| *** kaganos has joined #openstack-dev | 00:01 | |
| *** kaganos_ has joined #openstack-dev | 00:01 | |
| *** markmc has quit IRC | 00:02 | |
| *** dani4571 has quit IRC | 00:03 | |
| *** salv-orlando has quit IRC | 00:03 | |
| *** salv-orlando has joined #openstack-dev | 00:03 | |
| *** kaganos has quit IRC | 00:05 | |
| *** kaganos_ is now known as kaganos | 00:05 | |
| *** hemna has quit IRC | 00:07 | |
| *** jcooley is now known as jcooley_afk | 00:09 | |
| *** jergerber has quit IRC | 00:10 | |
| *** samkottler is now known as samkottler|afk | 00:11 | |
| *** sthaha has joined #openstack-dev | 00:12 | |
| *** nati_ueno has quit IRC | 00:13 | |
| *** nati_ueno has joined #openstack-dev | 00:14 | |
| *** mrodden has quit IRC | 00:15 | |
| *** rpedde is now known as rpedde_away | 00:15 | |
| zykes- | anyone, which of the two: moniker <resource>-get or <resource>-show to get details for a resource? | 00:16 |
|---|---|---|
| *** ewindisch has joined #openstack-dev | 00:18 | |
| *** TerryH has quit IRC | 00:20 | |
| *** lloydde has quit IRC | 00:22 | |
| *** jcooley_afk is now known as jcooley | 00:23 | |
| *** ewindisch has quit IRC | 00:23 | |
| *** kaganos has quit IRC | 00:24 | |
| *** TerryH has joined #openstack-dev | 00:26 | |
| *** joshuamckenty has quit IRC | 00:26 | |
| *** statik has quit IRC | 00:27 | |
| *** statik has joined #openstack-dev | 00:28 | |
| *** gongysh has quit IRC | 00:28 | |
| *** edmund has quit IRC | 00:31 | |
| *** mrodden has joined #openstack-dev | 00:34 | |
| *** patelna has quit IRC | 00:37 | |
| *** dingus9 has quit IRC | 00:38 | |
| *** flaper87 has quit IRC | 00:39 | |
| *** lloydde has joined #openstack-dev | 00:40 | |
| *** winston-d has joined #openstack-dev | 00:48 | |
| *** maxiz has quit IRC | 00:48 | |
| *** lloydde_ has joined #openstack-dev | 00:52 | |
| *** lloydde has quit IRC | 00:54 | |
| *** joshuamckenty has joined #openstack-dev | 00:56 | |
| *** edygarcia has joined #openstack-dev | 00:56 | |
| *** joshuamckenty has quit IRC | 00:56 | |
| *** woorea has joined #openstack-dev | 01:01 | |
| *** adjohn has quit IRC | 01:03 | |
| *** winston-d has quit IRC | 01:04 | |
| *** Dr_Who has joined #openstack-dev | 01:06 | |
| *** Dr_Who has joined #openstack-dev | 01:06 | |
| *** jaypipes has quit IRC | 01:07 | |
| *** andrewbogott is now known as andrewbogott_afk | 01:08 | |
| *** ewindisch has joined #openstack-dev | 01:08 | |
| *** alunduil has joined #openstack-dev | 01:08 | |
| *** alexpilotti has quit IRC | 01:09 | |
| *** ewindisch has quit IRC | 01:09 | |
| *** halj has quit IRC | 01:14 | |
| *** Dr_Who has quit IRC | 01:15 | |
| *** winston-d has joined #openstack-dev | 01:16 | |
| *** utlemming has quit IRC | 01:19 | |
| *** TerryH has quit IRC | 01:19 | |
| *** jeremielim has joined #openstack-dev | 01:20 | |
| *** pabelanger has joined #openstack-dev | 01:20 | |
| *** gyee has quit IRC | 01:21 | |
| *** sacharya has quit IRC | 01:21 | |
| *** samkottler|afk is now known as samkottler | 01:22 | |
| *** jakedahn_zz is now known as jakedahn | 01:22 | |
| *** jakedahn is now known as jakedahn_zz | 01:22 | |
| *** markmcclain has quit IRC | 01:23 | |
| *** adjohn has joined #openstack-dev | 01:25 | |
| *** maurosr has joined #openstack-dev | 01:29 | |
| *** woorea has quit IRC | 01:31 | |
| *** andrewbogott_afk is now known as andrewbogott | 01:32 | |
| *** jeremielim has quit IRC | 01:33 | |
| *** lloydde_ has quit IRC | 01:36 | |
| *** xchu has joined #openstack-dev | 01:36 | |
| *** sthaha has quit IRC | 01:37 | |
| *** novas0x2a|laptop has quit IRC | 01:38 | |
| *** kaganos has joined #openstack-dev | 01:39 | |
| *** dims has quit IRC | 01:40 | |
| *** Ryan_Lane has quit IRC | 01:41 | |
| *** otherwiseguy has quit IRC | 01:44 | |
| *** jog0 has quit IRC | 01:44 | |
| *** esp1 has quit IRC | 01:46 | |
| *** jeremielim has joined #openstack-dev | 01:47 | |
| *** jcooley is now known as jcooley_afk | 01:48 | |
| *** vipul is now known as vipul|away | 01:49 | |
| *** salv-orlando has quit IRC | 01:51 | |
| *** maxiz has joined #openstack-dev | 01:54 | |
| *** ewindisch has joined #openstack-dev | 01:54 | |
| *** dims has joined #openstack-dev | 01:55 | |
| *** vipul|away is now known as vipul | 01:57 | |
| *** jeremyb has quit IRC | 01:57 | |
| *** jeremyb has joined #openstack-dev | 01:57 | |
| *** imsplitbit has quit IRC | 01:58 | |
| *** Mandell has quit IRC | 02:03 | |
| *** blamar has quit IRC | 02:04 | |
| *** roadTripper has quit IRC | 02:08 | |
| *** ywu_ has joined #openstack-dev | 02:11 | |
| *** dingus9 has joined #openstack-dev | 02:16 | |
| *** sgran has quit IRC | 02:20 | |
| *** sgran has joined #openstack-dev | 02:21 | |
| *** adjohn has quit IRC | 02:23 | |
| *** esp1 has joined #openstack-dev | 02:30 | |
| *** sayjay has quit IRC | 02:30 | |
| *** danwent has quit IRC | 02:37 | |
| *** dolphm has joined #openstack-dev | 02:38 | |
| *** annegentle_ has joined #openstack-dev | 02:43 | |
| *** blamar has joined #openstack-dev | 02:45 | |
| *** ahardy_ has quit IRC | 02:46 | |
| *** sayjay has joined #openstack-dev | 02:52 | |
| *** adjohn has joined #openstack-dev | 02:58 | |
| *** nubi3_ has joined #openstack-dev | 02:59 | |
| *** jeremielim has quit IRC | 02:59 | |
| nubi3_ | hay guys, any ones can help my about arp missing on diablo ??? | 02:59 |
| openstackgerrit | A change was merged to openstack/cinder: Improve provider_location cleanup code for RBD. https://review.openstack.org/17023 | 03:00 |
| *** annegentle_ has quit IRC | 03:00 | |
| *** annegentle_ has joined #openstack-dev | 03:02 | |
| *** nubi3_ has quit IRC | 03:05 | |
| *** maurosr has quit IRC | 03:05 | |
| *** blamar has quit IRC | 03:07 | |
| *** pixelbeat has quit IRC | 03:07 | |
| *** woorea has joined #openstack-dev | 03:09 | |
| *** blamar has joined #openstack-dev | 03:09 | |
| *** nati_ueno_2 has joined #openstack-dev | 03:14 | |
| *** sthaha has joined #openstack-dev | 03:14 | |
| *** nati_ueno has quit IRC | 03:16 | |
| *** mattstep has quit IRC | 03:16 | |
| *** mattstep has joined #openstack-dev | 03:18 | |
| *** darjeeling has quit IRC | 03:19 | |
| *** annegentle_ has quit IRC | 03:20 | |
| *** woorea has quit IRC | 03:22 | |
| *** ewindisch has quit IRC | 03:24 | |
| *** SuYan has joined #openstack-dev | 03:24 | |
| *** andrewbogott is now known as andrewbogott_afk | 03:24 | |
| *** kaganos has quit IRC | 03:28 | |
| *** Mandell has joined #openstack-dev | 03:30 | |
| *** otherwiseguy has joined #openstack-dev | 03:31 | |
| *** ywu__ has joined #openstack-dev | 03:32 | |
| *** ywu_ has quit IRC | 03:32 | |
| *** Mandell has quit IRC | 03:34 | |
| *** jcmartin has joined #openstack-dev | 03:37 | |
| *** hemna has joined #openstack-dev | 03:37 | |
| *** jeremyb has quit IRC | 03:40 | |
| *** jeremyb has joined #openstack-dev | 03:40 | |
| *** ywu__ has quit IRC | 03:42 | |
| *** esp1 has quit IRC | 03:48 | |
| *** jtran has joined #openstack-dev | 03:48 | |
| openstackgerrit | A change was merged to openstack/cinder: Fix import order to make it alphabetical. https://review.openstack.org/16981 | 03:54 |
| *** aeperezt has quit IRC | 03:59 | |
| *** thingee has joined #openstack-dev | 04:01 | |
| *** Mandell has joined #openstack-dev | 04:03 | |
| *** edygarcia has quit IRC | 04:06 | |
| *** nati_ueno has joined #openstack-dev | 04:08 | |
| *** nati_ueno has quit IRC | 04:08 | |
| *** nati_ueno has joined #openstack-dev | 04:09 | |
| *** nati_ueno_2 has quit IRC | 04:11 | |
| *** jaypipes has joined #openstack-dev | 04:11 | |
| *** aeperezt has joined #openstack-dev | 04:11 | |
| *** pabelanger has quit IRC | 04:12 | |
| *** jcooley_afk is now known as jcooley | 04:26 | |
| *** jcooley is now known as jcooley_afk | 04:28 | |
| *** sthaha has quit IRC | 04:31 | |
| *** rmk has quit IRC | 04:32 | |
| *** rmk` has joined #openstack-dev | 04:32 | |
| *** rmk` is now known as rmk | 04:32 | |
| *** rmk has joined #openstack-dev | 04:32 | |
| *** zbitter has joined #openstack-dev | 04:40 | |
| *** zaneb has quit IRC | 04:43 | |
| *** aeperezt has quit IRC | 04:44 | |
| *** dingus9 has quit IRC | 04:45 | |
| *** jcooley_afk is now known as jcooley | 04:46 | |
| *** melwitt has quit IRC | 04:46 | |
| *** nati_ueno_2 has joined #openstack-dev | 05:03 | |
| *** rohitk has joined #openstack-dev | 05:03 | |
| *** nati_ueno has quit IRC | 05:06 | |
| *** nati_ueno_2 has quit IRC | 05:08 | |
| *** Dr_Who has joined #openstack-dev | 05:08 | |
| *** Dr_Who has joined #openstack-dev | 05:08 | |
| *** nati_ueno has joined #openstack-dev | 05:09 | |
| *** rohitk has quit IRC | 05:16 | |
| *** samkottler is now known as samkottler|out | 05:17 | |
| *** nati_ueno has quit IRC | 05:17 | |
| *** rohitk has joined #openstack-dev | 05:17 | |
| *** nati_ueno has joined #openstack-dev | 05:19 | |
| openstackgerrit | A change was merged to openstack/cinder: Use auth_token middleware from keystoneclient. https://review.openstack.org/16959 | 05:21 |
| *** winston-d has quit IRC | 05:21 | |
| *** winston-d has joined #openstack-dev | 05:26 | |
| *** colinmcnamara has joined #openstack-dev | 05:28 | |
| *** colinmcnamara has quit IRC | 05:28 | |
| *** Dr_Who has quit IRC | 05:28 | |
| *** Ryan_Lane has joined #openstack-dev | 05:30 | |
| *** colinmcnamara has joined #openstack-dev | 05:30 | |
| devananda | mikal: ping | 05:33 |
| mikal | devananda: pong | 05:33 |
| *** nati_ueno has quit IRC | 05:34 | |
| *** colinmcnamara1 has joined #openstack-dev | 05:35 | |
| *** colinmcnamara has quit IRC | 05:36 | |
| openstackgerrit | A change was merged to openstack/nova: Migration model update for multi-node resize fix. https://review.openstack.org/17001 | 05:36 |
| *** colinmcnamara1 has quit IRC | 05:40 | |
| *** sacharya has joined #openstack-dev | 05:40 | |
| *** colinmcnamara has joined #openstack-dev | 05:40 | |
| *** dolphm has quit IRC | 05:42 | |
| openstackgerrit | A change was merged to openstack/cinder: Provide i18n to those messages without _() https://review.openstack.org/16644 | 05:44 |
| *** colinmcnamara1 has joined #openstack-dev | 05:46 | |
| *** colinmcnamara has quit IRC | 05:48 | |
| *** Hitesh has joined #openstack-dev | 05:50 | |
| openstackgerrit | A change was merged to openstack/glance: Fixed deleted image being downloadable by admin. https://review.openstack.org/16750 | 05:52 |
| Hitesh | Hello guys, | 05:52 |
| Hitesh | I have stucked...I have to configure quantum without keystone, can anybody tell me the procudure? | 05:53 |
| Hitesh | mnewby: Hi, | 05:54 |
| mnewby | hitesh: hi | 05:54 |
| *** colinmcnamara has joined #openstack-dev | 05:54 | |
| mnewby | set auth_strategy = noauth in quantum.conf | 05:54 |
| Hitesh | mnewby: Have you tried Folsom Quantum without keystone? | 05:54 |
| Hitesh | mnewby: I did that | 05:55 |
| mnewby | Hitesh: you want no auth, or just not to use keystone? | 05:55 |
| Hitesh | mnewby: I want no auth | 05:55 |
| Hitesh | mnewby: how to call it from cli then? | 05:55 |
| Hitesh | mnewby: I have installed quantum-server and client from source and its up | 05:55 |
| mnewby | ah, the cli is validating regardless of the server setup... | 05:56 |
| Hitesh | mnewby: but I am confused about what to pass through cli for os_username | 05:56 |
| *** colinmcnamara1 has quit IRC | 05:56 | |
| Hitesh | mnewby: any suggestion? | 05:56 |
| mnewby | Are you using '--os-auth-strategy noauth' when invoking the cli? | 05:56 |
| Hitesh | mnewby: yes | 05:57 |
| mnewby | ok, so you won't be passing os_username | 05:57 |
| mnewby | you will have to pass --tenant-id, though | 05:57 |
| Hitesh | mnewby: I am passing becuase if I pass --os-auth-strategy, still it ask for os_username | 05:57 |
| mnewby | so that quantum knows who owns the resources you are attempting to create/modify | 05:57 |
| mnewby | huh | 05:57 |
| Hitesh | mnewby: so tenant id could be any thing right? | 05:57 |
| mnewby | hitesh: tenant id will have to correspond to a valid tenant id | 05:58 |
| mnewby | passing an invalid value should at least give a good error message, though | 05:58 |
| Hitesh | mnewby: since we not configure the keystone, then how tenant id could be valide? | 05:58 |
| mnewby | hmmm | 05:59 |
| Hitesh | mnewby: am I missing in config and paste file? | 05:59 |
| mnewby | i haven't used quantum without keystone... | 05:59 |
| *** colinmcnamara has quit IRC | 05:59 | |
| mnewby | Hitesh: just a second, let me try | 05:59 |
| Hitesh | mnewby: cool | 06:00 |
| winston-d | dhellmann, ping | 06:00 |
| Hitesh | mnewby: when I tried this command quantum net-create net1 --os-auth-strategy noauth, its giving me an error like this, (403, 'Forbidden') | 06:01 |
| *** colinmcnamara has joined #openstack-dev | 06:01 | |
| mnewby | Hitesh: can you execute with '-v'? | 06:02 |
| Hitesh | mnewby: ok | 06:02 |
| mnewby | Hitesh: and you've set OS_URL correctly (or are passing --os_url)? | 06:02 |
| Hitesh | mnewby: yes | 06:03 |
| Hitesh | DEBUG: quantumclient.quantum.v2_0.network.CreateNetwork get_data(Namespace(admin_state_down=True, columns=[], formatter='table', name='net1', prefix='', request_format='json', tenant_id=None, value_specs=[], variables=[])) DEBUG: quantumclient.client REQ: curl -i http://0.0.0.0:9696/v2.0/networks.json -X POST -H "Content-Type: application/json" -H "Accept: application/json" -H "User-Agent: python-quantumclient" DEBUG: quantumc | 06:03 |
| *** hemna has quit IRC | 06:03 | |
| mnewby | Hitesh: have you tried running net-list? | 06:03 |
| Hitesh | no | 06:03 |
| Hitesh | mnewby: let me try | 06:03 |
| *** colinmcnamara1 has joined #openstack-dev | 06:04 | |
| Hitesh | mnewby: DEBUG: quantumclient.quantum.v2_0.network.ListNetwork get_data(Namespace(columns=[], fields=[], filter_specs=[], formatter='table', quote_mode='nonnumeric', request_format='json', show_details=False)) DEBUG: quantumclient.quantum.v2_0.network.ListNetwork search options: {} DEBUG: quantumclient.client REQ: curl -i http://0.0.0.0:9696/v2.0/networks.json -X GET -H "Content-Type: application/json" -H "Accept: application/js | 06:04 |
| Hitesh | mnewby: These for net-list | 06:04 |
| mnewby | and does it succeed? | 06:04 |
| Hitesh | mnewby: nops | 06:04 |
| Hitesh | mnewby: same message | 06:05 |
| mnewby | Hitesh: error? | 06:05 |
| Hitesh | mnewby: DEBUG: quantumclient.shell got an error: (403, 'Forbidden') | 06:05 |
| mnewby | Hitesh: 0.0.0.0?? | 06:05 |
| mnewby | Hitesh: I'm not sure that's a valid ip | 06:05 |
| mnewby | Hitesh: You're trying to hit localhost? | 06:06 |
| Hitesh | mnewby: its the url that quantum-server gives, when its run succesfully | 06:06 |
| Hitesh | mnewby: yes | 06:06 |
| Hitesh | mnewby: I am trying to hit localhost | 06:06 |
| mnewby | Hitesh: so use localhost in the url | 06:06 |
| *** colinmcnamara has quit IRC | 06:06 | |
| Hitesh | mnewby: ok | 06:06 |
| mnewby | Hitesh: I'm pretty sure 0.0.0.0 is valid only as a source address | 06:06 |
| *** darjeeling has joined #openstack-dev | 06:06 | |
| Hitesh | mnewby: its giving me same error | 06:07 |
| Hitesh | mnewby: DEBUG: quantumclient.shell got an error: (403, 'Forbidden') | 06:07 |
| Hitesh | mnewby: tell me how to validate the setup has been up for right or wrong? | 06:08 |
| Hitesh | mnewby: I mean I have installed the service through source | 06:08 |
| *** colinmcnamara1 has quit IRC | 06:08 | |
| mnewby | Hitesh: 'installed service from source' isn't enough information for me to help you. | 06:08 |
| mnewby | Hitesh: installed from where? on what? | 06:08 |
| mnewby | Hitesh: and besides that, debugging network services tends to follow the following sequence: | 06:09 |
| Hitesh | mnewby: ok I have done the installation for quantum source, like this, python setup.py install | 06:09 |
| mnewby | Hitesh: is it reachable via telnet on the indicated port? Can a curl command be executed against the serer? And finally, does the command work from the cli? | 06:09 |
| Hitesh | mnewby: oh..let me check through telnet | 06:10 |
| Hitesh | mnewby: how to verify through curl ? any idea? | 06:10 |
| mnewby | Hitesh: I'm curious - why do you want to do this without keystone? | 06:10 |
| mnewby | Hitesh? And why not through devstack? | 06:10 |
| mnewby | Hitesh: If this is your first run at quantum, not having a good default configuration to work from is going to slow you down. | 06:11 |
