Earlier  
Posted Nick Remark
#openstack-sdks - 2017-02-02
20:02:35 ankur-gupta-f1 dtroyer: sindhu who has done quite a bit of Neutron command migration will be at PTG.
20:03:53 dtroyer ok, good
20:05:35 ankur-gupta-f1 dtroyer: nvm. I was wrong. None of us :'(
20:05:43 ankur-gupta-f1 we are all quite bitter about it
20:08:14 stevemar ankur-gupta-f1: grrr
20:08:15 dtroyer I know there has been some question about how to decide between PTG and Forum, I've tried to help clear that up as I find it
20:08:29 stevemar dtroyer: so i *think* reno will support renaming files too
20:09:22 stevemar but the changes may have to be backported? i'm not sure
20:09:48 stevemar changing locations
20:10:07 dtroyer to have the changes show up both in 'unreleased' and 'ocata' (for example) I think you are right
20:10:39 dtroyer all-in-all much easier to get them clean before releasing
20:17:38 filow hi there, i need some help with the openstackclient. I'm trying to run the example but it seems there is something not updated with the last commit. Does anyone try to run the example from the git?
20:19:38 dtroyer filow: the examples are likely out of date (why stevemar did not want me to merge them in the first place)
20:19:47 dtroyer what were you trying to accomplish with them?
20:20:10 filow well i'm trying to run the command openstack user list --project <id or name>
20:20:30 filow i saw there is still the key verify that is not anymore supported
20:21:36 filow but then i can't go ahead debugging, i got from debug "'OpenStackConfig' object has no attribute 'load_auth_plugin'"
20:22:56 dtroyer I am not sure what you mean by "key verify"
20:23:45 filow the constructor clientmanager.ClientManager doesn't require verify as parameter, while it is still there in the example
20:24:09 dtroyer ah, yes. that is what I men by out of date
20:24:32 filow i'm talking about osc-lib.py filew in example
20:24:39 filow yep i got it indeed
20:24:40 dtroyer I was using those as a place to work out some of the changes we made to osc around 3.0 and the osc-lib split, and also with os-client-config
20:25:10 filow infact the second error comes from osc-lib
20:26:20 filow from what i see the usage once you have the clientManager is straight forward, but it is not clear how to set it up (if the flow is still up to date)
20:26:47 filow i got that in common the parser get the environment vars you usually set up with your project-openrc.sh
20:35:51 filow does anyone know an example of how to configure the clientManager that actually works?
20:43:06 openstackgerrit Ankur proposed openstack/python-openstacksdk master: Introduce Base for Octavia (load balancing) https://review.openstack.org/428414
20:49:12 dtroyer filow: besides OSC itself, I doubt it. but those examples were created with a lot of copy-n-paste from the OSC code, fixing them shouldn't be too hard
20:52:20 filow dtroyer: ok, coz first i tried with keystone client and i have some unexpected behaviour, indeed i wanted to use the same approach of python-openstack. Thanks anyway
21:43:04 openstackgerrit Ankur proposed openstack/python-openstacksdk master: Introduce Base for Octavia (load balancing) https://review.openstack.org/428414
22:06:14 ankur-gupta-f1 briancurtin: ping you still around? how does the sdk know the endpoints to hit...?
22:10:00 openstackgerrit Merged openstack/python-openstackclient master: Update reno for stable/ocata https://review.openstack.org/427513
22:14:42 openstackgerrit Ankur proposed openstack/python-openstacksdk master: Introduce Base for Octavia (load balancing) https://review.openstack.org/428414
22:16:55 openstackgerrit Ankur proposed openstack/python-openstacksdk master: Introduce Base for Octavia (load balancing) https://review.openstack.org/428414
22:23:14 briancurtin ankur-gupta-f1: we look in the service catalog for the service_name you set on the ServiceFilter for that service, which in your case you set load_balancingv2 in load_balancer_service.py, and then we make a request to the root of that service to find the versions it supports and see that they match up with ones that we have in the library. if it finds a v2,
22:23:14 briancurtin we have a v2 in the SDK, so it takes that as the endpoint
22:24:02 briancurtin ankur-gupta-f1: i’m about to leave for the day, but if you need to know more about endpoint selection, it happens in openstack/session.py
22:25:00 ankur-gupta-f1 briancurtin: okay will take a look. Currently octavia only has v1 but in middle of implementing v2.0 so thats probably the way to go. Will dig in. Thanks for your help. Enjoy the rest of your evening
22:25:38 briancurtin oh yeah, if they only have v1 out then we’ll have to just support that to start with, and then adding in v2 once it’s out will be pretty easy
22:26:09 ankur-gupta-f1 running 'openstack service list' i see octavia. so thats a good sign :D
22:27:58 dtroyer ankur-gupta-f1: one side note, if you have any ability to suggest it, let's try to keep octavia from adding a new service catalog entry with 'v2' in the name/type. Since they haven't shipped it yet, they really should do proper version discovery up front.
