| *** trozet has quit IRC | 00:02 | |
| *** mbound has quit IRC | 00:46 | |
| *** lhcheng has quit IRC | 00:56 | |
| *** lhcheng has joined #tacker | 01:15 | |
| *** amotoki has joined #tacker | 01:39 | |
| *** ahelkhou1 has joined #tacker | 01:41 | |
| *** ahelkhou2 has quit IRC | 01:41 | |
| *** mbound has joined #tacker | 01:46 | |
| *** mbound has quit IRC | 01:51 | |
| *** jaff_cheng has joined #tacker | 01:57 | |
| *** trozet has joined #tacker | 02:11 | |
| *** hang has joined #tacker | 02:30 | |
| *** jaff_cheng has quit IRC | 02:32 | |
| *** sridhar_ram1 has quit IRC | 02:59 | |
| *** ahelkhou has quit IRC | 03:04 | |
| *** amotoki has quit IRC | 03:10 | |
| *** trozet has quit IRC | 03:11 | |
| *** amotoki has joined #tacker | 03:21 | |
| *** 18WABVZ4V has quit IRC | 03:27 | |
| *** bobh has joined #tacker | 03:28 | |
| *** amotoki has quit IRC | 03:30 | |
| *** ahelkhou has joined #tacker | 03:31 | |
| *** amotoki has joined #tacker | 03:35 | |
| *** bobh has quit IRC | 03:42 | |
| *** amotoki has quit IRC | 04:01 | |
| *** ahelkhou1 has quit IRC | 04:22 | |
| *** jaff_cheng has joined #tacker | 04:30 | |
| *** hang has quit IRC | 04:33 | |
| *** lhcheng has quit IRC | 04:57 | |
| *** ahelkhou1 has joined #tacker | 05:27 | |
| *** jaff_cheng has quit IRC | 05:33 | |
| *** vishwanathj has joined #tacker | 05:37 | |
| *** vishwanathj has quit IRC | 05:37 | |
| *** jaff_cheng has joined #tacker | 05:54 | |
| *** ahelkhou1 has quit IRC | 06:03 | |
| *** hang has joined #tacker | 06:18 | |
| *** jaff_cheng has quit IRC | 06:21 | |
| *** hang has quit IRC | 06:31 | |
| *** lhcheng has joined #tacker | 06:46 | |
| *** lhcheng has quit IRC | 06:50 | |
| *** gongysh has joined #tacker | 07:08 | |
| *** amotoki_ has joined #tacker | 07:09 | |
| *** gongysh has quit IRC | 07:13 | |
| *** lhcheng has joined #tacker | 07:21 | |
| *** gongysh has joined #tacker | 07:26 | |
| *** vishwanathj_ has joined #tacker | 07:28 | |
| *** vishwanathj_ has left #tacker | 07:28 | |
| *** amotoki_ is now known as amotoki | 07:56 | |
| kluka_ | Hello, I have a question about monitoring. Is it possible to integrate Tacker with Zabbix or some similar software? | 08:19 |
|---|---|---|
| kluka_ | Are there any plans to provide this functionality? | 08:19 |
| *** amotoki has quit IRC | 08:21 | |
| *** gongysh has quit IRC | 08:33 | |
| *** zeih has joined #tacker | 08:38 | |
| *** amotoki has joined #tacker | 08:40 | |
| *** amotoki has quit IRC | 08:49 | |
| *** gongysh has joined #tacker | 08:52 | |
| *** amotoki has joined #tacker | 08:52 | |
| *** amotoki has quit IRC | 08:55 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/tacker: Fixing default mutable argument issue. https://review.openstack.org/271952 | 08:57 |
| *** amotoki has joined #tacker | 09:08 | |
| *** amotoki has quit IRC | 09:18 | |
| *** mbound has joined #tacker | 09:18 | |
| openstackgerrit | Martin Oemke proposed openstack/tacker-horizon: Update setup.py https://review.openstack.org/271968 | 09:30 |
| *** amotoki has joined #tacker | 09:30 | |
| *** zeih has quit IRC | 10:03 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/tacker: Code cleanup at multiple places. https://review.openstack.org/271993 | 10:11 |
| gongysh | sripriya, hi | 10:40 |
| *** gongysh has quit IRC | 11:01 | |
| *** gubouvier_ has joined #tacker | 12:52 | |
| *** lhcheng has quit IRC | 12:53 | |
| *** gubouvier has quit IRC | 12:53 | |
| *** zeih has joined #tacker | 13:04 | |
| *** bobh has joined #tacker | 13:10 | |
| *** gubouvier_ is now known as gubouvier | 13:12 | |
| *** bobh has quit IRC | 13:14 | |
| *** amotoki has quit IRC | 13:19 | |
| *** ahelkhou1 has joined #tacker | 13:24 | |
| *** zeih has quit IRC | 13:25 | |
| *** sridhar_ram has joined #tacker | 13:30 | |
| *** zeih has joined #tacker | 13:42 | |
| *** gongysh has joined #tacker | 13:53 | |
| *** ahelkhou1 has quit IRC | 13:58 | |
| *** dkushwaha has left #tacker | 14:00 | |