| Hitesh | mnewby: Actually we have a requirement to test the quantum without keystone | 06:11 |
| mnewby | Hitesh: It's not that what you're trying to do isn't supported, but by deviating from the common way of doing things before you know what you're doing, it's going to be harder to learn. | 06:11 |
| Hitesh | mnewby: I have isntalled it on the other server through devstack | 06:11 |
| Hitesh | mnewby: but currently I can't access that server | 06:12 |
| Hitesh | mnewby: we did customization as well..I mean own plugin | 06:12 |
| mnewby | Hitesh: If that's the case, I would encourage you to get noauth working under devstack, and once you know how to configure things in that environment, try to do it from source. | 06:12 |
| Hitesh | mnewby: Oh..ok | 06:12 |
| mnewby | Hitesh: Otherwise, how do you debug problems? Without a known good configuration, things can be very challenging. | 06:12 |
| Hitesh | mnewby: generally what I use to do, I use to check the user, tenant from keystone, then I use to pass it to cli | 06:13 |
| Hitesh | mnewby: that case it works fine.. | 06:13 |
| Hitesh | mnewby: yes you are right, without good config, its hard to debug | 06:13 |
| *** otherwiseguy has quit IRC | 06:13 | |
| Hitesh | mnewby: let me validate the service setup..through curl and telnet | 06:14 |
| mnewby | Hitesh: it sounds an awful lot like your config is not being picked up | 06:14 |
| Hitesh | mnewby: do you know any configuration that can be done in api.past.ini file? | 06:14 |
| mnewby | Hitesh: Does /etc/quantum/quantum.conf exist, and have auth_strategy = noauth set? | 06:15 |
| Hitesh | mnewby: Yes | 06:15 |
| mnewby | Hitesh: You may want to double check that. 403 indicates an auth failure, which simply shouldn't occur if noauth is configured. | 06:15 |
| Hitesh | mnewby: its showing in debug message as well when we run quantum-server start | 06:15 |
| *** colinmcnamara has joined #openstack-dev | 06:16 | |
| mnewby | Hitesh: and what are you seeing in the quantum server log when the auth failure occurs from the client? | 06:16 |
| Hitesh | mnewby: 2012-11-27 23:53:02 DEBUG [quantum.service] auth_strategy noauth | 06:16 |
| *** nati_ueno has joined #openstack-dev | 06:16 | |
| Hitesh | mnewby: there is no log file under /var/log/quantum | 06:17 |
| mnewby | Hitesh: So where is it logging to? Is it being run as a daemon, or are you invoking manually and having it log to screen? | 06:18 |
| *** colinmcnamara1 has joined #openstack-dev | 06:18 | |
| Hitesh | mnewby: I am seeing it manually on screen | 06:18 |
| mnewby | Hitesh: So what are you seeing when the cli receives a 403? Anything? | 06:19 |
| Hitesh | mnewby: but no message while we hit that url | 06:19 |
| Hitesh | mnewby: no message on log screen | 06:19 |
| mnewby | Hitesh: I'll try to repro, but it sounds an awful lot like you aren't hitting the right service. | 06:19 |
| mnewby | Hitesh: Especially since the service seems to be configured to do DEBUG logging | 06:20 |
| mnewby | Hitesh: Double check with telnet | 06:20 |
| Hitesh | mnewby: yes | 06:20 |
| *** colinmcnamara has quit IRC | 06:20 | |
| Hitesh | mnewby: need to check if that port open or not | 06:20 |
| mnewby | Hitesh: Given that you're seeing a 403, the port is open | 06:20 |
| mnewby | Hitesh: Whether it's the right port is the issue | 06:20 |
| *** colinmcnamara has joined #openstack-dev | 06:21 | |
| *** jeremielim has joined #openstack-dev | 06:22 | |
| mnewby | Hitesh: can you please give me the full url that is failing? | 06:23 |
| Hitesh | mnewby: ok | 06:23 |
| *** lloydde has joined #openstack-dev | 06:23 | |
| *** colinmcnamara1 has quit IRC | 06:23 | |
| mnewby | Hitesh: example: quantum --os-auth-strategy=noaauth --os-url=http://quantum-dev:9696/ net-list | 06:23 |
| mnewby | Hitesh: This works fine | 06:23 |
| Hitesh | mnewby: here is the output of telnet | 06:23 |
| Hitesh | telnet: connect to address ::1: Connection refused | 06:23 |
| mnewby | Hitesh: Not telnet | 06:23 |
| Hitesh | mn | 06:23 |
| Hitesh | mnewby: ok | 06:23 |
| mnewby | Hitesh: Use something like the following: quantum --os-auth-strategy=noaauth --os-url=http://quantum-dev:9696/ net-list | 06:23 |
| *** colinmcnamara1 has joined #openstack-dev | 06:24 | |
| mnewby | Hitesh: Substituting your server name for quantum-dev, of course | 06:24 |
| Hitesh | mnewby: quantum --os-auth-strategy=noaauth --os-url=http://localhost:9696/ net-list | 06:24 |
| mnewby | Hitesh: And? | 06:24 |
| Hitesh | mnewby: same message | 06:25 |
| Hitesh | mnewby: (403, 'Forbidden') | 06:25 |
| mnewby | Hitesh: You're not hitting the write service, then. | 06:25 |
| Hitesh | mnewby: netstat -tulpn I have run this command | 06:25 |
| *** colinmcnamara has quit IRC | 06:25 | |
| Hitesh | mnewby: output is tcp 0 0 0.0.0.0:9696 0.0.0.0:* LISTEN 940/python | 06:25 |
| Hitesh | mnewby: also I ran this | 06:26 |
| Hitesh | mnewby: telnet localhost 9696 | 06:26 |
| Hitesh | mnewby: output is , Trying ::1... telnet: connect to address ::1: Connection refused Trying 127.0.0.1... Connected to localhost. | 06:26 |
| *** colinmcnamara has joined #openstack-dev | 06:26 | |
| Hitesh | mnewby: can you give me telnet command that you use to check? | 06:26 |
| mnewby | Hitesh: What is the bind_host value in quantum.conf? | 06:27 |
| Hitesh | bind_host=0.0.0.0 | 06:27 |
| mnewby | Hitesh: well, change it to something sane. | 06:27 |
| *** colinmcnamara has left #openstack-dev | 06:27 | |
| Hitesh | mnewby: do you want me to change it? | 06:27 |
| Hitesh | mnewby: machine ip address? | 06:27 |
| mnewby | Hitesh: Like, say, localhost. machine ip. whatever. | 06:27 |
| Hitesh | mnewby: ok | 06:27 |
| *** jeremielim_ has joined #openstack-dev | 06:28 | |
| *** jeremielim has quit IRC | 06:28 | |
| *** jeremielim_ is now known as jeremielim | 06:28 | |
| *** colinmcnamara1 has quit IRC | 06:28 | |
| Hitesh | mnewby: I have changed it to machine ip, still getting same message | 06:29 |
| Hitesh | mnewby: give me telnet command that you use to try for port? | 06:30 |
| mnewby | Hitesh: I don't know what to tell you. If you've configured quantum with noauth, and executed the command as I described, you should not be getting a 403. | 06:30 |
| mnewby | Hitesh: If you are, you're hitting the wrong service, somehow. | 06:30 |
| mnewby | Hitesh: And as to how, well, that's not something I can help you debug. | 06:30 |
| mnewby | Hitesh: The particulars of the box in question aren't amenable to debugging by proxy. | 06:31 |
| *** rushiagr has joined #openstack-dev | 06:31 | |
| Hitesh | mnewby: oh.. | 06:31 |
| mnewby | Hitesh: Make sure no other service is running and that nothing is preventing you from hitting the service (redirection via firewall rules, for instance). | 06:31 |
| Hitesh | mnewby: Ok..it seems port has been open | 06:31 |
| Hitesh | mnewby: ok. | 06:32 |
| Hitesh | mnewby: my VM is on DC..I am connecting that VM thrugh VPN | 06:32 |
| Hitesh | mnewby: will that be a problem? | 06:32 |
| mnewby | Hitesh: No idea. | 06:32 |
| Hitesh | mnewby: Ok. | 06:33 |
| mnewby | Hitesh: sorry. | 06:33 |
| Hitesh | mnewby: when I tried at my home lab through devstack | 06:33 |
| Hitesh | mnewby: its working fine..also debug messages appreare on screen when we hit the url | 06:33 |
| Hitesh | mnewby: let check code though | 06:34 |
| mnewby | Hitesh: every indication is that the cli invocation is not hitting the right service. no log message, an unexpected 403… figure out how to hit the right service, and things will work as expected. | 06:34 |
| Hitesh | mnewby: exactly | 06:35 |
| Hitesh | mnewby: let me hit that url through curl | 06:35 |
| mnewby | Hitesh: are you invoking quantum on the same box as is running the service? | 06:35 |
| *** lloydde has quit IRC | 06:35 | |
| *** lloydde has joined #openstack-dev | 06:36 | |
| Hitesh | mnewby: yes | 06:36 |
| Hitesh | mnewby: I am on the same box | 06:36 |
| mnewby | iptables -L ? | 06:36 |
| Hitesh | mnewby: quantum-server and cli are on the same box | 06:37 |
| *** colinmcnamara1 has joined #openstack-dev | 06:37 | |
| Hitesh | Chain INPUT (policy ACCEPT) target prot opt source destination ACCEPT all -- anywhere anywhere state RELATED,ESTABLISHED ACCEPT icmp -- anywhere anywhere ACCEPT all -- anywhere anywhere ACCEPT tcp -- anywhere anywhere state NEW tcp dpt:ssh REJECT all -- anywhere anywhere reject-with icmp-ho | 06:37 |
| mnewby | Hitesh: Please don't paste so much - clutters it up for everyone. I was just prompting you to check whether there were any relevant iptables rules that would be blocking or redirecting packets. | 06:40 |
| *** lloydde has quit IRC | 06:41 | |
| *** colinmcnamara has joined #openstack-dev | 06:41 | |
| *** colinmcnamara1 has quit IRC | 06:41 | |
| *** eglynn_ has quit IRC | 06:41 | |
| Hitesh | mnewby: Ok..I will take care of it :) .. do you feel any config in policy.json? I think it should not invoke if we configure noauth | 06:41 |
| mnewby | Hitesh: No. Your first task is to get log output on the server in response to a cli call. | 06:42 |
| Hitesh | mnewby: Ok.. | 06:43 |
| *** colinmcnamara has quit IRC | 06:46 | |
| *** jcooley is now known as jcooley_afk | 06:46 | |
| *** yjiang5_away is now known as yjiang5 | 06:49 | |
| *** colinmcnamara has joined #openstack-dev | 06:49 | |
| *** maxiz has quit IRC | 06:52 | |
| *** jcooley_afk is now known as jcooley | 06:52 | |
| Hitesh | mnewby: give me your emai lid? I will send the respective efforts for this on your id | 06:52 |
| mnewby | Hitesh: I'm not sure I see the point. This isn't a quantum issue. | 06:53 |
| Hitesh | mnewby: ok | 06:53 |
| *** maxiz has joined #openstack-dev | 06:53 | |
| *** jeremielim has quit IRC | 06:54 | |
| Hitesh | mnewby: You are right, since its not hitting the service, it must be configuration problem.. | 06:54 |
| *** colinmcnamara has quit IRC | 06:54 | |
| Hitesh | mnewby: some where its blocking | 06:54 |
| *** jeremielim has joined #openstack-dev | 06:57 | |
| Hitesh | mnewby: Thanks mnewby for help :) | 06:58 |
| Hitesh | mnewby: Really appreciated | 06:58 |
| mnewby | Hitesh: Not sure I did much of anything but I hope you figure it out. :) | 06:58 |
| *** adjohn has quit IRC | 06:58 | |
| Hitesh | mnewby: Yes..I will ask internal team about it....from Grizzly perspective, what are you looking at currently | 06:59 |
| *** jcooley is now known as jcooley_afk | 06:59 | |
| mnewby | Hitesh: I haven't been contributing much to core lately. Some back-burnered efforts to get quantum ovs working on xcp/xs, working with xcp on devstack, and testable via tempest. Slow going, though, since xcp isn't a platform most quantum devs use. | 07:00 |
| Hitesh | mnewby: oh..sounds intersting | 07:01 |
| Hitesh | mnewby: I will start contributing from next week | 07:01 |
| mnewby | Hitesh: Always appreciated :) | 07:01 |
| *** Tross has quit IRC | 07:02 | |
| Hitesh | mnewby: :) | 07:02 |
| *** Mandell has quit IRC | 07:03 | |
| *** nati_ueno has quit IRC | 07:04 | |
| *** garyk has joined #openstack-dev | 07:08 | |
| *** markwash has quit IRC | 07:13 | |
| *** nati_ueno has joined #openstack-dev | 07:16 | |
| *** mindpixel has joined #openstack-dev | 07:18 | |
| mnewby | garyk: question regardling cli... | 07:19 |
| *** jcmartin has quit IRC | 07:19 | |
| *** zhiteng has joined #openstack-dev | 07:21 | |
| *** zhiteng has quit IRC | 07:22 | |
| *** sacharya has quit IRC | 07:23 | |
| mnewby | garyk: I can't seem to figure out how to create a subnet with multiple allocation pools. And I notice that, despite docs suggesting allocation_pools should be posted as a list of dicts in a json request, it is actually being passed as a raw string, which would seem to preclude multiple pools being provided. | 07:23 |
| garyk | mnewby: give me a sec | 07:24 |
| mnewby | garyk: no rush :) | 07:24 |
| garyk | mnewby: i have a cheat sheet somewhere that i need to dig up... | 07:24 |
| *** mrunge has joined #openstack-dev | 07:25 | |
| *** jeremielim has quit IRC | 07:27 | |
| *** jeremielim has joined #openstack-dev | 07:27 | |
| garyk | mnewby: quantum subnet-create --allocation-pool start=13.0.0.10,end=13.0.0.20 --allocation-pool start=13.0.0.40,end=13.0.0.60 private 13.0.0.0/24 | 07:31 |
| mnewby | garyk: ah, so it requires providing the options in two different ways? | 07:31 |
| garyk | mnewby: i guess. originally i thought that it needed a "list=true type=dict" but this did not work (that is how i configured the fixed ips). | 07:32 |
| *** eglynn has joined #openstack-dev | 07:32 | |
| garyk | mnewby: i think that gongys is the best guys to ask about the reasoning here | 07:32 |
| *** Mandell has joined #openstack-dev | 07:34 | |
| mnewby | garyk: the ability to specify things in multiple ways confused me, thanks for the working example :) | 07:36 |
| *** darjeeling has quit IRC | 07:42 | |
| *** ondergetekende has joined #openstack-dev | 07:42 | |
| *** suraj_ has joined #openstack-dev | 07:46 | |
| suraj_ | hi all | 07:46 |
| *** maelfius has quit IRC | 07:46 | |
| suraj_ | I want to know abt VMware ESX support in openstack? | 07:46 |
| *** ondergetekende has quit IRC | 07:47 | |
| *** ondergetekende has joined #openstack-dev | 07:47 | |
| *** ondergetekende has joined #openstack-dev | 07:48 | |
| *** darjeeling has joined #openstack-dev | 07:50 | |
| *** flaper87 has joined #openstack-dev | 07:55 | |
| *** flaper87 has joined #openstack-dev | 07:55 | |
| *** flaper87 has quit IRC | 07:55 | |
| *** flaper87 has joined #openstack-dev | 07:56 | |
| *** flaper87 has joined #openstack-dev | 07:56 | |
| *** rafaduran has joined #openstack-dev | 08:03 | |
| *** nati_ueno has quit IRC | 08:05 | |
| *** nati_ueno has joined #openstack-dev | 08:06 | |
| *** nati_ueno has quit IRC | 08:10 | |
| *** Tross has joined #openstack-dev | 08:14 | |
| *** Tross has left #openstack-dev | 08:14 | |
| *** maploin has joined #openstack-dev | 08:17 | |
| *** maploin has quit IRC | 08:17 | |
| *** maploin has joined #openstack-dev | 08:17 | |
| *** jpich has joined #openstack-dev | 08:18 | |
| *** Tross1 has joined #openstack-dev | 08:18 | |
| *** markwash has joined #openstack-dev | 08:19 | |
| *** Tross has joined #openstack-dev | 08:22 | |
| *** Tross1 has quit IRC | 08:22 | |
| *** Tross has left #openstack-dev | 08:23 | |
| *** markwash has quit IRC | 08:24 | |
| *** afazekas has joined #openstack-dev | 08:25 | |
| *** Tross1 has joined #openstack-dev | 08:25 | |
| *** eglynn_ has joined #openstack-dev | 08:25 | |
| *** reidrac has joined #openstack-dev | 08:27 | |
| *** eglynn has quit IRC | 08:28 | |
| *** maelfius has joined #openstack-dev | 08:34 | |
| *** vishy is now known as vishy_zz | 08:36 | |
| *** winston-d_ has quit IRC | 08:39 | |
| *** EmilienM has quit IRC | 08:40 | |
| *** salv-orlando has joined #openstack-dev | 08:48 | |
| *** mrodden has quit IRC | 08:52 | |
| *** ondergetekende has quit IRC | 08:53 | |
| *** zing has joined #openstack-dev | 08:55 | |
| *** winston-d has quit IRC | 08:56 | |
| *** jrodom has joined #openstack-dev | 08:57 | |
| *** Tross1 has quit IRC | 08:58 | |
| *** woorea has joined #openstack-dev | 08:58 | |
| *** Tross has joined #openstack-dev | 09:00 | |
| *** shardy_afk is now known as shardy | 09:06 | |
| *** vipul has quit IRC | 09:08 | |
| *** vipul has joined #openstack-dev | 09:08 | |
| *** epim has quit IRC | 09:09 | |
| *** kpavel has quit IRC | 09:11 | |
| *** derekh has joined #openstack-dev | 09:13 | |
| *** darraghb has joined #openstack-dev | 09:24 | |
| *** danpb has joined #openstack-dev | 09:28 | |
| *** AnilV4 has quit IRC | 09:28 | |
| *** maelfius has quit IRC | 09:28 | |
| *** kpavel has joined #openstack-dev | 09:29 | |
| *** shang has quit IRC | 09:29 | |
| *** shang has joined #openstack-dev | 09:30 | |
| *** AnilV4 has joined #openstack-dev | 09:30 | |
| *** pixelbeat has joined #openstack-dev | 09:31 | |
| *** EmilienM has joined #openstack-dev | 09:39 | |
| *** aloga has quit IRC | 09:40 | |
| *** aloga has joined #openstack-dev | 09:41 | |
| *** ondergetekende has joined #openstack-dev | 09:42 | |
| *** markmc has joined #openstack-dev | 09:42 | |
| *** huats_ has quit IRC | 09:43 | |
| *** maxiz has quit IRC | 09:43 | |
| *** arbrandes has joined #openstack-dev | 09:45 | |
| *** eglynn__ has joined #openstack-dev | 09:46 | |