22:29:18 ankur-gupta-f1 dtroyer: no problem. Im working on their v2 implementation. will make sure it happens
#openstack-sdks - 2017-02-03
02:31:35 openstackgerrit Steve Martinelli proposed openstack/python-openstackclient master: mention the final ocata osc version in releasenotes https://review.openstack.org/428545
03:37:39 openstackgerrit Ankur proposed openstack/python-openstacksdk master: Introduce Base for Octavia (load balancing) https://review.openstack.org/428414
04:37:53 reedip sindhu : Just came to office, have a bad hangover without being drunk ! :P ( Lack of sleep evidently ) . Will look into it :)
04:42:50 sindhu reedip: can you check if router set command is working for setting the external_gateway_info
04:43:00 sindhu reedip: thanks!
04:43:14 reedip sindhu : in your patch ? Just wait :)
04:43:51 sindhu reedip: no it's not there in my patch. I was trying it locally
04:44:00 reedip Okay , just wait
04:50:24 reedip sindhu : u awake ?
04:50:39 reedip I dont know but m guessing your TZ is opposite of mine :)
04:51:33 reedip sindhu : the external-gateway-info works well
04:51:37 sindhu reedip: I m in US
04:52:21 sindhu reedip: can u send the command ur testing
04:52:33 reedip sindhu : doing just that
04:53:15 stevemar huanxuan: dtroyer i've updated the cinder vs osc spreadsheet -- but extended it to all projects: https://docs.google.com/spreadsheets/d/18ZtWC75BNCwFqLfFpCGGJ9uPVBvUXX0xuXP1yYG0NDA/edit?usp=sharing
04:53:21 reedip sindhu : there you go : http://paste.openstack.org/show/597483/
04:55:05 reedip sindhu : whats failing in your patch, havent seen it yet
04:58:14 sindhu reedip: I tried doing the same, but with different external network that I created
04:58:21 sindhu reedip: it didn't work
04:58:58 reedip sindhu : did your network have Subnets?
04:59:08 sindhu reedip: yes
04:59:22 sindhu reedip: I did those 8 steps I mentioned earlier
04:59:32 reedip sindhu : ok, send me the paste of the commands , and the error which you saw
04:59:52 openstackgerrit Ankur proposed openstack/python-openstackclient master: OSC Extension Show https://review.openstack.org/382023
05:00:25 sindhu https://www.irccloud.com/pastebin/wb9EARvJ/
05:01:43 reedip sindhu : what was the Subnet associated with the port created in step 6 ?
05:02:43 sindhu reedip: ^^^
05:03:10 sindhu reedip: the private network's subnet
05:09:57 reedip sindhu : checking
05:12:57 reedip sindhu : you created a port with internal Fixed IP , and then you tried to attach the port to the router using external subnet, right ?
05:15:24 sindhu reedip: http://paste.openstack.org/show/597485/
05:15:36 reedip sindhu : checking
05:25:55 reedip sindhu : did you run it using --debug ?
05:25:59 reedip the last CLI
05:26:26 reedip I am attempting to replicate the issue at my end, but meanwhile you can try it using --0debug
05:26:30 reedip I am attempting to replicate the issue at my end, but meanwhile you can try it using --debug *
05:26:51 sindhu reedip: will do
05:27:28 reedip sindhu : when I ran the CLI, I got the response HttpException: Bad Request, Bad router request: Router already has a port on subnet 3f9ccf79-de05-4caf-8b1e-2d6f80f42561.
05:28:33 sindhu reedip: yeah.. even I am getting the same
05:30:03 reedip sindhu : oops, you already added the subnet in step 7
05:30:20 reedip so the router would already have a port for that subnet
05:37:48 reedip sindhu : you can run "openstack port list --router" to see if the router already has a port
05:38:28 reedip and then run port show for the ports connected to the router and you will see that the router already has a port connected to the external subnet, so you cannot set the gateway to that port
05:40:49 sindhu reedip: oh ok ... but if I create a floating ip in the external_net and try to set that to the port it says: External network e01d4875-68dd-44bf-a871-cba53519d20f is not reachable from subnet 9154650e-aa76-44bd-9135-1c7d8ff4534d.
05:41:53 reedip Thats bcz its the Subnet ID of the Pvt network
05:42:31 reedip do one thing, remove the port which is connected to the external network
05:42:39 reedip and execute step 8 once again
05:43:35 sindhu reedip: you are right ... the router has two ports connected to it now. One in private network and one in external network
05:43:55 reedip sindhu : and your router currently has the gateway set at the Pvt network
05:44:04 reedip IIRC
05:44:48 sindhu reedip: correct
05:44:58 reedip the floating IP basically lies on one side of the router's gateway, and the ports of the router lie on the other side of the gateway
05:45:49 reedip so if the gateway is incorrectly directed, it should not be able to route, so try to remove the port which is connected to the external network, try setting the external gateway and then associate floating IP
05:46:26 sindhu reedip: so I went wrong in doing: openstack router add subnet <external_subnet> ??
05:46:42 reedip sindhu : seems so ,
05:47:16 sindhu reedip: ok will try removing the port connected to external net
07:25:52 openstackgerrit Huanxuan Ao proposed openstack/python-openstackclient master: Fix properties format for volume qos in volume v1 https://review.openstack.org/428610

Earlier   Later