| openstackgerrit | Andreas Jaeger proposed openstack/tacker-horizon: Update setup.py https://review.openstack.org/271968 | 14:22 |
| *** gongysh has quit IRC | 14:34 | |
| *** vishwanathj has joined #tacker | 14:42 | |
| *** trozet has joined #tacker | 14:52 | |
| *** bobh has joined #tacker | 15:02 | |
| *** ahelkhou1 has joined #tacker | 15:02 | |
| *** ahelkhou2 has joined #tacker | 15:03 | |
| *** mbound has quit IRC | 15:07 | |
| *** ahelkhou1 has quit IRC | 15:07 | |
| *** ahelkhou2 has quit IRC | 15:34 | |
| *** zeih has quit IRC | 15:43 | |
| *** mbound has joined #tacker | 16:05 | |
| *** mbound has quit IRC | 16:22 | |
| *** ahelkhou1 has joined #tacker | 16:26 | |
| *** zeih has joined #tacker | 16:59 | |
| *** mbound has joined #tacker | 17:23 | |
| *** lhcheng has joined #tacker | 17:24 | |
| *** mbound has quit IRC | 17:27 | |
| openstackgerrit | Martin Oemke proposed openstack/tacker-horizon: Fix pep8 errors in tacker-horizon https://review.openstack.org/272193 | 17:29 |
| *** mbound has joined #tacker | 17:33 | |
| *** sripriya has quit IRC | 17:49 | |
| *** prashantD_ has joined #tacker | 17:52 | |
| *** mbound has quit IRC | 17:55 | |
| *** sripriya has joined #tacker | 18:14 | |
| *** gubouvier has quit IRC | 18:31 | |
| *** santoshk has joined #tacker | 18:31 | |
| *** lhcheng_ has joined #tacker | 18:34 | |
| *** lhcheng has quit IRC | 18:36 | |
| *** zeih has quit IRC | 18:47 | |
| *** openstackgerrit has quit IRC | 19:02 | |
| *** openstackgerrit has joined #tacker | 19:03 | |
| *** zeih has joined #tacker | 19:08 | |
| *** zeih has quit IRC | 19:14 | |
| *** mbound has joined #tacker | 19:21 | |
| sripriya | kluka_: reg: zabbix, the tool looks interesting. yes it should be possible to integrate any monitoring tool within tacker with the mgmt driver architecture. are you looking at a specific use case where this monitoring tool needs to be used? | 19:23 |
| *** mbound has quit IRC | 19:45 | |
| openstackgerrit | Santosh Kodicherla proposed openstack/tacker: Developers guide for functional tests https://review.openstack.org/246183 | 20:14 |
| openstackgerrit | Santosh Kodicherla proposed openstack/tacker: Developers guide for functional tests https://review.openstack.org/246183 | 20:20 |
| *** mbound has joined #tacker | 20:22 | |
| openstackgerrit | Santosh Kodicherla proposed openstack/tacker: Developers guide for functional tests https://review.openstack.org/246183 | 20:32 |
| openstackgerrit | Santosh Kodicherla proposed openstack/tacker: Developers guide for functional tests https://review.openstack.org/246183 | 20:33 |
| *** kluka__ has joined #tacker | 20:48 | |
| kluka__ | sripriya, thanks for the info about Zabbix. I am not looking for specific use case, just a general question :) | 20:50 |
| kluka__ | sripriya, but if there is more info available somewhere I would be happy see it :) | 20:51 |
| kluka__ | sripriya, if I understand Tacker monitoring corrctly each VDU needs a public mgmt IP (floating IP)? Please correct me if I am wrong. | 20:52 |
| kluka__ | sripriya, so that Monitor process can reach the VDU. | 20:53 |
| sripriya | kluka__: thats correct | 20:53 |
| sripriya | https://github.com/openstack/tacker/blob/master/doc/source/devref/monitor-api.rst | 20:53 |
| sripriya | kluka__: if this can help to undertand the monitoring framework | 20:53 |
| kluka__ | sripriya, Zabbix has an agent that connects to a server that can be outside Openstack. So there is no need to give each VDU public IP address and they can be hidden behind the NAT. | 20:55 |
| kluka__ | sripriya, I can see some benefits with this approach. | 20:56 |
| *** uck has joined #tacker | 20:56 | |
| *** amotoki has joined #tacker | 20:56 | |
| kluka__ | Ok, I will check your link. Thanks again ;) | 20:57 |
| sripriya | kluka__: i was just looking at the features. so is the server responsible for controlling the actual monitoring policies? | 21:03 |
| sripriya | kluka__: the actual management of VNFs is still under Tacker control whatever the backend monitoring tool is | 21:07 |