| *** eglynn_ has quit IRC | 09:49 | |
| *** salv-orlando has quit IRC | 09:49 | |
| *** salv-orlando_ has joined #openstack-dev | 09:49 | |
| *** reed has joined #openstack-dev | 09:51 | |
| *** woorea has quit IRC | 09:51 | |
| *** EmilienM has quit IRC | 09:53 | |
| *** EmilienM has joined #openstack-dev | 09:55 | |
| *** suraj_ has quit IRC | 09:56 | |
| *** woorea has joined #openstack-dev | 09:57 | |
| *** salv-orlando_ has quit IRC | 09:59 | |
| *** SuYan has quit IRC | 09:59 | |
| *** jeremielim has quit IRC | 10:02 | |
| *** xchu has quit IRC | 10:10 | |
| *** alobbs has joined #openstack-dev | 10:13 | |
| *** gongysh has joined #openstack-dev | 10:19 | |
| *** ewindisch has joined #openstack-dev | 10:23 | |
| *** adalbas has joined #openstack-dev | 10:24 | |
| *** woorea has quit IRC | 10:32 | |
| garyk | markmc: ping | 10:39 |
| markmc | garyk, yes? | 10:39 |
| garyk | markmc: regarding the quantum stable there is one outstanding patch - https://review.openstack.org/#/c/17033/. | 10:39 |
| markmc | garyk, yes | 10:39 |
| garyk | markmc: i have tested this with the stable env and it works. | 10:39 |
| garyk | markmc: once this is in quantum is good to go | 10:40 |
| markmc | garyk, ok, add that as a comment to the review | 10:40 |
| garyk | markmc: sure. np. | 10:40 |
| markmc | garyk, I don't see much in the review, commit message or bug to help folks decide whether it's vital for 2012.2.1 or not | 10:41 |
| garyk | markmc: without it dhcp will not work when namespaces is disabled. i can add this to the commit message if necessary | 10:41 |
| markmc | garyk, add as a review comment | 10:42 |
| markmc | garyk, and it's not about "what the bug is" but "how serious is that bug, who does it impact, etc." | 10:42 |
| *** woorea has joined #openstack-dev | 10:42 | |
| garyk | markmc: ok. when it was opened it should have been set as critical or a blocker | 10:44 |
| markmc | garyk, ok, change it then | 10:45 |
| *** mmagr has joined #openstack-dev | 10:48 | |
| *** maploin has quit IRC | 10:49 | |
| *** maurosr has joined #openstack-dev | 10:53 | |
| *** arbrandes has quit IRC | 11:01 | |
| *** mindpixel has quit IRC | 11:01 | |
| *** yjiang5 is now known as yjiang5_away | 11:05 | |
| *** woorea has quit IRC | 11:06 | |
| *** jrodom has quit IRC | 11:08 | |
| *** salv-orlando has joined #openstack-dev | 11:11 | |
| *** alexpilotti has joined #openstack-dev | 11:15 | |
| *** maxiz has joined #openstack-dev | 11:15 | |
| *** kmartin has quit IRC | 11:16 | |
| *** jeremielim has joined #openstack-dev | 11:19 | |
| *** jeremielim has quit IRC | 11:21 | |
| *** mrunge has quit IRC | 11:35 | |
| *** shardy is now known as shardy_afk | 11:38 | |
| garyk | salv-orlando: hi, i saw your comments on the l3. i did things a little different. thanks! | 11:39 |
| salv-orlando | garyk: which comments are you referring to? I am commenting on a lot of stuff recently. | 11:41 |
| salv-orlando | You mean the l3 agent rpc? | 11:42 |
| salv-orlando | ok, I've seen your update on gerrit | 11:42 |
| garyk | salv-orlando: ack | 11:44 |
| *** roz has joined #openstack-dev | 11:52 | |
| *** roadTripper has joined #openstack-dev | 11:56 | |
| *** sthaha has joined #openstack-dev | 11:57 | |
| *** darjeeling has quit IRC | 12:02 | |
| *** maploin has joined #openstack-dev | 12:05 | |
| *** maploin has quit IRC | 12:05 | |
| *** maploin has joined #openstack-dev | 12:05 | |
| *** mrunge has joined #openstack-dev | 12:07 | |
| *** Tross has quit IRC | 12:12 | |
| *** mrunge has quit IRC | 12:20 | |
| *** Mandell has quit IRC | 12:23 | |
| *** timello has joined #openstack-dev | 12:29 | |
| *** dhellmann is now known as dhellmann-afk | 12:32 | |
| *** EmilienM has quit IRC | 12:33 | |
| *** EmilienM has joined #openstack-dev | 12:34 | |
| *** adalbas has quit IRC | 12:35 | |
| *** sandywalsh has joined #openstack-dev | 12:47 | |
| *** shardy_afk is now known as shardy | 12:49 | |
| *** almaisan-away is now known as al-maisan | 12:49 | |
| *** 17SAB5KKR is now known as ayoung | 12:54 | |
| *** markvoelker has joined #openstack-dev | 12:55 | |
| *** darjeeling has joined #openstack-dev | 12:57 | |
| *** rushiagr has quit IRC | 13:03 | |
| *** boden has joined #openstack-dev | 13:05 | |
| *** Elwell_ has joined #openstack-dev | 13:09 | |
| *** Elwell has quit IRC | 13:09 | |
| *** dprince has joined #openstack-dev | 13:11 | |
| *** jpich has quit IRC | 13:12 | |
| *** jpich has joined #openstack-dev | 13:12 | |
| rohitk | hi folks: Anybody tried installing Quantum with the latest Devstack? | 13:14 |
| rohitk | Quantum service fails to start with CRITICAL quantum [-] [Errno 13] Permission denied: '/var/lib/quantum' | 13:14 |
| *** adalbas has joined #openstack-dev | 13:14 | |
| *** tmishler has joined #openstack-dev | 13:15 | |
| *** dolphm has joined #openstack-dev | 13:16 | |
| *** jrodom has joined #openstack-dev | 13:18 | |
| amotoki | rohitk: please apply https://review.openstack.org/17042. This will fix your problem. | 13:18 |
| rohitk | amotoki: Thanks, will try that | 13:20 |
| openstackgerrit | A change was merged to openstack/quantum: Fixes update router gateway successful with existed floatingip association https://review.openstack.org/16880 | 13:20 |
| *** alunduil has quit IRC | 13:23 | |
| rohitk | amotoki: The patch worked | 13:24 |
| amotoki | rohitk: good to hear that | 13:24 |
| *** samkottler|out is now known as samkottler | 13:28 | |
| *** colinmcnamara has joined #openstack-dev | 13:29 | |
| *** dingus9 has joined #openstack-dev | 13:30 | |
| afazekas | jaypipes: ping | 13:32 |
| soren | markmc: Do you have a ball park estimate of how many changes have landed in all the stable/folsom branches so far? | 13:36 |
| soren | markmc: Total, I mean. | 13:36 |
| soren | markmc: Oh, you mentioned that yesterday in the meeting, didn't you? | 13:37 |
| *** dingus9 has quit IRC | 13:38 | |
| *** dingus9 has joined #openstack-dev | 13:39 | |
| trapni | is it possible to run folsom horizon UI against an Essex Cloud ? | 13:41 |
| *** p4tux has quit IRC | 13:42 | |
| *** nati_ueno has joined #openstack-dev | 13:45 | |
| *** dprince_ has joined #openstack-dev | 13:48 | |
| *** dprince has quit IRC | 13:48 | |
| *** nati_ueno has quit IRC | 13:48 | |
| markmc | soren, http://lists.openstack.org/pipermail/openstack-stable-maint/2012-November/000131.html | 13:49 |
| *** rushiagr has joined #openstack-dev | 13:49 | |
| soren | markmc: Wicked, thanks. | 13:50 |
| *** rushiagr has left #openstack-dev | 13:50 | |
| *** pabelanger has joined #openstack-dev | 13:52 | |
| *** e1mer has quit IRC | 13:52 | |
| *** sayjay has quit IRC | 14:04 | |
| dims | clarkb, markmc, any word on problem with gate-oslo-incubator-pyflakes? | 14:09 |
| *** imsplitbit has joined #openstack-dev | 14:10 | |
| *** salv-orlando has quit IRC | 14:10 | |
| *** salv-orlando has joined #openstack-dev | 14:10 | |
| markmc | dims, this needs to land: https://review.openstack.org/16671 | 14:10 |
| dims | thanks markmc | 14:15 |
| *** Dr_Who has joined #openstack-dev | 14:16 | |
| *** Dr_Who has joined #openstack-dev | 14:16 | |
| *** al-maisan is now known as almaisan-away | 14:18 | |
| dims | sdague, how do i recreate this problem? https://bugs.launchpad.net/oslo/+bug/1083218 | 14:19 |
| uvirtbot | Launchpad bug 1083218 in oslo "oslo changes to logging_context_format_string not compatible with nova" [Critical,Triaged] | 14:19 |
| *** ahardy_ has joined #openstack-dev | 14:19 | |
| afazekas | jaypipes: ping | 14:19 |
| *** trapni has quit IRC | 14:22 | |
| jaypipes | afazekas: pong, but back in 20.. need to reboot. | 14:24 |
| *** jaypipes has quit IRC | 14:25 | |
| *** imsplitbit has quit IRC | 14:27 | |
| *** imsplitbit has joined #openstack-dev | 14:28 | |
| *** alunduil has joined #openstack-dev | 14:29 | |
| *** kbringard has joined #openstack-dev | 14:31 | |
| *** alunduil has quit IRC | 14:31 | |
| *** alunduil has joined #openstack-dev | 14:32 | |
| garyk | russellb: from what it seems the nitification queue has entries when then service starts. hence http://paste.openstack.org/show/26941/ | 14:32 |
| *** trapni has joined #openstack-dev | 14:33 | |
| *** gongysh has quit IRC | 14:33 | |
| *** trapni has quit IRC | 14:34 | |
| *** trapni has joined #openstack-dev | 14:34 | |
| *** radez_g0n3 is now known as radez | 14:34 | |
| *** Dr_Who has quit IRC | 14:35 | |
| russellb | garyk: looking around at code ... | 14:36 |
| *** amotoki has quit IRC | 14:36 | |
| garyk | russellb: if i clear the queues prior to running the quantum service then all is good. it just cropped up today (from i see we did not update any common code). | 14:37 |
| russellb | so that's interesting, i wonder why there's anything in the queue | 14:38 |
| russellb | it looks like whatever set up the queue to send something to it didn't set the same parameters as this code trying to consume from it | 14:38 |
| *** imsplitbit_ has joined #openstack-dev | 14:40 | |
| *** zhiteng has joined #openstack-dev | 14:40 | |
| *** imsplitbit_ has quit IRC | 14:41 | |
| *** mrodden has joined #openstack-dev | 14:41 | |
| *** imsplitbit_ has joined #openstack-dev | 14:43 | |
| *** imsplitbit has quit IRC | 14:43 | |
| russellb | garyk: actually, there shouldn't be anything in the queue unless something created it, and only the consumer side (this code) should be creating it | 14:43 |
| *** jcooley_afk is now known as jcooley | 14:44 | |
| *** maploin has quit IRC | 14:45 | |
| russellb | garyk: not sure what to tell you ... i'd dig into finding out what is in the queue when you start up, that will probably give a hint | 14:45 |
| garyk | russellb: thanks. that is what i am doing. just was crossing my fingers that you could have had a ace up your sleeve | 14:47 |
| *** imsplitbit_ is now known as imsplitbit | 14:47 | |
| russellb | nope sorry :) | 14:47 |
| *** maploin has joined #openstack-dev | 14:48 | |
| *** aloga has quit IRC | 14:49 | |
| *** maploin has quit IRC | 14:49 | |
| *** maploin has joined #openstack-dev | 14:49 | |
| *** aloga has joined #openstack-dev | 14:49 | |
| *** zhiteng has quit IRC | 14:49 | |
| *** colinmcnamara has quit IRC | 14:51 | |
| *** aeperezt has joined #openstack-dev | 14:51 | |
| *** rohitk has quit IRC | 14:51 | |
| *** winston-d has joined #openstack-dev | 14:52 | |
| *** jchiles has joined #openstack-dev | 14:52 | |
| *** jaustinpage|want has joined #openstack-dev | 14:52 | |
| *** Dr_Who has joined #openstack-dev | 14:53 | |
| *** Dr_Who has joined #openstack-dev | 14:53 | |
| *** nunosantos has joined #openstack-dev | 14:56 | |
| *** flaper87_ has joined #openstack-dev | 14:58 | |
| *** ahardy_ has quit IRC | 14:58 | |
| *** jaypipes has joined #openstack-dev | 14:59 | |
| *** flaper87_ has quit IRC | 14:59 | |
| *** Dr_Who has quit IRC | 15:00 | |
| *** Hitesh has quit IRC | 15:01 | |
| soren | markmc: Just for giggles, do you have the same number for stable/essex? | 15:01 |
| afazekas | jaypipes: I would like to add http://docs.openstack.org/developer/swift/overview_object_versioning.html - related test case(s) to tempest. Is the allow_versions = true now ? | 15:01 |
| markmc | soren, http://wiki.openstack.org/ReleaseNotes/2012.1.{1,2,3} | 15:01 |
| jaypipes | afazekas: looking... | 15:02 |
| *** flaper87 has quit IRC | 15:02 | |
| * soren hugs markmc | 15:02 | |
| jaypipes | afazekas: so, are you asking about the Swift setting in the devstack gate? If so, I actually have no idea... | 15:03 |
| markmc | soren, np :) | 15:03 |
| jaypipes | afazekas: I'm not even sure we have swift enabled in the devstack gate, to be honest. | 15:03 |
| jaypipes | afazekas: gimme some time to check it out... going to grab some coffee and I will ping you about this later. | 15:03 |
| *** colinmcnamara has joined #openstack-dev | 15:03 | |
| afazekas | jaypipes: thank you | 15:04 |
| *** sthaha has quit IRC | 15:04 | |
| *** markmcclain has joined #openstack-dev | 15:04 | |
| *** TheEstablishment has joined #openstack-dev | 15:05 | |
| *** kpavel has quit IRC | 15:06 | |
| *** jrodom has quit IRC | 15:06 | |
| *** edmund has joined #openstack-dev | 15:09 | |
| *** ndipanov has quit IRC | 15:09 | |
| *** eharney has joined #openstack-dev | 15:09 | |
| *** andrewbogott_afk is now known as andrewbogott | 15:10 | |
| *** rpedde_away is now known as rpedde | 15:10 | |
| *** edmund1 has joined #openstack-dev | 15:12 | |
| *** ndipanov has joined #openstack-dev | 15:12 | |
| *** sacharya has joined #openstack-dev | 15:13 | |
| *** edmund has quit IRC | 15:13 | |
| *** koolhead17 has joined #openstack-dev | 15:14 | |
| *** colinmcnamara has quit IRC | 15:17 | |
| *** jcru has joined #openstack-dev | 15:18 | |
| *** salv-orlando has quit IRC | 15:18 | |
| *** salv-orlando has joined #openstack-dev | 15:19 | |
| *** ondergetekende has quit IRC | 15:19 | |
| *** sacharya has quit IRC | 15:19 | |
| *** colinmcnamara has joined #openstack-dev | 15:19 | |
| *** eglynn__ has quit IRC | 15:19 | |
| *** TerryH has joined #openstack-dev | 15:20 | |
| davidkranz | afazekas, jaypipes : I believe https://review.openstack.org/#/c/14689/ (EC2 stuff) has caused a bunch of error messages in the logs but no test failure. | 15:22 |
| davidkranz | See http://logs.openstack.org/14689/18/gate/gate-tempest-devstack-vm/20140/console.html.gz | 15:22 |
| davidkranz | We need to get rid of those messages if there is not actually anything wrong. | 15:22 |
| *** eglynn__ has joined #openstack-dev | 15:22 | |
| *** salv-orlando_ has joined #openstack-dev | 15:23 | |
| *** salv-orlando has quit IRC | 15:23 | |
| *** salv-orlando_ is now known as salv-orlando | 15:23 | |
| *** kpavel has joined #openstack-dev | 15:24 | |
| afazekas | davidkranz: boto library need to be upgraded | 15:24 |
| *** glauaguiar has joined #openstack-dev | 15:24 | |
| davidkranz | afazekas: Can you do that or is ci team needed? | 15:24 |
| afazekas | davidkranz: I do not have access | 15:25 |
| *** Dr_Who has joined #openstack-dev | 15:25 | |
| davidkranz | afazekas: Can you ping fungi or clarkb and explain what is needed? | 15:26 |
| *** colinmcnamara has quit IRC | 15:26 | |
| openstackgerrit | A change was merged to openstack/nova: Make resize and multi-node work properly together https://review.openstack.org/17002 | 15:27 |
| *** benner has quit IRC | 15:28 | |
| *** anteaya has joined #openstack-dev | 15:28 | |
| *** benner has joined #openstack-dev | 15:30 | |
| *** mtreinish has joined #openstack-dev | 15:32 | |
| *** winston-d has quit IRC | 15:32 | |
| afazekas | davidkranz: tools/pip-requires says: boto>=2.2.1 is required. I can provide shell script part for manifest preparation if it missing too | 15:34 |
| fungi | afazekas: davidkranz: i'll look into what's involved in increasing the minimum version for a dependency | 15:36 |
| *** annegentle_ has joined #openstack-dev | 15:36 | |
| *** timello has quit IRC | 15:37 | |
| afazekas | fungi: I am ok, with much newer version like, boto-2.5.2 too | 15:38 |
| *** timello has joined #openstack-dev | 15:38 | |
| fungi | afazekas: if it's used by multiple subprojects, that can get tricky i think (depending on how much backward compatability the library maintains) | 15:39 |
| fungi | but i'll find out | 15:39 |
| *** maxiz has quit IRC | 15:39 | |
| *** datsun180b has joined #openstack-dev | 15:43 | |
| *** timello has quit IRC | 15:43 | |
| *** timello has joined #openstack-dev | 15:43 | |
| *** timello has quit IRC | 15:44 | |
| *** eglynn__ has quit IRC | 15:46 | |
| *** eglynn__ has joined #openstack-dev | 15:46 | |
| *** winston-d has joined #openstack-dev | 15:47 | |
| *** timello has joined #openstack-dev | 15:48 | |
| jaypipes | davidkranz: k, looking into it. | 15:49 |
| *** woodspa has joined #openstack-dev | 15:49 | |
| davidkranz | jaypipes: I think fungi is on it now. ^^^^ | 15:50 |
| jaypipes | davidkranz: ya. | 15:50 |
| jaypipes | davidkranz: when did we start filling up the log with ERRORs that aren't errors? :( | 15:50 |
| jaypipes | davidkranz: IMHO, a 404 is not an ERROR log level... | 15:51 |
| openstackgerrit | A change was merged to openstack/keystone: Ensure token expiration is maintained (bug 1079216) https://review.openstack.org/17051 | 15:51 |
