28] Starting node "iosv-5"
(INFO) [Nov/12/2015 19:23:35] Starting node "iosv-7"
(INFO) [Nov/12/2015 19:23:41] Starting node "iosv-6"
(INFO) [Nov/12/2015 19:23:54] Starting node "iosxrv-1"
(INFO) [Nov/12/2015 19:24:00] Starting node "iosv-3"
(INFO) [Nov/12/2015 19:24:06] Starting node "iosv-2"
(INFO) [Nov/12/2015 19:24:15] Starting node "iosxrv-2"
(INFO) [Nov/13/2015 03:22:01] Node "iosv-2" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosv-3" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosv-5" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosv-6" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosv-7" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosv-8" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosxrv-1" state changed from ABSENT to ACTIVE
(INFO) [Nov/13/2015 03:22:01] Node "iosxrv-2" state changed from ABSENT to ACTIVE

Thoughts/ideas? I thought it was the clock at first, double-checked that the timezone was correct, even then, I am GMT -6 so the 8 doesn't fit.

Thanks in advance for any ideas. Remember that if I need to test it, it could be 8 hours before I can test again.

Posts: 1

Participants: 1

Read full topic