| kluka__ | sripriya: zabbix can be configured to monitor VDU with installed zabbix-agent, and if something goes wrong zabbix can notify Tacker about what happend... After that Tacker needs to decide what to do according to monitoring policy. | 21:09 |
| kluka__ | At least that's how I imagine it :) | 21:10 |
| kluka__ | But I am not sure how can this be implemented with current monitoring framework. | 21:11 |
| kluka__ | Guys on Openbaton project (http://openbaton.github.io/) are using this approach. You can check zabbix plugin. | 21:12 |
| sripriya | kluka__: Tacker currently does not have a way to consume notifications within monitoring drivers, that will be an area to explore | 21:15 |
| sripriya | kluka__: we did see openbaton project. for some other features, good to know they support this approach | 21:17 |
| *** sridhar_ram has quit IRC | 21:19 | |
| kluka__ | sripriya: Okey thanks for the info :) | 21:19 |
| kluka__ | Maybe it would be possible to write custom monitoring driver that would pool zabbix server instead of the acctual VDU or something like that :) | 21:19 |
| kluka__ | I have to try this some day. | 21:20 |
| *** sridhar_ram1 has joined #tacker | 21:21 | |
| kluka__ | The down side of this approach is that you need a zabbix-agent on VDU. And it also needs to be configured correctly so that it can connect to zabbix-server. | 21:21 |
| *** lhcheng has joined #tacker | 21:21 | |
| kluka__ | Guys at Openbaton are using cloud-init for zabbix-agent configuration.. | 21:22 |
| sripriya | kluka__: thats what i was thinking, i do see they have agentless monitoring, what about that | 21:22 |
| sripriya | klula__: isee | 21:22 |
| kluka__ | Hm, I haven't check this out yet... | 21:23 |
| sripriya | kluka__: that actually avoids installing agent on VDU and goes through other ways to monitor VDU like icmp ping and stuff | 21:23 |
| *** lhcheng_ has quit IRC | 21:24 | |
| sripriya | and Tacker could query the server appropriately | 21:24 |
| sripriya | server--> zabbix server | 21:24 |
| kluka__ | sripriya: tabbix server has a REST api, so I think it would be possible to query it about VDUs... but I also haven't tested this yet :) | 21:26 |
| kluka__ | zabbix* | 21:26 |
| sripriya | kluka__: i was also looking into their APIs, they have an API to query hosts | 21:28 |
| sripriya | https://www.zabbix.com/documentation/2.2/manual/api#performing_requests | 21:28 |
| kluka__ | I am just looking that :) It would also be possible to configure monitoring according to monitoring policy given in VNFD... | 21:30 |
| *** gongysh has joined #tacker | 21:30 | |
| kluka__ | Zabbix is an interesing and useful project | 21:30 |
| gongysh | vishwanathj, hi | 21:31 |
| sripriya | gongysh: pinging you back, i missed your ping earlier | 21:32 |
| gongysh | sripriya, hi | 21:32 |
| kluka__ | here is more detailed list of all the APIs that coud be used: https://www.zabbix.com/documentation/2.2/manual/api/reference | 21:32 |
| gongysh | I want to know why the status of vnf with monitor is active-dead-active? | 21:33 |
| sripriya | kluka__: yes | 21:33 |
| kluka__ | sripriya: ok, I have to run now, but thanks for this chat ;) | 21:34 |
| kluka__ | Have a nice day, yall | 21:34 |
| sripriya | gongysh: to show the intermittent unreachability of the VNF | 21:34 |
| sripriya | kluka__: thank you for this discussion | 21:34 |
| gongysh | sripriya, what is the first active? doesn't it mean the vdu is up? | 21:35 |
| sripriya | gongysh:yes,it means it is up, and healthy | 21:36 |
| gongysh | so why is there dead then? | 21:36 |
| sripriya | gongysh: anytime during VNF life cycle, it becomes unreachable is when its status is set to DEAD state and then back to ACTIVE, once the background VM is respun | 21:37 |
| sripriya | gongysh: VNF once it is up is constantly monitored for reachability | 21:37 |
| vishwanathj | gongysh, hi | 21:38 |
| gongysh | sripriya, I am just wondering why our function tests with monitor has the fixed pattern "active-dead-active"? | 21:38 |