| uvirtbot | Launchpad bug 1079216 in keystone "token expires time incorrect for auth by one token" [High,In progress] https://launchpad.net/bugs/1079216 | 15:51 |
| fungi | davidkranz: jaypipes: fungi isn't entirely sure what he's doing and is attempting to muddle through. found references to boto in devstack files/{apts,rpms}/* | 15:51 |
| *** imsplitbit has quit IRC | 15:52 | |
| *** maxiz has joined #openstack-dev | 15:53 | |
| jaypipes | fungi: yes, the problem is that the euca.sh exercise in devstack may depend on a different version of boto than the tempest ec2 tests do... it may be that for the tempest runs specifically (in devstack-gate project), we install the updated 2.5.2 lib. | 15:53 |
| fungi | doesn't look like we're doing versioning of any of those though, since that's distro packages | 15:53 |
| fungi | jaypipes: ahh, okay | 15:53 |
| *** dhellmann-afk is now known as dhellmann | 15:53 | |
| jaypipes | fungi: right, and not pip installs, which tempest relies on, IIRC | 15:53 |
| jaypipes | fungi: it might be as simple as putting a call to pip install -r requirements.txt in the tempest dir from the devstack-gate-tempest-vm.sh script? | 15:54 |
| fungi | jaypipes: maybe. i'm looking in there now | 15:54 |
| jaypipes | fungi: coolio, thx for your help jeremy | 15:55 |
| fungi | jaypipes: i take it a versioned entry for it in tools/test-requires wouldn't be appropriate? | 15:55 |
| jaypipes | fungi: sure, that's totally fine. | 15:56 |
| *** ahardy_ has joined #openstack-dev | 15:57 | |
| fungi | jaypipes: okay, i'll try to skin it that way then | 15:57 |
| * jaypipes thought there was something in there for boto already...? | 15:57 | |
| jaypipes | kk | 15:57 |
| fungi | no references to the string "boto" anywhere in the devstack-gate repo (yet) | 15:57 |
| jaypipes | fungi: no, it's in tempest.. | 15:57 |
| fungi | oh! i'll pull that one and have a look then | 15:58 |
| jaypipes | fungi: and since devstack-gate installs tempest, should be easy enough :) | 15:58 |
| *** maploin has quit IRC | 15:58 | |
| fungi | yep | 15:58 |
| *** kmartin has joined #openstack-dev | 15:59 | |
| fungi | ahh, yeah, updating my tempest tree i see a bunch of boto pieces just got pulled | 15:59 |
| fungi | jaypipes: right, so tempest already has tools/pip-requires:boto>=2.2.1 | 16:00 |
| fungi | which appears to be what afazekas was already saying | 16:00 |
| jaypipes | fungi: and what I was saying too :) | 16:01 |
| *** avishay has joined #openstack-dev | 16:01 | |
| *** andrewbogott is now known as andrewbogott_afk | 16:02 | |
| *** p4tux has joined #openstack-dev | 16:02 | |
| alexpilotti | dansmith: ping | 16:02 |
| jaypipes | fungi: I believe we just need to call pip install -r tools/test-requires right before the call to do nosetests -sv blah blah | 16:02 |
| winston-d | dhellmann, ping | 16:02 |
| jaypipes | winston-d: hey man, long time to chat :) how are ya? | 16:02 |
| fungi | jaypipes: ah! now it makes sense. as opposed to later | 16:02 |
| *** mmagr has quit IRC | 16:03 | |
| fungi | i'll work up a patch in that case | 16:03 |
| avishay | winston-d: thanks for your help with CHAP. I was under the impression that there was support for two-way CHAP, and that's why I was confused by the code. | 16:03 |
| winston-d | jaypipes, hey, jay. I'm fine, thx. how about you? you seems to have been quiet for a while. :) | 16:03 |
| dansmith | alexpilotti: be with you in a sec | 16:03 |
| *** Dr_Who has quit IRC | 16:03 | |
| jaypipes | winston-d: yes, indeed. moved to AT&T and been heads down building deployment automation stuff. fun :) | 16:04 |
| *** Elwell_ is now known as Elwell | 16:04 | |
| winston-d | avishay, ur welcome. | 16:04 |
| alexpilotti | dansmith: ok, tx! | 16:04 |
| *** reidrac has quit IRC | 16:05 | |
| winston-d | jaypipes, cool! is att providing public cloud service? | 16:05 |
| *** tmishler has quit IRC | 16:05 | |
| *** sacharya has joined #openstack-dev | 16:06 | |
| *** tmishler has joined #openstack-dev | 16:06 | |
| jaypipes | winston-d: nope, internal customers only | 16:06 |
| jaypipes | winston-d: which has its upside and downside :) upside is no billing, downside is everyone just grabs m2.xlarges ;) | 16:07 |
| winston-d | jaypipes, private cloud, even better! i love enterprise workloads. :) | 16:07 |
| jaypipes | winston-d: lol, you would be the first. :P | 16:07 |
| winston-d | jaypipes, r u still in texas? or moved to NY, i know Yun Mao is in NY. | 16:09 |
| jaypipes | winston-d: yeah, Yun and Simon are in New Jersey actually... I'm still in Ohio, but the whole team is spread over the US, from Austin to colorado to LA to Maryland to SFO :) | 16:10 |
| *** markwash has joined #openstack-dev | 16:10 | |
| *** utlemming has joined #openstack-dev | 16:10 | |
| *** Tross has joined #openstack-dev | 16:10 | |
| *** gyee has joined #openstack-dev | 16:10 | |
| *** Tross has left #openstack-dev | 16:11 | |
| *** woorea has joined #openstack-dev | 16:11 | |
| winston-d | jaypipes, ahh, cool. deploy & operate remotely is fun, isn't it? :) | 16:12 |
| jaypipes | winston-d: you know it! | 16:12 |
| *** jcooley is now known as jcooley_afk | 16:12 | |
| *** jdurgin1 has joined #openstack-dev | 16:13 | |
| winston-d | jaypipes, anyway, what are you working on besides tempest? anything interesting? | 16:15 |
| *** benner has quit IRC | 16:16 | |
| jaypipes | winston-d: a whole lot of chef and python glue code. | 16:16 |
| jaypipes | winston-d: hopefully sooner rather than later, I can get back to working on some of the core projects more. | 16:17 |
| jaypipes | winston-d: we have a small team and a big deploy schedule, so getting the automation right is essential.. | 16:17 |
| *** TheEstablishment has quit IRC | 16:17 | |
| jaypipes | dhellmann, nijaba: is anyone working on a Ceilometer Horizon plugin/dashboard, by any chance? | 16:19 |
| *** andrewbogott_afk is now known as andrewbogott | 16:19 | |
| jeblair | jaypipes: hi there! seems like devstack should probably be patched to install whatever deps tempest needs... | 16:20 |
| jeblair | jaypipes: does 'python setup.py install' work for tempest? | 16:20 |
| jaypipes | jeblair: lol, no. | 16:20 |
| *** jaustinpage|want has quit IRC | 16:20 | |
| jeblair | jaypipes: if not, maybe just have devstack "pip install -r tools/pip-requires" as a special case | 16:20 |
| jaypipes | jeblair: yeah, that's what we suggested to fungi | 16:20 |
| jeblair | jaypipes: (that way all the "set up and configure apps and dependencies" logic stays in devstack) | 16:20 |
| *** benner has joined #openstack-dev | 16:20 | |
| jaypipes | jeblair: would be great, of course, to have real setup.py-ability. | 16:21 |
| jeblair | jaypipes: yeah, just saying devstack vs devstack-gate. | 16:21 |
| jaypipes | oh! yeah, sure | 16:21 |
| *** dolphm_ has joined #openstack-dev | 16:21 | |
| fungi | cool. i'll wedge it into devstack instead of devstack-gate then | 16:21 |
| fungi | thanks! | 16:22 |
| *** dolphm has quit IRC | 16:22 | |
| afazekas | jaypipes: fungi: http://pastebin.com/qseHF6t9 I do not know this script parts are merged jet or the manifest.xml provided in an another way. The dependent tests cases are SKIPing automatically when its missing. | 16:22 |
| *** garyk has quit IRC | 16:23 | |
| nijaba | jaypipes: we have a bp, but it is not assigned to anyone: https://blueprints.launchpad.net/ceilometer/+spec/horizon-plugin | 16:23 |
| *** jcooley_afk is now known as jcooley | 16:23 | |
| jaypipes | nijaba: cheers, thx! | 16:23 |
| *** sacharya1 has joined #openstack-dev | 16:23 | |
| jaypipes | afazekas: yup, fungi is looking into the fix... | 16:24 |
| * fungi nods | 16:24 | |
| *** jcru is now known as jcru|away | 16:25 | |
| *** sacharya has quit IRC | 16:25 | |
| *** maoy has joined #openstack-dev | 16:27 | |
| *** dprince_ has quit IRC | 16:29 | |
| *** dprince has joined #openstack-dev | 16:29 | |
| *** danwent has joined #openstack-dev | 16:30 | |
| *** thingee has quit IRC | 16:30 | |
| *** thingee has joined #openstack-dev | 16:31 | |
| *** thingee has quit IRC | 16:31 | |
| *** thingee has joined #openstack-dev | 16:33 | |
| *** aabes has joined #openstack-dev | 16:34 | |
| winston-d | jaypipes, have to get off to bed. talk 2 u later. | 16:35 |
| jaypipes | winston-d: cya! | 16:35 |
| winston-d | jaypipes, bye~ | 16:36 |
| *** winston-d is now known as winston-d_zzZ | 16:36 | |
| *** andrewbogott is now known as andrewbogott_afk | 16:37 | |
| *** timello has quit IRC | 16:38 | |
| BLZbubba | morning | 16:41 |
| *** timello has joined #openstack-dev | 16:44 | |
| *** jcru|away is now known as jcru | 16:48 | |
| *** jcru has quit IRC | 16:50 | |
| *** colinmcnamara has joined #openstack-dev | 16:50 | |
| *** anteaya has quit IRC | 16:50 | |
| *** jcru has joined #openstack-dev | 16:50 | |
| *** samkottler is now known as samkottler|bbl | 16:52 | |
| *** joshuamckenty has joined #openstack-dev | 16:53 | |
| *** woorea has quit IRC | 16:54 | |
| *** woorea has joined #openstack-dev | 16:54 | |
| *** colinmcnamara has quit IRC | 16:55 | |
| *** colinmcnamara has joined #openstack-dev | 16:57 | |
| *** datsun180b_ has joined #openstack-dev | 16:58 | |
| *** halj has joined #openstack-dev | 16:59 | |
| *** brucer has quit IRC | 17:00 | |
| *** brucer_afk has joined #openstack-dev | 17:00 | |
| *** datsun180b has quit IRC | 17:01 | |
| *** joe5081 has joined #openstack-dev | 17:01 | |
| *** colinmcnamara has quit IRC | 17:01 | |
| *** joe5081 has quit IRC | 17:01 | |
| openstackgerrit | A change was merged to openstack/nova: Move imagecache code from nova.virt.libvirt.utils https://review.openstack.org/16703 | 17:02 |
| *** joe5081 has joined #openstack-dev | 17:02 | |
| *** datsun180b_ has quit IRC | 17:02 | |
| *** datsun180b has joined #openstack-dev | 17:02 | |
| *** joe5081 has quit IRC | 17:02 | |
| openstackgerrit | A change was merged to openstack/nova: Move ec2 config opts to nova.api.ec2.cloud https://review.openstack.org/16706 | 17:02 |
| *** joe5081 has joined #openstack-dev | 17:03 | |
| BLZbubba | ok is it possible to get the ramfilter working in essex; i can't quite tell from this bug report: https://bugs.launchpad.net/nova/+bug/1016273 | 17:03 |
| uvirtbot | Launchpad bug 1016273 in nova/essex "ram_allocation_ratio does not work" [High,Fix released] | 17:03 |
| *** jcru has quit IRC | 17:03 | |
| BLZbubba | not sure which package should have the filter code | 17:03 |
| *** garyk has joined #openstack-dev | 17:06 | |
| *** cp16net is now known as cp16net|away | 17:06 | |
| *** cp16net|away is now known as cp16net | 17:08 | |
| *** woorea has quit IRC | 17:08 | |
| *** avishay has quit IRC | 17:11 | |
| afazekas | jaypipes: , fungi: I will be on IRC in 1.5h again | 17:12 |
| fungi | afazekas: okay, cool | 17:12 |
| *** andrewbogott_afk is now known as andrewbogott | 17:12 | |
| *** jtran has joined #openstack-dev | 17:13 | |
| *** Mandell has joined #openstack-dev | 17:13 | |
| *** afazekas has quit IRC | 17:13 | |
| *** woorea has joined #openstack-dev | 17:13 | |
| *** woorea has quit IRC | 17:15 | |
| *** otherwiseguy has joined #openstack-dev | 17:15 | |
| *** epim has joined #openstack-dev | 17:17 | |
| *** Tross has joined #openstack-dev | 17:17 | |
| *** martine has joined #openstack-dev | 17:18 | |
| *** jaustinpage|want has joined #openstack-dev | 17:19 | |
| markmc | adam_g, nice work on bug #1083818 | 17:23 |
| uvirtbot | Launchpad bug 1083818 in nova/folsom "Detached and deleted RBD volumes remain associated with insance" [Critical,In progress] https://launchpad.net/bugs/1083818 | 17:23 |
| *** sacharya1 has quit IRC | 17:23 | |
| markmc | adam_g, isn't it needed on cinder stable/folsom too ? | 17:23 |
| *** sacharya has joined #openstack-dev | 17:23 | |
| dims | markmc, looks like the gate job worked! - https://review.openstack.org/#/c/16671/ | 17:23 |
| markmc | dhellmann, russellb, maybe one of you could approve https://review.openstack.org/16789 and https://review.openstack.org/16671 ? | 17:24 |
| mordred | whee! | 17:24 |
| mordred | jaypipes: why don't you have real setup.py ability now? | 17:25 |
| jaypipes | mordred: just never did it. | 17:25 |
| jaypipes | mordred: laziness. | 17:26 |
| mordred | jaypipes, fungi: we could just add setup.py ability to tempest real quick | 17:26 |
| *** woorea has joined #openstack-dev | 17:27 | |
| jaypipes | mordred: that would be awesome. | 17:28 |
| fungi | mordred: it has a setup.py, but i guess it's insufficient or something | 17:28 |
| fungi | looks like jaypipes added that file back in july | 17:29 |
| jaypipes | fungi: lol, but I'm not sure it works ;_) | 17:29 |
| fungi | ha! got it | 17:29 |
| *** dingus9 has quit IRC | 17:29 | |
| * mordred making patch | 17:30 | |
| fungi | well, i'm waiting on jenkins to thumbs-up the boto fix for tempest in devstack, but i can just reset that to wip | 17:30 |
| openstackgerrit | A change was merged to openstack/nova: Use flags() helper method to override config in tests https://review.openstack.org/17044 | 17:31 |
| *** vishy_zz is now known as vishy | 17:32 | |
| *** aabes has left #openstack-dev | 17:32 | |
| openstackgerrit | A change was merged to openstack/nova: Remove default_image config option https://review.openstack.org/16707 | 17:34 |
| mordred | jaypipes, fungi: https://review.openstack.org/17063 | 17:35 |
| *** derekh has quit IRC | 17:35 | |
| *** jdurgin1 has quit IRC | 17:38 | |
| *** martine has quit IRC | 17:41 | |
| *** jog0 has joined #openstack-dev | 17:42 | |
| dims | anyone have a few mins, can you please review https://review.openstack.org/#/c/16950/ | 17:42 |
| *** dolphm_ has quit IRC | 17:43 | |
| markmc | vishy, there? | 17:44 |
| vishy | markmc: yup | 17:44 |
| markmc | vishy, remove_unused_base_images = False | 17:44 |
| markmc | vishy, is that enough for 2012.2.1 ? | 17:44 |
| markmc | vishy, to avoid the image cache cleanup thing | 17:45 |
| *** dolphm has joined #openstack-dev | 17:45 | |
| vishy | yeah although I think setting the interval=0 might be better | 17:45 |
| markmc | vishy, whyso? just generally nervous about the cache manager? | 17:45 |
| vishy | since it will be going crazy md5summing on shared storage and still running all the logic otherwise | 17:45 |
| vishy | so it may not delete images but I suspect the performance will still be horrible | 17:46 |
| markmc | ok, fair enough | 17:46 |
| vishy | that's why i picked setting the interval to 0 in the release notes | 17:46 |
| *** Elwell has quit IRC | 17:47 | |
| *** mrodden has quit IRC | 17:47 | |
| markmc | vishy, cool, just checking | 17:47 |
| mordred | jaypipes: what the heck just failed with that tempest change? | 17:49 |
| *** jaustinpage|want has quit IRC | 17:50 | |
| mordred | jaypipes: oh. ffs. there's an openstack.py file in there | 17:50 |
| *** jog0_ has joined #openstack-dev | 17:50 | |
| mordred | makes it kinda hard to drop openstack.common dirs in the tree | 17:50 |
| *** AlanClark has joined #openstack-dev | 17:50 | |
| * mordred cries | 17:50 | |
| *** winston-d_zzZ has quit IRC | 17:50 | |
| jaypipes | lol | 17:51 |
| mordred | jaypipes: I will come back to this | 17:51 |
| *** imsplitbit has joined #openstack-dev | 17:51 | |
| *** woorea has quit IRC | 17:51 | |
| *** kaganos has joined #openstack-dev | 17:52 | |
| *** colinmcnamara has joined #openstack-dev | 17:53 | |
| *** jog0 has quit IRC | 17:53 | |
| *** jog0_ is now known as jog0 | 17:53 | |
| *** utlemming has quit IRC | 17:54 | |
| *** markmc is now known as mcaway | 17:54 | |
| *** dingus9 has joined #openstack-dev | 17:55 | |
| vishy | markmc: I assume you will add something to the release notes saying it has been turned off and how to re-enable it? | 17:56 |
| *** kpavel has quit IRC | 17:57 | |
| *** mrodden has joined #openstack-dev | 17:58 | |
| *** colinmcnamara1 has joined #openstack-dev | 17:58 | |
| *** colinmcnamara has quit IRC | 17:58 | |
| *** kpavel has joined #openstack-dev | 17:58 | |
| *** mnaser has quit IRC | 17:59 | |
| *** samkottler|bbl is now known as samkottler | 17:59 | |
| *** shardy is now known as shardy_afk | 18:01 | |
| *** mnaser has joined #openstack-dev | 18:02 | |
| *** colinmcnamara has joined #openstack-dev | 18:02 | |
| *** colinmcnamara1 has quit IRC | 18:03 | |
| *** rushiagr has joined #openstack-dev | 18:03 | |
| *** dingus91 has joined #openstack-dev | 18:03 | |
| *** martine has joined #openstack-dev | 18:04 | |
| *** dingus9 has quit IRC | 18:04 | |
| *** joshuamckenty has quit IRC | 18:04 | |
| *** jaustinpage|want has joined #openstack-dev | 18:04 | |
| *** colinmcnamara1 has joined #openstack-dev | 18:05 | |
| *** salv-orlando has quit IRC | 18:05 | |
| *** joshuamckenty has joined #openstack-dev | 18:07 | |
| *** colinmcnamara has quit IRC | 18:07 | |
| *** colinmcnamara has joined #openstack-dev | 18:08 | |
| *** colinmcnamara1 has quit IRC | 18:10 | |
| *** Ruetobas has quit IRC | 18:11 | |
| *** colinmcnamara1 has joined #openstack-dev | 18:11 | |
| doude | vishy: thanks to validate the review https://review.openstack.org/#/c/16153/. I forget to update small unit tests. I just pushed them on the review. The review needs to be revalidated | 18:12 |
| *** afazekas has joined #openstack-dev | 18:12 | |
| *** dingus91 has quit IRC | 18:12 | |
| doude | dansmith: thanks to validate the review https://review.openstack.org/#/c/16153/. I forget to update small unit tests. I just pushed them on the review. The review needs to be revalidated | 18:12 |
| *** colinmcnamara has quit IRC | 18:13 | |
| * afazekas re | 18:13 | |
| *** colinmcnamara has joined #openstack-dev | 18:13 | |
| dansmith | doude: okay, I just -2'd it to make sure it won't merge | 18:13 |
| dansmith | alexpilotti: sorry, I forgot to get back to you | 18:13 |
| alexpilotti | dansmith: np! | 18:14 |
| *** zing has quit IRC | 18:14 | |
| *** darraghb has quit IRC | 18:14 | |
| alexpilotti | dansmith: I wanted to ask your opinion about a cloud-init for WIndows project we are working on | 18:14 |
| dansmith | alexpilotti: as someone who knows nearly nothing about windows, I'm honored! :D | 18:15 |
| alexpilotti | dansmith: lol :-) | 18:15 |
| alexpilotti | dansmith: the question is about where to commit this projject | 18:15 |
| *** colinmcnamara1 has quit IRC | 18:15 | |
| *** woorea has joined #openstack-dev | 18:15 | |
| alexpilotti | dansmith: would you suggest to create an ad hoc project to be incubated in OpenStack (if possible) or to see if it's possible to commit it directly to cloud-init? | 18:16 |
| alexpilotti | dansmith: considering in particular the Apache2 vd GPL licensing involved | 18:17 |
| dansmith | alexpilotti: I dunno, I think smoser is probably a better person to ask about that | 18:17 |
| dansmith | or at least, he'd have a pointer for you I think | 18:17 |
| alexpilotti | dansmith: another question related to cloud-init | 18:17 |
| alexpilotti | dansmith: in nova, AFAIK there's no way to communicate with the guest after boot (e.g. after the metadata has been read on boot) | 18:18 |
| dansmith | alexpilotti: I think that's the job of an "agent" | 18:19 |
| alexpilotti | dansmith: on Azure for example is very useful to communicate with a local service and do actions like change password, inject files / configurations etc at any moment | 18:19 |
| dansmith | alexpilotti: which is what xen, vmware, virtualbox do | 18:19 |
| *** dingus9 has joined #openstack-dev | 18:20 | |
| vishy | doude: thanks. Did you see my new note? | 18:20 |
| alexpilotti | dansmith: sure. but still afaik, there's no way in nova to issue a command like "update metadata for that instance", or am I wrong? | 18:21 |
| alexpilotti | dansmith: I just started digging into this, so I still have to explore quite a bit of options | 18:21 |
| *** jcooley is now known as jcooley_afk | 18:22 | |
| alexpilotti | dansmith: especially the authentication part, is tightly coupled with SSH keypairs | 18:22 |
| dansmith | alexpilotti: updating metadata is a job for the hypervisor, not the guest, right? | 18:23 |
| dansmith | alexpilotti: updating the admin password is the job of the agent | 18:23 |
| alexpilotti | dansmith: making it quite difficult to integrate OSs (like ehm… Windows) which don't support this type of asymmetric key auth | 18:23 |
| dansmith | alexpilotti: although vishy had some thoughts on doing that through the cloud-init type interface | 18:23 |
| alexpilotti | dansmith: sure, what I see is the missing link between nova and the agent | 18:24 |
| *** dolphm has quit IRC | 18:24 | |
| dansmith | alexpilotti: what's missing? | 18:24 |
| alexpilotti | dansmith: the agent can pick up from the metadata the updated config, communicate with the hypervisor in a driver specific way, and the guest can apply the update (eg. set a password) | 18:25 |
| dansmith | alexpilotti: right, that's how it works in nova with xen now, right? | 18:25 |
| alexpilotti | dansmith: what I don't get, is: what nova command should I issue to update a configuration after boot so that the agent can pick it up? | 18:26 |
| alexpilotti | dansmith: for exaple a new userdata script | 18:26 |
| *** rkukura has quit IRC | 18:27 | |
| *** rushiagr1 has joined #openstack-dev | 18:27 | |
| dansmith | alexpilotti: honestly, I really don't know the details of that sort of plumbing because I only use nova with kvm | 18:28 |
| *** alobbs has quit IRC | 18:28 | |
| dansmith | alexpilotti: so you should talk to one of the folks that knows the details of the (much more functional) xen driver | 18:28 |
| *** dolphm has joined #openstack-dev | 18:28 | |
| *** rushiagr has quit IRC | 18:28 | |
| dansmith | alexpilotti: it's entirely possible that there *isn't* such a call right now, and that you'd want to write an extension to add it | 18:28 |
| dansmith | which I think would be fine | 18:28 |
| alexpilotti | dansmith: that was what I was fearing. Wouldn't it be better to have an hypervisor independent feature? (handled by each driver as usual of course) | 18:30 |
| dansmith | alexpilotti: being an extension doesn't mean it's hypervisor-specific | 18:30 |
| alexpilotti | dansmith: ok, but wouldn't it be easier to have it in nova "out of the box"? :-) | 18:31 |
| alexpilotti | dansmith: I see it as a quite useful feature. Or at least my customers do :-D | 18:31 |
| *** colinmcnamara has quit IRC | 18:31 | |
| dansmith | alexpilotti: you mean "in the core nova api" not "in nova out of the box" | 18:31 |
| *** lloydde has joined #openstack-dev | 18:31 | |
| *** corXi has quit IRC | 18:31 | |
| dansmith | because an extension would satisfy the latter | 18:31 |
| *** colinmcnamara has joined #openstack-dev | 18:32 | |
| alexpilotti | dansmith: yes, sorry for my lack of formal terminology! | 18:32 |
| dansmith | alexpilotti: I don't know that libvirt/kvm would ever support that feature in the same say (for copying files) so I don't know that it *does* make sense to be in the core api | 18:32 |
| dansmith | alexpilotti: there is a ton of critical functionality that is technically an extension (floating ips for example) | 18:32 |
| dansmith | so I don't think that this is really more important than that, no | 18:32 |
| *** rushiagr1 is now known as rushiagr | 18:33 | |
| alexpilotti | dansmith: got it | 18:33 |
| dansmith | alexpilotti: feel free to bring it up with the nova core team if you want a real ruling on it though, I'm just one unimportant person :D | 18:33 |
| alexpilotti | dansmith: I'd like to talk about this with the xen guys to see how they approached it. Do you have somebody to suggest that I should bug.. ehm contact? | 18:34 |
| dansmith | alexpilotti: I don't really know, but you might ask jerdfelt_, comstud, etc | 18:35 |
| dansmith | xen is almost as icky as windows for me, so I don't really even know who uses it :P | 18:35 |
| *** glauaguiar has quit IRC | 18:35 | |
| alexpilotti | dansmith: loool I liked the "almost" part :-D | 18:36 |
| *** vishy is now known as vishy_zz | 18:36 | |
| alexpilotti | dansmith: thanks a lot as usual!! :-) | 18:36 |
| dansmith | alexpilotti: yep :) | 18:36 |
| alexpilotti | smoser: ping | 18:37 |
| *** vishy_zz is now known as vishy | 18:42 | |
| *** adalbas has quit IRC | 18:43 | |
| *** colinmcnamara has quit IRC | 18:47 | |
| *** colinmcnamara has joined #openstack-dev | 18:48 | |
| *** ChanServ sets mode: +v jeblair | 18:48 | |
| *** jcooley_afk is now known as jcooley | 18:48 | |
| vishy | jgriffith: ping | 18:49 |
| jgriffith | vishy: hey | 18:49 |
| *** garyk has quit IRC | 18:49 | |
| vishy | did i tell you that you can remove the monkeypatch workaround? | 18:50 |
| *** dolphm has quit IRC | 18:50 | |
| vishy | jgriffith: we fixed the underlying issue in keystone to work with and without os monkeypatched | 18:50 |
| *** ChanServ sets mode: -v pandemicsyn | 18:51 | |
| jgriffith | vishy: No... did the eventlet patch get released? | 18:51 |
| jgriffith | vishy: Ahh... excellent | 18:51 |
| jgriffith | vishy: Ok, thanks for the update | 18:51 |
| *** colinmcnamara1 has joined #openstack-dev | 18:51 | |
| *** vipul is now known as vipul|away | 18:51 | |
| jgriffith | vishy: I'll remove it now... we never approved the Backport IIRC so that's good | 18:52 |
| *** colinmcnamara has quit IRC | 18:52 | |
| vishy | jgriffith: note that api will still need it if you ever port over the multiple api worker code from nova | 18:53 |
| vishy | (at least until the eventlet patch is in the distros) | 18:53 |
| *** eglynn__ has quit IRC | 18:53 | |
| *** eglynn has joined #openstack-dev | 18:53 | |
| jgriffith | vishy: I'll add the comment | 18:53 |
| jgriffith | vishy: thanks | 18:53 |
| clarkb | vishy: speaking of monkeypatching http://logs.openstack.org/15078/7/check/gate-nova-python27/10319/console.html.gz has a bunch of failures which appear to be related to the monkey patching of libvirt_utils | 18:55 |
| clarkb | this is the run nova unittests under testr change so things are run in different orders and is less forgiving of unit tests that aren't really independent | 18:55 |
| *** colinmcnamara1 has quit IRC | 18:56 | |
| *** salv-orlando has joined #openstack-dev | 18:56 | |
| clarkb | any ideas? https://github.com/openstack/nova/blob/master/nova/tests/test_libvirt.py#L584 appears to need an unmonkeypatch for snapshots.libvirt_utils, but I don't think that is causing the breakage | 18:57 |
| *** colinmcnamara has joined #openstack-dev | 18:57 | |
| *** AlanClark has quit IRC | 18:59 | |
| vishy | clarkb: any way to repro locally? | 18:59 |
| vishy | preferrably with a small number of tests :) | 19:00 |
| *** novas0x2a|laptop has joined #openstack-dev | 19:00 | |
| clarkb | vishy: I have not been able to reproduce locally, and the py26 tests pass | 19:00 |
| clarkb | (those py26 tests pass) I am still trying to sort out how to reproduce it | 19:00 |
| *** datsun180b has quit IRC | 19:00 | |
| *** colinmcnamara has quit IRC | 19:02 | |
| *** cp16net is now known as cp16net|away | 19:05 | |
| *** dolphm has joined #openstack-dev | 19:06 | |
| *** esp1 has joined #openstack-dev | 19:07 | |
| ayoung | dolphm, any opinions on where I should place the trusts backend? | 19:09 |
| *** adalbas has joined #openstack-dev | 19:10 | |
| dolphm | ayoung: identity or a new whole new backend? | 19:10 |
| ayoung | yeahm or possibly token | 19:11 |
| *** alexpilotti has quit IRC | 19:11 | |
| *** garyk has joined #openstack-dev | 19:11 | |
| dolphm | ayoung: token makes more sense than identity | 19:11 |
| ayoung | dolphm, like it in identity, but I don't think I can make the LDAP side work. | 19:11 |
| *** colinmcnamara has joined #openstack-dev | 19:11 | |
| Ryan_Lane | I'm not sure why identity can only have one backend | 19:12 |
| ayoung | dolphm, Ryan_Lane because no one has written anything to make multiple, but thanks for volunteering | 19:12 |
| Ryan_Lane | wasn't there discussion about having portions of it stored in mysql? | 19:12 |
| Ryan_Lane | heh. if I had time I'd gladly do so | 19:12 |
| ayoung | Ryan_Lane, yeah, there is the idea that you have it ins SQL, but fetch on demand certain data from LDAp, but there are some issues to dance around in there | 19:13 |
| dolphm | Ryan_Lane: yeah, that conversation stalled at what pieces should go where... there's conflicting opinions / use cases | 19:13 |
| Ryan_Lane | well, for the most part it should be clear-ish what can live in ldap, unless we want to have a non-custom schema | 19:13 |
| Ryan_Lane | err | 19:14 |
| Ryan_Lane | non-standard | 19:14 |
| vishy | Ryan_Lane, ayoung: while I have you guys here | 19:14 |
| ayoung | Ryan_Lane, yep, which is why I don't think Trusts can live there, unless you know an analogue we can use? | 19:14 |
| vishy | thoughts on tenant_name vs. tenant_id ? | 19:14 |
| * ayoung has left | 19:14 | |
| *** danpb has quit IRC | 19:15 | |
| Ryan_Lane | vishy: in my implementation they are the same thing | 19:15 |
| *** colinmcnamara1 has joined #openstack-dev | 19:15 | |
| Ryan_Lane | I think the uuid is useless in LDAP | 19:15 |
| *** adjohn has joined #openstack-dev | 19:15 | |
| *** rushiagr1 has joined #openstack-dev | 19:15 | |
| ayoung | vishy, context? | 19:16 |
| *** colinmcnamara has quit IRC | 19:16 | |
| vishy | Ryan_Lane: that is one option I'm considering | 19:16 |
| *** rushiagr1 has left #openstack-dev | 19:16 | |
| vishy | ayoung: we only store tenant_id in nova | 19:16 |
| *** adalbas has quit IRC | 19:16 | |
| vishy | but that means we expose a random string to the user in a bunch of places | 19:16 |
| vishy | where the tenant_name would be much more useful. | 19:17 |
| *** rushiagr has quit IRC | 19:17 | |
| Ryan_Lane | we expect tenant name to be unique | 19:17 |
| Ryan_Lane | what's the point of the uuid? | 19:17 |
| dolphm | vishy: ideally the ID's are only used internally, and the name is exposed to the user, because it's user-defined | 19:17 |
| Ryan_Lane | right now we have two unique identifiers | 19:17 |
| *** andrewbogott is now known as andrewbogott_afk | 19:17 | |
| vishy | Ryan_Lane: I think the purpose is so that the name can be changed | 19:18 |
| Ryan_Lane | ah | 19:18 |
| vishy | without breaking relationships | 19:18 |
| Ryan_Lane | that does make sense | 19:18 |
| dolphm | Ryan_Lane: id: service-defined, immutable, url-friendly, globally unique; name: user-defined, mutable, freeform text, deployment unique | 19:18 |
| Ryan_Lane | except it doesn't really work when using the ldap backend | 19:18 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Replace try: import with extras.try_import https://review.openstack.org/16789 | 19:19 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Make project pyflakes clean. https://review.openstack.org/16671 | 19:21 |
| Ryan_Lane | I'm sure it's doable in the ldap backend, though | 19:21 |
| *** joshuamckenty has quit IRC | 19:22 | |
| *** maelfius has joined #openstack-dev | 19:23 | |
| creiht | mordred: ping | 19:25 |
| *** rkukura has joined #openstack-dev | 19:25 | |
| openstackgerrit | A change was merged to openstack/cinder: Unpin lxml requirements https://review.openstack.org/16852 | 19:25 |
| creiht | any idea why: https://review.openstack.org/#/c/16234/ didn't get merged after approval? | 19:25 |
| *** bitblt has joined #openstack-dev | 19:26 | |
| clarkb | creiht: that change depends on https://review.openstack.org/#/c/16233/2 | 19:26 |
| clarkb | the dependency is not approved so the first change will nto merge | 19:27 |
| creiht | clarkb: ahh thanks | 19:27 |
| creiht | missed the depends on | 19:27 |
| ayoung | dolphm, so, separate backend or token? You seem to favor token. The thing is, it feels different. Like, token data it would almost make sense to store in memcached, but trusts are more contracts: keep forever type stuff | 19:27 |
| creiht | been a while since I have done this stuff :) | 19:27 |
| ayoung | But, unlike most of the stuff in identity, they are defined by the end user. | 19:27 |
| ayoung | Although only a user can create a trust for themselves | 19:28 |
| dolphm | ayoung: can trusts not expire? | 19:29 |
| clarkb | creiht: np | 19:29 |
| ayoung | dolphm, not by default. | 19:30 |
| *** cp16net|away is now known as cp16net | 19:30 | |
| ayoung | dolphm, no more than a user account can expire. doesn't really fit the abstraction | 19:31 |
| ayoung | instead, if a trust is no longer needed, it should be deleted at that point | 19:31 |
| *** colinmcnamara1 has quit IRC | 19:31 | |
| *** colinmcnamara has joined #openstack-dev | 19:31 | |
| *** jp_at_hp has joined #openstack-dev | 19:37 | |
| *** krtaylor has quit IRC | 19:43 | |
| *** Elwell has joined #openstack-dev | 19:45 | |
| *** tmishler__ has joined #openstack-dev | 19:46 | |