| gongysh | vishwanathj, hi | 21:38 |
| gongysh | vishwanathj, do u update the topic? | 21:38 |
| sripriya | gongysh: because, to test that feature, we forcefully bring the VNF down to test if it actually goes to DEAD from being ACTIVE and then finally going back to ACTIVE | 21:39 |
| vishwanathj | gongysh, I am working on it and hope to have an update by Wednesday | 21:41 |
| gongysh | vishwanathj, great, thanks | 21:42 |
| gongysh | sripriya, I cannot see this codes at:https://github.com/openstack/tacker/blob/master/tacker/tests/functional/vnfm/base.py#L100 | 21:42 |
| gongysh | sripriya, do you mean we forcefully bring the VNF down, but I cannot locate the codes doing this. | 21:43 |
| sripriya | gongyshL: it happens in the template, once the VNF boots up | 21:43 |
| sripriya | https://github.com/openstack/tacker/blob/master/tacker/tests/etc/samples/sample-vnfd-single-vdu-monitoring.yaml#L19 | 21:43 |
| sripriya | through the user_data part | 21:44 |
| sripriya | thats basically like a post init script that gets run once the VM comes up | 21:44 |
| gongysh | sripriya, https://github.com/openstack/tacker/blob/master/tacker/tests/etc/samples/sample-vnfd-single-vdu-monitoring.yaml#L19 | 21:46 |
| gongysh | sripriya, the 'sudo ifdown eth0'? | 21:47 |
| sripriya | gongysh: yes | 21:47 |
| gongysh | sripriya, so this is the dead. | 21:48 |
| sripriya | yup | 21:48 |
| gongysh | sripriya, then about repawn, what is our way, to boot a new VM? | 21:48 |
| sripriya | not necesary, depends on the Action Policy, | 21:49 |
| sripriya | if re boot is what is configured, then that will happen | 21:49 |
| sripriya | if you just need to log the status and take no further action that is also possible | 21:49 |
| gongysh | sripriya, https://github.com/openstack/tacker/blob/master/tacker/vm/monitor.py#L213 | 21:53 |
| gongysh | sripriya, ActionRespawn, it is not add the new device_id into monitor | 21:53 |
| gongysh | sripriya, https://github.com/openstack/tacker/blob/master/tacker/vm/monitor.py#L244 | 21:54 |
| gongysh | sripriya, @ActionPolicy.register('respawn', 'heat') | 21:54 |
| gongysh | class ActionRespawnHeat(ActionPolicy): | 21:54 |
| sripriya | https://github.com/openstack/tacker/blob/master/tacker/vm/monitor.py#L243 | 21:54 |
| gongysh | this one does. | 21:54 |
| sripriya | gongysh: yes that is the one | 21:54 |
| gongysh | so what is useful for L213 one? | 21:55 |
| gongysh | it seems it is a bugy one. | 21:55 |
| sripriya | from the code, it brings up a brand new VNF instead of updating existing VNF | 21:57 |
| sripriya | i have never tested that code though | 21:57 |
| sripriya | it basically sets the current VNF to DEAD and also brings up a new VNF | 21:58 |
| *** gongysh has quit IRC | 22:01 | |
| *** gongysh has joined #tacker | 22:01 | |
| gongysh | sripriya, sorry disconnected. back again. the L244(heat respawn one) also brings up a new VNF. | 22:04 |
| *** uck has quit IRC | 22:04 | |
| gongysh | sripriya, and can you help to review https://review.openstack.org/#/c/255160/, after that I can upload a new one patch to fix the conflict. | 22:05 |
| *** bobh has quit IRC | 22:06 | |
| gongysh | sripriya, I have to board. see you. thanks | 22:07 |
| *** gongysh has quit IRC | 22:07 | |
| *** bobh has joined #tacker | 22:11 | |
| sripriya | gongysh: sure | 22:11 |
| *** uck has joined #tacker | 22:14 | |
| *** bobh has quit IRC | 22:15 | |
| openstackgerrit | Santosh Kodicherla proposed openstack/tacker: Developers guide for functional tests https://review.openstack.org/246183 | 22:30 |
| *** s3wong has joined #tacker | 22:34 | |
| *** bobh has joined #tacker | 23:02 | |
| *** amotoki has quit IRC | 23:12 | |
| *** amotoki has joined #tacker | 23:15 | |
| *** amotoki has quit IRC | 23:16 | |
| *** trozet has quit IRC | 23:22 | |
| *** vishwanathj has quit IRC | 23:23 | |
| *** bobh has quit IRC | 23:26 | |
| *** ahelkhou1 has quit IRC | 23:31 | |
| *** mbound has quit IRC | 23:49 | |
| *** santoshk has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!