| *** tmishler has quit IRC | 19:46 | |
| *** tmishler__ has quit IRC | 19:46 | |
| *** zaitcev has joined #openstack-dev | 19:48 | |
| *** woorea has quit IRC | 19:48 | |
| *** shardy_afk is now known as shardy | 19:49 | |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Add --version CLI opt and __version__ module attr https://review.openstack.org/16985 | 19:49 |
| *** colinmcnamara has quit IRC | 19:49 | |
| *** colinmcnamara has joined #openstack-dev | 19:50 | |
| dims | vishy, a question about image_cache_manager_interval=0, seems like it sets nova/manager.py may end up calling _run_image_cache_manager_pass more periodically. do we need to add a check in manager.py to skip tasks which have ticks_between_runs set to 0 as well? | 19:50 |
| *** tmishler has joined #openstack-dev | 19:51 | |
| *** tmishler has joined #openstack-dev | 19:52 | |
| *** networkstatic has joined #openstack-dev | 19:53 | |
| *** salv-orlando has quit IRC | 19:53 | |
| *** salv-orlando has joined #openstack-dev | 19:53 | |
| dims | anyone else know? :) | 19:54 |
| *** salv-orlando has quit IRC | 19:54 | |
| *** salv-orlando has joined #openstack-dev | 19:54 | |
| dprince | markmcclain: around? Would you mind having a look at this one: https://review.openstack.org/#/c/17070/ | 19:54 |
| dprince | markmcclain: would love to get that in so I can run some more tests! | 19:55 |
| *** flaper87 has joined #openstack-dev | 19:56 | |
| *** flaper87 has joined #openstack-dev | 19:56 | |
| markmcclain | dprince: looking | 19:56 |
| *** thingee is now known as thingee_zz | 19:57 | |
| *** uvirtbot has quit IRC | 19:57 | |
| *** tmishler has joined #openstack-dev | 19:57 | |
| *** hub_cap has quit IRC | 19:58 | |
| *** hub_cap has joined #openstack-dev | 19:58 | |
| *** uvirtbot has joined #openstack-dev | 19:58 | |
| Ryan_Lane | does anyone have contact info for Greg Althaus? | 19:58 |
| Ryan_Lane | (email address, preferably) | 19:59 |
| *** zbitter is now known as zaneb | 19:59 | |
| *** woorea has joined #openstack-dev | 20:00 | |
| vishy | dims: interesting question | 20:01 |
| vishy | dims: let me look | 20:01 |
| *** networkstatic has quit IRC | 20:01 | |
| mordred | bcwaldon: hey - on that thing that just merged to keystoneclient- you know there is an oslo function for retrieving the version, right? | 20:01 |
| bcwaldon | mordred: I will admit I didn't look into that too closely - I copied code out of python-glanceclient | 20:02 |
| vishy | dims: ugh, I think you are right | 20:02 |
| vishy | so we will get a log on every tick... | 20:02 |
| vishy | which is every 60 seconds by default | 20:02 |
| *** dolphm has quit IRC | 20:03 | |
| dims | vishy, ouch! | 20:03 |
| bcwaldon | mordred: if we can simplify that code in python-{nova|glance|keysotne}client, I would love to do so | 20:03 |
| mordred | bcwaldon: lovely. I would like to do that | 20:03 |
| mordred | bcwaldon: I will go look and make a patch for you at some point | 20:03 |
| vishy | dims: not hugely painful I suppose but in the future we shoudl probably allow None to disable it. | 20:03 |
| bcwaldon | mordred: fantastic | 20:04 |
| dims | vishy, worth logging a bug? | 20:05 |
| *** dolphm has joined #openstack-dev | 20:05 | |
| openstackgerrit | A change was merged to openstack/quantum: Make signing dir a subdir in /var/lib/quantum. https://review.openstack.org/17070 | 20:05 |
| vishy | dims: sure | 20:06 |
| *** alexpilotti has joined #openstack-dev | 20:06 | |
| *** colinmcnamara has quit IRC | 20:06 | |
| *** colinmcnamara has joined #openstack-dev | 20:07 | |
| *** maurosr has quit IRC | 20:07 | |
| *** bitblt has quit IRC | 20:08 | |
| alexpilotti | smoser: ping | 20:08 |
| dims | vishy, "Allow None for ticks_between_runs for periodic tasks" sound good? | 20:08 |
| alexpilotti | jerdfelt_: ping | 20:09 |
| dims | going with "Allow None for ticks_between_runs when configuring periodic tasks" | 20:09 |
| *** colinmcnamara1 has joined #openstack-dev | 20:11 | |
| *** colinmcnamara has quit IRC | 20:11 | |
| ayoung | Ryan_Lane, do you know of anything that would be comparable to trusts/impersonation in the LDAP world? | 20:12 |
| Ryan_Lane | yes | 20:12 |
| Ryan_Lane | proxy authentication | 20:12 |
| Ryan_Lane | which is different than what people call a proxy agent | 20:12 |
| Ryan_Lane | though maybe people originally used proxy authentication with proxy agents | 20:13 |
| Ryan_Lane | proxy auth is fucking evil | 20:13 |
| Ryan_Lane | it's specifically impersonation | 20:13 |
| ayoung | Ryan_Lane, yes. | 20:14 |
| Ryan_Lane | I'd imagine all servers support it | 20:14 |
| Ryan_Lane | I don't know many people that actually grant that permission to anyone | 20:14 |
| *** markwash has quit IRC | 20:14 | |
| ayoung | Ryan_Lane I'd want to be able to do that, but limit it to a single role and/or end point | 20:14 |
| Ryan_Lane | most implementations aren't fine grained. if you allow proxy auth for a user it allows that user to impersonate anyone | 20:15 |
| *** jcooley is now known as jcooley_afk | 20:15 | |
| ayoung | So, say, the high availaibilty agent can migrate your vm for you from one host to another | 20:15 |
| * Ryan_Lane nods | 20:15 | |
| Ryan_Lane | wouldn't delegation be better for that? | 20:15 |
| ayoung | No. | 20:15 |
| ayoung | That is, of course it would, but we can't use it | 20:15 |
| Ryan_Lane | ah | 20:16 |
| ayoung | the resources need to be owned by the actual user, not the service user that is performing that action for them | 20:16 |
| Ryan_Lane | impersonation seems like a great code path to find security vulnerabilities | 20:17 |
| Ryan_Lane | eh? why? | 20:17 |
| ayoung | So, yeah, you don;t want to give away the keys to the kingdom, just the keys to the latrine at the royal hunting house | 20:17 |
| *** colinmcnamara1 has quit IRC | 20:17 | |
| Ryan_Lane | resources are owned by tenants, not users | 20:17 |
| *** colinmcnamara has joined #openstack-dev | 20:17 | |
| ayoung | So now we make the service user a memeber of every tenant? | 20:17 |
| Ryan_Lane | oauth seems to handle this fairly well | 20:18 |
| ayoung | Ryan_Lane, great, lets chuck Keystone and use oAuth | 20:18 |
| *** sacharya1 has joined #openstack-dev | 20:18 | |
| Ryan_Lane | heh | 20:18 |
| Ryan_Lane | that's not what I mean | 20:18 |
| Ryan_Lane | I'm talking about looking at oauth's model | 20:18 |
| *** halj1 has joined #openstack-dev | 20:18 | |
| Ryan_Lane | the service asks to act on behalf of the user | 20:19 |
| Ryan_Lane | requesting specific permissions | 20:19 |
| ayoung | right | 20:19 |
| Ryan_Lane | it gets a token | 20:19 |
| *** halj has quit IRC | 20:19 | |
| ayoung | that is what this | 20:19 |
| ayoung | is | 20:19 |
| Ryan_Lane | that token is then given delegated permissions | 20:19 |
| ayoung | and by "asks" | 20:19 |
| *** colinmcnamara1 has joined #openstack-dev | 20:19 | |
| Ryan_Lane | that's different than impersonation | 20:19 |
| ayoung | I mean requests a token | 20:19 |
| ayoung | no | 20:19 |
| ayoung | the service user needs to pass a token along with the operation | 20:19 |
| ayoung | that token has, as the user_id, the user that needs the action performed | 20:20 |
| openstackgerrit | A change was merged to openstack/nova: Add methods to Host operations to fake hypervisor https://review.openstack.org/16567 | 20:20 |
| ayoung | it is not really any different than handing a token to nova compute, which nova then hands off to glance | 20:20 |
| ayoung | it just happens later, a lot later | 20:20 |
| Ryan_Lane | will the token have limited scope? | 20:21 |
| *** colinmcnamara has quit IRC | 20:22 | |
| *** sacharya has quit IRC | 20:22 | |
| ayoung | Ryan_Lane, yes. yes. yes. Very limited . Limitedississimo | 20:22 |
| Ryan_Lane | :D | 20:22 |
| *** cp16net is now known as cp16net|away | 20:22 | |
| ayoung | That is the whole poinmt | 20:22 |
| ayoung | point | 20:22 |
| Ryan_Lane | ah. ok. that seems sane enough. | 20:22 |
| *** cp16net|away is now known as cp16net | 20:22 | |
| ayoung | Ryan_Lane, so I could avoid this whole thing if I were not being such a purist | 20:22 |
| openstackgerrit | A change was merged to openstack/glance: Use auth_token middleware from python-keystoneclient. https://review.openstack.org/16966 | 20:23 |
| ayoung | I like it when databases, relational type, are allwed to do things like integrity checking | 20:23 |
| Ryan_Lane | I still don't know if I'd call that impersonation, but that's semantics :D | 20:23 |
| *** cp16net is now known as cp16net|away | 20:23 | |
| *** cp16net|away is now known as cp16net | 20:23 | |
| ayoung | Ryan_Lane, that is why I called it Trusts | 20:23 |
| guitarzan | russellb: are you the mastermind behind the conductor stuff? | 20:23 |
| ayoung | The question is whether I can store the trust data in LDAP | 20:24 |
| russellb | guitarzan: heh, one of them i guess | 20:24 |
| russellb | what's up | 20:24 |
| guitarzan | russellb: :) | 20:24 |
| Ryan_Lane | I doubt it | 20:24 |
| guitarzan | I've got an environment using sqlite | 20:24 |
| *** colinmcnamara1 has quit IRC | 20:24 | |
| Ryan_Lane | I don't think it belongs there, thoguh | 20:24 |
| Ryan_Lane | *though | 20:24 |
| ayoung | Ryan_Lane, why do you get my hopes up just to dash them? Why oh Why? | 20:24 |
| russellb | guitarzan: so ... you shouldn't use sqlite with nova, it's going to break. we really don't recommend it | 20:24 |
| guitarzan | and sqlite isn't happy about throwing strings into datetime fields | 20:24 |
| Ryan_Lane | you only want to store something in ldap if it's going to be used by multiple applications | 20:24 |
| guitarzan | russellb: fair enough, if that's the answer | 20:25 |
| Ryan_Lane | this is only ever going to be used by keystone | 20:25 |
| Ryan_Lane | it belongs in the keystone database | 20:25 |
| guitarzan | russellb: but it has worked for a long time | 20:25 |
| guitarzan | and this is just a test env | 20:25 |
| ayoung | Ryan_Lane, and then you lose all referential integrity | 20:25 |
| russellb | guitarzan: though that's a good point, though. that's an unexpected side effect of these changes. | 20:25 |
| *** colinmcnamara has joined #openstack-dev | 20:25 | |
| Ryan_Lane | that's true, but that's not abnormal | 20:25 |
| Ryan_Lane | this is one of the downsides of ldap integration | 20:26 |
| Ryan_Lane | you have to ensure that yourself | 20:26 |
| russellb | guitarzan: want to file a bug and send me the link? | 20:26 |
| Ryan_Lane | we don't have that right now with tokens either | 20:26 |
| russellb | guitarzan: and would you be willing to test a patch? | 20:26 |
| guitarzan | sure on both counts | 20:26 |
| russellb | k | 20:26 |
| ayoung | Ryan_Lane, I am starting to think that LDAP was a mistake. We should probably have done only SQL, with LDAP calls made out to populate in bulk, and maybe to authenticate | 20:26 |
| guitarzan | I don't want to suggest a sqlite only migration though... | 20:26 |
| Ryan_Lane | nah | 20:27 |
| guitarzan | I'm not sure how else we'd want to marshal stuff over the queue though... | 20:27 |
| ayoung | MOst people can't change their LDAP servers | 20:27 |
| Ryan_Lane | yeah, that's not a problem, though | 20:27 |
| russellb | guitarzan: nah ... I think we can convert the appropriate fields back into a datetime object | 20:27 |
| ayoung | we've made a solution that only works for a minority | 20:27 |
| russellb | before doing the db call | 20:27 |
| guitarzan | russellb: sounds good to me | 20:27 |
| Ryan_Lane | I don't see why | 20:27 |
| Ryan_Lane | we don't need to store most things in ldap | 20:28 |
| Ryan_Lane | store them in the database | 20:28 |
| Ryan_Lane | use ldap as the source of truth for specific things and the database for others | 20:28 |
| ayoung | Ryan_Lane, that is the feedback I get from most people. THey want external authentication, but then the Keystone admins want to control authorization for openstack components | 20:28 |
| Ryan_Lane | use the database as a cache of the source of truth stuff | 20:28 |
| ayoung | Ryan_Lane, right | 20:28 |
| vishy | clarkb: if you end up getting a repro on those tests I'm happy to help track down the issue | 20:29 |
| clarkb | vishy: thanks. lifeless and I are hunting down a way to reproduce | 20:29 |
| Ryan_Lane | honestly, if there's NSS/PAM support for keystone I think a lot of this becomes a non-issue | 20:29 |
| lifeless | Ryan_Lane: +10000 | 20:29 |
| ayoung | Ryan_Lane, explain. | 20:30 |
| Ryan_Lane | except we'd still need to deal with AD | 20:30 |
| lifeless | Ryan_Lane: I was saying we should do a pam + userdir module for that | 20:30 |
| Ryan_Lane | fucking microsoft | 20:30 |
| lifeless | Ryan_Lane: why would you need AD ? | 20:30 |
| Ryan_Lane | ayoung: one of the biggest necessities of ldap is tying instance auth together with openstack auth | 20:30 |
| ayoung | why would AD be any different than any other LDAP module? | 20:30 |
| Ryan_Lane | lifeless: so that windows auth could be tied with openstack auth | 20:31 |
| ayoung | OK, I suspect I need to push trusts onto the stack and deal with this issue up front. | 20:32 |
| Ryan_Lane | why not a different backend? | 20:32 |
| Ryan_Lane | token also kind of makes sense, since it's a form of token | 20:32 |
| ayoung | Ryan_Lane, because it belongs in identity | 20:33 |
| Ryan_Lane | isn't it just a token with a limited scope? | 20:33 |
| ayoung | Ryan_Lane, OK, let me put my DBA hat on | 20:33 |
| ayoung | Nope | 20:33 |
| ayoung | a trust by itself is more like a user than like a token | 20:34 |
| ayoung | it is something that has enough information to create a token | 20:34 |
| * Ryan_Lane nods | 20:34 | |
| ayoung | a trust is a contract. you cannot present a trust to a remote service and say "let me do something" you still need a token in order to do that | 20:34 |
| *** adjohn has quit IRC | 20:35 | |
| ayoung | So, LDAP actually maps better to this than SQL, in that the trust object needs to be owned by the user creating the trust | 20:35 |
| ayoung | I should be able to create an destroy my own trusts | 20:35 |
| *** flaper87 has quit IRC | 20:36 | |
| Ryan_Lane | that seems odd to me | 20:36 |
| ayoung | then, audit is maintained by linking the token to the trust to the trustee | 20:36 |
| ayoung | Ryan_Lane, well, you do it all the time | 20:36 |
| Ryan_Lane | since rights are at the role level, not the user level | 20:36 |
| ayoung | Ryan_Lane, correct | 20:36 |
| Ryan_Lane | and roles are tenant specific | 20:36 |
| ayoung | think about your key ring | 20:36 |
| ayoung | each key is tied to a set of actions | 20:37 |
| ayoung | open front door, start car, check mail | 20:37 |
| *** krtaylor has joined #openstack-dev | 20:37 | |
| ayoung | If I give you a key from my ring, you can only do the actions that are specified by it. If you then give me the key back, someone that saw my car driving around yesterday can balme me for running a red light. And in fact, since you no longer have the kye, I can't even prove it was you | 20:38 |
| ayoung | I realize this is stretching the analogy, but you get the point? | 20:38 |
| Ryan_Lane | not really | 20:38 |
| Ryan_Lane | since the key is a right | 20:38 |
| Ryan_Lane | and it's shared by a number of people in my household | 20:39 |
| ayoung | yeah, yeah... | 20:39 |
| Ryan_Lane | I'm giving away a right that affects others | 20:39 |
| ayoung | as I said, it is stretching the analogy | 20:39 |
| Ryan_Lane | it shouldn't be associated with a user | 20:39 |
| Ryan_Lane | it should be associated with the tenant | 20:39 |
| ayoung | the key is the role in the organization. But by adding a user on to the token, you can say who authorized the action | 20:40 |
| Ryan_Lane | ah | 20:40 |
| Ryan_Lane | now I see what you mean | 20:40 |
| ayoung | yeah. audit | 20:40 |
| Ryan_Lane | you're making the trust into a user so that you can track the usage | 20:40 |
| Ryan_Lane | that makes complete sense | 20:40 |
| *** colinmcnamara has quit IRC | 20:40 | |
| kbringard | vishy: ping | 20:41 |
| vishy | kbringard: sup? | 20:41 |
| Ryan_Lane | now I see why you'd want it in identity | 20:41 |
| Ryan_Lane | it's not a real user, though | 20:41 |
| kbringard | quick ?, in essex stable, do you know if there is an equivalent flag to libvirt_snapshots_directory to set where snapshots get stored before upload? | 20:41 |
| ayoung | so if action A was performed under the auspices of token K which was created under Trust T where the trustor was user U then we have a chain for audit: U->T->K->A | 20:41 |
| Ryan_Lane | right | 20:41 |
| Ryan_Lane | but.... | 20:41 |
| ayoung | Ryan_Lane, it is a real user. It is who authorized the action | 20:41 |
| kbringard | I don't seem to see that flag when grepping the stable/essex code | 20:41 |
| ayoung | and, since it is a trust token, we also know who actually performed the action | 20:42 |
| Ryan_Lane | you could do the same by listing a set of delegations with the user | 20:42 |
| Ryan_Lane | which could be a different table | 20:42 |
| Ryan_Lane | when the token is used, it would be as user A, with delegation D | 20:42 |
| Ryan_Lane | user U, that is | 20:42 |
| ayoung | Ryan_Lane, not without changing all of the downstream consumers of tokens | 20:43 |
| Ryan_Lane | how so? | 20:43 |
| ayoung | this mechanism works without changing glance, nova, etc | 20:43 |
| Ryan_Lane | it would be an attribute of the token | 20:43 |
| ayoung | but they look at the userid associated with the tokne | 20:43 |
| ayoung | now they would have to look at some other attribute | 20:43 |
| Ryan_Lane | why? | 20:43 |
| ayoung | and, would it really make a difference? | 20:43 |
| ayoung | I mean, semantically, it is the same thing. | 20:44 |
| Ryan_Lane | ah, for auditing | 20:44 |
| Ryan_Lane | yes, but from an implementation POV it's cleaner | 20:44 |
| *** garyk has quit IRC | 20:44 | |
| ayoung | what really matters is the the scope of the trust is as limited as possible. | 20:44 |
| Ryan_Lane | it's a separate table and doesn't involve creating what are essentially fake users | 20:44 |
| *** shardy is now known as shardy_afk | 20:45 | |
| ayoung | Hence, when you create a trust, you specify the set of roles. Those are a subset of the roles of the trustor | 20:45 |
| Ryan_Lane | yes | 20:45 |
| Ryan_Lane | and that's associated with the token | 20:45 |
| ayoung | But do you see why I think it is more tightly linked to identity than tokens? | 20:45 |
| Ryan_Lane | I do, but I don't think it needs to live in either token or identity | 20:46 |
| openstackgerrit | A change was merged to openstack/swift: Upgrade pep8 to 1.3.3. https://review.openstack.org/16947 | 20:46 |
| Ryan_Lane | I think it's kind of being forced into one or the other because it seems simpler at first glace | 20:46 |
| Ryan_Lane | *glance | 20:46 |
| *** vipul|away is now known as vipul | 20:47 | |
| ayoung | I want the referential integrity. A trust is basically all foreign keys. The amount of other work necessary to verify what is done by the database makes me say that it would be better off in the same schema as the identity tables | 20:47 |
| Ryan_Lane | referential integrity still exists | 20:47 |
| Ryan_Lane | even when using ldap | 20:47 |
| Ryan_Lane | but, when using ldap, you'd need to look things up, and resources would be left around | 20:48 |
| *** annegentle_ has quit IRC | 20:48 | |
| Ryan_Lane | it's the same case for tokens, though | 20:49 |
| Ryan_Lane | if a user is deleted in ldap, you don't know about it | 20:49 |
| *** thingee_zz is now known as thingee | 20:49 | |
| kbringard | vishy: or perhaps put another way, can you change what nova uses as tmpdir via flag in essex? | 20:49 |
| *** sacharya1 has quit IRC | 20:49 | |
| Ryan_Lane | there's going to be lots of cases like this in future, as well | 20:50 |
| *** sacharya has joined #openstack-dev | 20:50 | |
| Ryan_Lane | ayoung: so, with ldap, the only way you're going to get referential integrity with this is if you are able to write to ldap | 20:52 |
| ayoung | Ryan_Lane, yeah, and we still need to check if a user is enable/disabled at trust activation time. You can make the argument that this is more application specific data than the identity stuff, but I'd like to validate that there is no LDAP solution before I yank the SQL one into a separate schema | 20:52 |
| ayoung | I get that. | 20:52 |
| Ryan_Lane | well... | 20:52 |
| ayoung | I want ref-integ in SQL, and a possible implementation in LDAP as a first step | 20:52 |
| rmk | kbringard: No, there's no way. You have to override it as an environment variable. | 20:52 |
| kbringard | vishy: nevermind, found it in the mailing list archives | 20:52 |
| *** dolphm has quit IRC | 20:52 | |
| Ryan_Lane | openam stores all of this in LDAP | 20:52 |
| kbringard | well, gmi found it | 20:52 |
| Ryan_Lane | but, they require that you import the schema | 20:53 |
| rmk | I raised this bug a while ago. | 20:53 |
| ayoung | Ryan_Lane, any idea what object classes? | 20:53 |
| Ryan_Lane | and they use a secondary internal ldap server | 20:53 |
| rmk | I think I'm going to submit a patch again. | 20:53 |
| rmk | It should absolutely be configurable. | 20:53 |
| Ryan_Lane | it's openam specific schema | 20:53 |
| ayoung | ah, so no way using standard schema? | 20:53 |
| kbringard | rmk: yea, snapshot dir is now a flag | 20:53 |
| Ryan_Lane | no | 20:53 |
| ayoung | meh | 20:53 |
| kbringard | but not overall tmpdir | 20:53 |
| Ryan_Lane | and you'd be requiring people to add data into their ldap directories that is keystone specific | 20:54 |
| kbringard | for reference: http://openstack.markmail.org/search/?q=snapshot%20and%20tmp#query:snapshot%20and%20tmp+page:1+mid:3ofblc6kpvd76zrb+state:results | 20:54 |
| Ryan_Lane | and it looks like you are going to be wanting full user accounts added | 20:54 |
| Ryan_Lane | and if you think people are hesitant to change schema, ask them to add fake users ;) | 20:54 |
| kbringard | rmk: I think what vishy proposed there is a good idea | 20:54 |
| ayoung | Ryan_Lane, yeah, that is what I was thinking | 20:54 |
| Ryan_Lane | this is why I still think it's cleaner to make this a separate table with delegations | 20:55 |
| ayoung | BTW, one problem with the LDAP back end is the Service users...fake users again. | 20:55 |
| kbringard | or something similar, rather than relying on environment vars | 20:55 |
| Ryan_Lane | I actually use admin token for this specific reason | 20:55 |
| ayoung | Well since there is no ref-integrity between the token and the trust, I guess I should put it in its own back end. It is the most work, but probably the right solution | 20:56 |
| Ryan_Lane | yeah. I think in the long run it's going to be easier | 20:56 |
| Ryan_Lane | it's harder short-term, for sure | 20:56 |
| *** alobbs has joined #openstack-dev | 20:58 | |
| *** markwash has joined #openstack-dev | 21:01 | |
| *** dolphm has joined #openstack-dev | 21:02 | |
| *** dingus91 has joined #openstack-dev | 21:03 | |
| kbringard | rmk: it sounds like you're familiar with the snapshot process? | 21:04 |
| *** dingus9 has quit IRC | 21:06 | |
| *** dingus91 has quit IRC | 21:06 | |
| *** cp16net is now known as cp16net|away | 21:06 | |
| *** dingus9 has joined #openstack-dev | 21:06 | |
| *** dingus9 has quit IRC | 21:09 | |
| *** maurosr has joined #openstack-dev | 21:09 | |
| *** dingus9 has joined #openstack-dev | 21:09 | |
| *** koolhead17 has quit IRC | 21:09 | |
| zykes- | danwent: around for a pm ? | 21:09 |
| *** dingus9 has quit IRC | 21:10 | |
| *** dingus9 has joined #openstack-dev | 21:10 | |
| openstackgerrit | A change was merged to openstack/nova: Update migration so it supports PostgreSQL. https://review.openstack.org/17073 | 21:11 |
| *** jaustinpage|want has quit IRC | 21:12 | |
| *** dingus91 has joined #openstack-dev | 21:12 | |
| *** andrewbogott_afk is now known as andrewbogott | 21:12 | |
| *** blamar_ has joined #openstack-dev | 21:14 | |
| *** sayjay has joined #openstack-dev | 21:14 | |
| *** dingus92 has joined #openstack-dev | 21:15 | |
| *** dingus92 has quit IRC | 21:15 | |
| *** dingus9 has quit IRC | 21:15 | |
| *** epim_ has joined #openstack-dev | 21:15 | |
| *** dingus9 has joined #openstack-dev | 21:15 | |
| ayoung | Ryan_Lane, dolphm Ok, I think I know the next step on the token refactoring. It ties in with my musings about trusts as well. We need to make the Token API aware of the identity API, pass it in at construction time, so that it can validate what it needs to when creating tokens. That is why service.py has grown out of proportion: it is trying to do things that belong in the token API, but the token couldn't do those things be | 21:16 |
| ayoung | cause it didn't have access to the identity API. Make sense? | 21:16 |
| *** dingus9 has quit IRC | 21:16 | |
| *** brucer_out has joined #openstack-dev | 21:16 | |
| Ryan_Lane | yeah | 21:16 |
| Ryan_Lane | that makes sense | 21:16 |
| *** jamey has joined #openstack-dev | 21:17 | |
| *** dingus91 has quit IRC | 21:17 | |
| *** mikal_ has joined #openstack-dev | 21:18 | |
| ayoung | I mean, we could pass the identity API as a parameter to the call if we needed to make it stateless, but that would be dumb. | 21:18 |
| *** dprince has quit IRC | 21:18 | |
| *** tmishler__ has joined #openstack-dev | 21:18 | |
| *** sandywalsh has quit IRC | 21:18 | |
| *** tmishler__ has quit IRC | 21:19 | |
| *** stevebake has quit IRC | 21:19 | |
| *** tmishler has quit IRC | 21:19 | |
| *** cp16net|away is now known as cp16net | 21:20 | |
| *** epim has quit IRC | 21:22 | |
| *** brucer_afk has quit IRC | 21:22 | |
| *** blamar has quit IRC | 21:22 | |
| *** mikal has quit IRC | 21:22 | |
| *** jmeredit has quit IRC | 21:22 | |
| *** blamar_ is now known as blamar | 21:22 | |
| *** epim_ is now known as epim | 21:22 | |
| *** tmishler has joined #openstack-dev | 21:23 | |
| *** jtran_ has joined #openstack-dev | 21:23 | |
| *** tmishler has quit IRC | 21:24 | |
| afazekas | fungi: is the allow_versions = true configured for tempest http://docs.openstack.org/developer/swift/overview_object_versioning.html ? | 21:24 |
| ayoung | Ryan_Lane, OK, so why would adding "from keystone import identity" blow up inside keystone/token/core.py ? | 21:26 |
| ayoung | ImportError: cannot import name identity | 21:26 |
| *** lifeless_ has joined #openstack-dev | 21:27 | |
| fungi | afazekas: i'm not sure. i don't know a lot about tempest yet, but i'll take a look and see if i can find that out | 21:27 |
| Ryan_Lane | hm. | 21:27 |
| dolphm | ayoung: probably a circular import | 21:27 |
| dolphm | ayoung: extend the v3 controller class and you'll get self.identity_api for free | 21:28 |
| ayoung | dolphm, doesn't look like it. identity know nothing about tokens | 21:28 |
| *** mikal has joined #openstack-dev | 21:28 | |
| ayoung | dolphm, I don't want to do this at the controller level | 21:28 |
| ayoung | I want to to it at the manager | 21:28 |
| *** Ng_ has joined #openstack-dev | 21:28 | |
| fungi | afazekas: i don't find any references to allow_versions in any of tempest, devstack or devstack-gate repositories currently | 21:29 |
| ayoung | dolphm, nope, you are right | 21:29 |
| ayoung | core.py:759: self.token_api.revoke_tokens(context, user_id, tenant_id) | 21:30 |
| ayoung | hrm | 21:30 |
| afazekas | fungi: It is a swift configuration option normali it is in the container-server.conf [app:container-server] section | 21:30 |
| *** tmishler has joined #openstack-dev | 21:30 | |
| clarkb | fungi: I think devstack would be setting that option | 21:31 |
| clarkb | or it would be the thing to do that if it was happening | 21:31 |
| dolphm | ayoung: oh yeah... so, why the manager and not the controller? | 21:31 |
| *** tmishler has quit IRC | 21:32 | |
| fungi | afazekas: clarkb i also don't see any reference to a container-server.conf in tempest, devstack or devstacg-gate | 21:32 |
| *** mikal_ has quit IRC | 21:33 | |
| *** krtaylor has quit IRC | 21:33 | |
| *** jtran has quit IRC | 21:33 | |
| *** mtreinish has quit IRC | 21:33 | |
| *** pabelanger has quit IRC | 21:33 | |
| *** roz has quit IRC | 21:33 | |
| *** lifeless has quit IRC | 21:33 | |
| *** Ng has quit IRC | 21:33 | |
| *** jtran_ is now known as jtran | 21:33 | |
| *** lifeless_ is now known as lifeless | 21:33 | |
| *** roz has joined #openstack-dev | 21:33 | |
| *** openstack_newbe has quit IRC | 21:33 | |
| *** mtreinish has joined #openstack-dev | 21:35 | |
| ayoung | dolphm, this is logic that is not HTTP specific, but rather logic that is common to anything creating tokens or creating trusts. I want to validate referential integrity. If I do that in the controller, why even bother having a manager? | 21:35 |
| *** jpich has quit IRC | 21:35 | |
| fungi | afazekas: clarkb: earlier today jaypipes noted that he wasn't even sure swift is being exercised in there and was going to find out | 21:35 |
| fungi | if it isn't, i suspect there's a lot more to be added than just that one setting | 21:35 |
| ayoung | I mean, we are already seeing that the create token logic fragmented between the admin apis, and all of the contrib code. shouldn't "create token" be done as a protocol neutral business logic piece? | 21:36 |
| afazekas | fungi: in the stack.sh I see swift install steps | 21:36 |
| ayoung | the same way that revoke tokens is done? | 21:36 |
| fungi | afazekas: do you know if swift creates a container-server.conf when it installs in that case? | 21:36 |
| ayoung | dolphm, ah, looking at what you are doing, I might have been mischaracterizing the controller | 21:37 |
| dolphm | ayoung: well, the managers wrap a specific driver... i think we're missing a layer of abstraction between controllers and managers, since controllers are overloaded with http+business | 21:37 |
| afazekas | fungi: generate_swift_configuration looks like does it | 21:38 |
| *** alop has joined #openstack-dev | 21:38 | |
| ayoung | dolphm, yeah, or we have the managers doing too much work that belongs in the drivers | 21:38 |
| dolphm | ayoung: (managers aren't doing much today, except wrapping exceptions, no?) | 21:39 |
| *** sacharya has quit IRC | 21:39 | |
| *** pabelanger has joined #openstack-dev | 21:39 | |
| *** dingus9 has joined #openstack-dev | 21:40 | |
| *** krtaylor has joined #openstack-dev | 21:40 | |
| ayoung | dolphm, the sql one is doing a lot. probably too much. Look at the identity/backends/sql.py file | 21:40 |
| fungi | afazekas: looks like that's doing something with a swift.conf instead of container-server.conf (unless i'm misreading) | 21:40 |
| dolphm | ayoung: maybe we're referring to different things ... that's the "driver" impl itself to me, the "manager" is in identity/core.py | 21:41 |
| *** mtreinish is now known as 65MABJ9VG | 21:42 | |
| *** mtreinish has joined #openstack-dev | 21:42 | |
| *** AlanClark has joined #openstack-dev | 21:42 | |
| zodiak | given an instance_uuid (or name label), what's the "approved" way to get a vdi back from cinder ? this has to be called from inside the xenapi, so I would rather avoid having to do rpc/http | 21:42 |
| ayoung | in identity core, there are Driver and Manager classes defined, that are both abstract base classes. Both are implemented in backends/sql.py | 21:42 |
| fungi | afazekas: oh, i see. ${server_type} is getting filled to "container" | 21:42 |
| ayoung | Ah, disregard | 21:43 |
| *** annegentle_ has joined #openstack-dev | 21:43 | |
| fungi | afazekas: so probably needs another iniset call in there to add that option | 21:43 |
| ayoung | I was confusing myself between driver and the calsses for the individual files | 21:43 |
| ayoung | classes | 21:43 |
| dolphm | ayoung: Manager is concrete, it's just dynamic | 21:43 |
| ayoung | right | 21:43 |
| *** aeperezt has quit IRC | 21:44 | |
| afazekas | fungi: looks like it is the per node config, IMHO there is a global conf too | 21:44 |
| *** mtreinish has quit IRC | 21:44 | |
| ayoung | dolphm, still, I'd still have to deal with the circular dependency problem the way we have the code structured | 21:45 |
| dolphm | ayoung: are you writing this on v3 or v2? | 21:45 |
| ayoung | since modifying user deletes all tokens, and creating a token needs to verify the users | 21:45 |
| ayoung | this is trusts, so v3, but I would love to continue to clean up v2 without breaking things | 21:46 |
| fungi | afazekas: so anyway, to answer your earlier question, no it doesn't look like swift's allow_versions option is being set anywhere in devstack that i can find, unless it's already present in swift's swift.conf-sample or something | 21:46 |
| fungi | i'm cloning a copy of swift now to take a look at that | 21:47 |
| fungi | looks like etc/container-server.conf-sample has it commented out with the implication that it defaults to False | 21:47 |
| *** stevebake has joined #openstack-dev | 21:49 | |
| *** colinmcnamara has joined #openstack-dev | 21:49 | |
| afazekas | fungi: note the "True" is not a true value as document says , the "true" is true | 21:49 |
| *** aeperezt has joined #openstack-dev | 21:49 | |
| fungi | so when devstack's stack.sh does "generate_swift_configuration container 6011 2" the container is passed into generate_swift_configuration()'s server_type causing container-server.conf-sample to get copied for the node config and modified | 21:50 |
| dolphm | ayoung: if you extend keystone.common.controller.V3Controller you shouldn't have any circular dependencies unless you speficically call something recursively | 21:51 |
| *** jcooley_afk is now known as jcooley | 21:51 | |
| fungi | afazekas: and i'm *guessing* an iniset in there could set that option for each node's container-server.conf that way | 21:51 |
| ayoung | dolphm, the problem is the v2 stuff is in the file, though | 21:51 |
| *** 65MABJ9VG is now known as mtreinish | 21:51 | |
| *** colinmcnamara1 has joined #openstack-dev | 21:52 | |
| dolphm | ayoung: like what? (and which file?) | 21:52 |
| clarkb | vishy: I think we sorted it out. But we did it with a shotgun method and did the monkey patching with the fixtures lib | 21:52 |
| clarkb | that way things get cleaned up properly each time | 21:52 |
| fungi | afazekas: something like... iniset ${swift_node_config} DEFAULT allow_versions true | 21:52 |
| vishy | clarkb: nice | 21:53 |
| *** colinmcnamara has quit IRC | 21:53 | |
| *** adjohn has joined #openstack-dev | 21:54 | |
| ayoung | dolphm, if I add "from keystone import identity" it blows up inside keystone/token/core.py | 21:54 |
| ayoung | as identity already has from keystone import token | 21:55 |
| *** colinmcnamara has joined #openstack-dev | 21:55 | |
| dolphm | ayoung: where do you need to reference identity in keystone/token/core.py? | 21:55 |
| ayoung | dolphm, I was looking to see, first off, if I could move the guts of the authenticate method in service.py to the token api, as that is where it belongs | 21:55 |
| *** radez is now known as radez_g0n3 | 21:56 | |
| ayoung | bascially, to create the token, we need to collect and validate data from the identity api. DOing it in service.api is not really the right place. | 21:56 |
| *** colinmcnamara1 has quit IRC | 21:56 | |
| fungi | afazekas: you might want to double-check that assertion with one of the devstack developers | 21:57 |
| ayoung | dolphm, I don't think the controller code should be in the same file as the driver defintions. what about identity/controllers.py and token/controllers.py instead? | 21:57 |
| afazekas | fungi: probably yes, but with the app:container-server section. I check how the swift reads the config files | 21:58 |
| *** woorea has quit IRC | 21:58 | |
| *** colinmcnamara has quit IRC | 21:58 | |
| dolphm | ayoung: ah, i see -- what's left of v2 service.py would have to use the same kind of dependency injection as V3Controller then | 21:58 |
| dolphm | ayoung: i'm down with that | 21:58 |
| *** colinmcnamara has joined #openstack-dev | 21:59 | |
| *** otherwiseguy has quit IRC | 21:59 | |
| *** dingus9 has quit IRC | 21:59 | |
| ayoung | dolphm, cool. | 22:00 |
| fungi | afazekas: ahh, right. then maybe... iniset ${swift_node_config} app:${server_type}-server allow_versions true | 22:01 |
| fungi | or something similar | 22:01 |
| *** boden has quit IRC | 22:04 | |
| *** colinmcnamara has quit IRC | 22:04 | |
| afazekas | fungi: it is just a container options, not for all server type | 22:04 |
| afazekas | fungi: I'll install new system, and check how the per node and global confs working | 22:05 |
| fungi | afazekas: gotcha, so maybe a separate conditional matching on ${server_type} to wrap that particular iniset i guess | 22:05 |
| *** danger_fo has joined #openstack-dev | 22:06 | |
| *** alop has quit IRC | 22:08 | |
| afazekas | fungi: so probably I just need send the modifications to the devstack gerrit and hope an approval | 22:10 |
| fungi | afazekas: yes, but feel free to add me as a reviewer on it and i'll provide what input i can muster | 22:10 |
| *** datsun180b has joined #openstack-dev | 22:10 | |
| afazekas | fungi: ok, thank you | 22:11 |
| fungi | afazekas: also i added you to the https://review.openstack.org/17062 fix we were discussing earlier if you have time to look at it | 22:12 |
| *** eglynn is now known as sleepy-eglynn | 22:12 | |
| *** colinmcnamara has joined #openstack-dev | 22:13 | |
| *** AlanClark has quit IRC | 22:13 | |
| *** e1mer has joined #openstack-dev | 22:13 | |
| *** martine has quit IRC | 22:13 | |
| *** colinmcnamara1 has joined #openstack-dev | 22:16 | |
| *** colinmcnamara has quit IRC | 22:17 | |
| *** dolphm has quit IRC | 22:17 | |
| *** donaldh has joined #openstack-dev | 22:20 | |
| *** vipul is now known as vipul|away | 22:20 | |
| *** vipul|away is now known as vipul | 22:21 | |
| openstackgerrit | A change was merged to openstack/nova: Allow multi_host compute nodes to share dhcp ip https://review.openstack.org/16578 | 22:21 |
| *** colinmcnamara1 has quit IRC | 22:21 | |
| *** networkstatic has joined #openstack-dev | 22:21 | |
| Madkiss | i think i'm having a quantum bug. | 22:22 |
| Madkiss | whatever I do on the compute-node, the ovs l2 agent is just not going to create any network namespaces. | 22:23 |
| *** colinmcnamara has joined #openstack-dev | 22:24 | |
| *** markvoelker has quit IRC | 22:25 | |
| *** sacharya has joined #openstack-dev | 22:25 | |
| *** alobbs has quit IRC | 22:27 | |
| ayoung | dolphm, hows this https://review.openstack.org/#/c/17089/ | 22:28 |
| *** flaper87 has joined #openstack-dev | 22:29 | |
| *** flaper87 has joined #openstack-dev | 22:29 | |
| *** colinmcnamara has quit IRC | 22:31 | |
| *** colinmcnamara has joined #openstack-dev | 22:32 | |
| *** pmyers has quit IRC | 22:35 | |
| *** rpedde is now known as rpedde_away | 22:35 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Fix blocking test for nova notifier https://review.openstack.org/17074 | 22:39 |
| *** pabelanger has quit IRC | 22:41 | |
| *** colinmcnamara has quit IRC | 22:42 | |
| openstackgerrit | A change was merged to openstack/quantum: Correct i18n message for api and db module https://review.openstack.org/17035 | 22:44 |
| openstackgerrit | A change was merged to openstack/quantum: Use openstack.common.logging in NEC OpenFlow plugin https://review.openstack.org/17053 | 22:44 |
| *** davidf_76 has joined #openstack-dev | 22:45 | |
| *** colinmcnamara has joined #openstack-dev | 22:45 | |
| *** jcooley is now known as jcooley_afk | 22:47 | |
| *** jcooley_afk is now known as jcooley | 22:48 | |
| *** brucer_out has quit IRC | 22:49 | |
| *** joe5081 has quit IRC | 22:49 | |
| *** brucer has joined #openstack-dev | 22:51 | |
| alexpilotti | sdague: ping | 22:52 |
| jk0 | C | 22:53 |
| jk0 | er | 22:53 |
| *** colinmcnamara1 has joined #openstack-dev | 22:53 | |
| *** ayoung has quit IRC | 22:53 | |
| *** jab416171|Cloud has joined #openstack-dev | 22:54 | |
| jab416171|Cloud | I just git clone'd and ran the init script for devstack, and entered the same password for everything. now how do I log in to the web interface? | 22:54 |
| *** eharney has quit IRC | 22:54 | |
| openstackgerrit | A change was merged to openstack/nova: Verify doc/api_samples files along with the templates https://review.openstack.org/16991 | 22:54 |
| openstackgerrit | A change was merged to openstack/nova: Include 'hosts' and 'metadetails' in aggregate. https://review.openstack.org/17009 | 22:55 |
| *** colinmcnamara has quit IRC | 22:55 | |
| clarkb | jab416171|Cloud: I think the users created are admin and demo | 22:55 |
| annegentle_ | jab416171|Cloud: at the completion of the devstack script, it should have returned a listing starting with "Horizon is now available at http://nnn.nnn.nnn.nnn/" with the IP address of where you can log into the Dashboard. Horizon = Dashboard. | 22:55 |
| clarkb | they should use the password you entered | 22:56 |
| alexpilotti | rmk: ping | 22:56 |
| annegentle_ | jab416171|Cloud: it also returns a line with "The password: passwordyouentered" | 22:56 |
| alexpilotti | russellb: ping | 22:57 |
| russellb | alexpilotti: pong | 22:57 |
| alexpilotti | russellb: hi thanks for the review! | 22:57 |
| russellb | np | 22:57 |
| alexpilotti | russellb: I wanted to discuss sdague comment on this review: https://review.openstack.org/#/c/16843/ | 22:57 |
| jab416171|Cloud | annegentle_: I don't think I saw that line | 22:58 |
| jab416171|Cloud | annegentle_: I can hit the log in interface, but admin/mypass doesn't work | 22:58 |
| alexpilotti | russellb: just wanted to make sure that you guys are ok with the fact that the Hyper-V tests don't have anything to do with interface changes :-) | 22:58 |
| russellb | alexpilotti: sounds like he may just misunderstand how the tests work then | 22:59 |
| russellb | as long as tests can be fixed without having to mess wwith that stuff i think we're ok | 22:59 |
| russellb | so just grab sdague when you can, because that was a misunderstanding on our part | 22:59 |
| alexpilotti | russellb: yep, do you think we should bring this up tomorrow at the Nova meeting? | 22:59 |
| russellb | sure, worth noting | 23:00 |
| russellb | will you be there? | 23:00 |
| alexpilotti | russellb: sure, I plan to attend regularly | 23:00 |
| russellb | ok, i'll put it on the agenda | 23:00 |
| alexpilotti | russellb: great tx! | 23:01 |
| russellb | done | 23:01 |
| *** colinmcnamara1 has quit IRC | 23:01 | |
| *** andrewbogott is now known as andrewbogott_afk | 23:01 | |
| alexpilotti | russellb: just in case you'd like to review this one as well… https://review.openstack.org/#/c/16843/ ;-) | 23:01 |
| *** colinmcnamara has joined #openstack-dev | 23:01 | |
| dansmith | alexpilotti: I was about to get that one | 23:01 |
| *** stevebake has quit IRC | 23:02 | |
| alexpilotti | dansmith: tx! :-) | 23:02 |
| *** datsun180b has quit IRC | 23:02 | |
| alexpilotti | dansmith: let me know if you have any question, as the committer is already asleep ;-) | 23:02 |
| *** jcooley is now known as jcooley_afk | 23:03 | |
| *** woodspa has quit IRC | 23:03 | |
| jab416171|Cloud | annegentle_: while setting up, it prompted me for like 5 passwords, so I entered a password for them all. Now it's done starting up and I can hit the login page, but I can't log in with admin/thatpassword | 23:03 |
| annegentle_ | jab416171|Cloud: I wonder if the dashboard install failed? You can get to the log by attaching to screen via screen -x as the user that ran the devstack script. | 23:04 |
| annegentle_ | jab416171|Cloud: the default users are admin and demo, did you try just admin? | 23:04 |
| jab416171|Cloud | yes | 23:04 |
| jab416171|Cloud | I'm logged in to the box I ran the script from | 23:05 |
| *** annegentle_ has quit IRC | 23:05 | |
| *** ahardy_ has quit IRC | 23:05 | |
| clarkb | jab416171|Cloud: if you run screen -x it will attach you to the screen session with all of the log output for the various services. you can then switch to the window for horizon | 23:06 |
| jab416171|Cloud | screen -x outputs there is no screen to be attached | 23:06 |
| *** jcooley_afk is now known as jcooley | 23:07 | |
| *** stevebake has joined #openstack-dev | 23:07 | |
| clarkb | does ps show screen? you may be a different user than the user running screen | 23:07 |
| *** alunduil has quit IRC | 23:08 | |
| *** ahardy has joined #openstack-dev | 23:09 | |
| *** colinmcnamara has quit IRC | 23:09 | |
| sdague | alexpilotti: sorry, here now | 23:09 |
| alexpilotti | sdague: hi! | 23:09 |
| sdague | alexpilotti: ok, so it's clear I had a missunderstanding from reading the backlog | 23:09 |
| *** colinmcnamara has joined #openstack-dev | 23:09 | |
| alexpilotti | sdague: np :-) | 23:09 |
| sdague | alexpilotti: do you understand the issue that was getting run into by folks? | 23:09 |
| sdague | where it wasn't possible to change the hyperv driver to pass it's tests when a virt interface needed to change | 23:10 |
| alexpilotti | sdague: did you see the commit: https://review.openstack.org/#/c/17090/ ? | 23:10 |
| *** kbringard has quit IRC | 23:11 | |
| openstackgerrit | A change was merged to openstack/swift: Fix bug with swob.Request.path_info_pop https://review.openstack.org/17020 | 23:11 |
| alexpilotti | sdague: IMO the fact that we are passing a callable to a utility method used by all the state change methods might have misleaded them | 23:11 |
| alexpilotti | sdague: I'd like to make sure that the Hyper-V tests don't affect any interface change | 23:12 |
| dansmith | alexpilotti: that's not really possible | 23:12 |
| sdague | alexpilotti: ok, so that solves this one issue, but it does raise a problem for how can the hyperv code be updated by someone else on an interface face change | 23:12 |
| *** mtreinish has quit IRC | 23:13 | |
| dansmith | although, sdague, it really should only affect folks if they are changing the order of the calls to WMI stuff, I think, | 23:13 |
| dansmith | which is definitely the minority of the situations | 23:13 |
| alexpilotti | sdague dansmith: it's only affecting situations in which you have to change the internal behavior of the driver | 23:14 |
| dansmith | right | 23:14 |
| vishy | smoser: ping | 23:14 |
| *** colinmcnamara has quit IRC | 23:14 | |
| dansmith | alexpilotti: which is sometimes going to have to be the case, I'm sure, in order to make a change that affects all of the drivers | 23:14 |
| alexpilotti | sdague dansmith: but do you think it's realistic that somebody will change that code without running it? | 23:14 |
| *** vipul is now known as vipul|away | 23:14 | |
| *** vipul|away is now known as vipul | 23:14 | |
| dansmith | alexpilotti: I change xenapi code constantly without running it | 23:14 |
| dansmith | alexpilotti: and I've changed yours (a little) | 23:14 |
| *** jchiles has quit IRC | 23:14 | |
| alexpilotti | dansmith: did you run into issues with the tests when you changed it? | 23:15 |
| dansmith | alexpilotti: no, and like I said, most of those changes won't be problematic | 23:15 |
| dansmith | alexpilotti: but they could be | 23:15 |
| *** dhellmann is now known as dhellmann-afk | 23:15 | |
| alexpilotti | dansmith: I agree 100% on the fact that we need to make those stubs manually editable | 23:15 |
| dansmith | alexpilotti: I think I'd be a lot happier with a FakeWMI implementation that could behave just like the real one, instead of just replaying static captures | 23:16 |
| dansmith | the xenapi driver goes to a lot of effort to do that, as does the libvirt one | 23:16 |
| alexpilotti | dansmith: we would be way happier as well with a FakeWMI driver | 23:16 |
| dansmith | just to enable testing | 23:16 |
| *** colinmcnamara has joined #openstack-dev | 23:17 | |
| sdague | alexpilotti: is that something you guys can prioritize? because I expect this is going to cause a lot of grief in the future. | 23:17 |
| alexpilotti | dansmith: but the WMI architecture is radically different from xenapi or libvirt | 23:17 |
| dansmith | alexpilotti: in what way (that makes it not doable) ? | 23:17 |
| alexpilotti | dansmith: are you familiar with WMI? | 23:17 |
| dansmith | alexpilotti: you know I'm not :D | 23:18 |
| alexpilotti | dansmith: ok I know that you think it's ickyer than Xen, but maybe you read about it, etc :-D | 23:18 |
| dansmith | alexpilotti: I said "windows" .. I really don't know anything about WMI, other than I've seen you guys issue quasi-SQL statements and CIM-like calls | 23:19 |
| alexpilotti | dansmith: correct. WMI is MS's take at CIM | 23:19 |
| sdague | alexpilotti: be forwarned, dansmith does know quite a bit about CIM :) | 23:19 |
| dansmith | muahahaha | 23:19 |
| * sdague throws dansmith under a bus | 23:19 | |
| alexpilotti | dansmith: loool | 23:19 |
| *** cyeoh has quit IRC | 23:20 | |
| *** cyeoh has joined #openstack-dev | 23:20 | |
| alexpilotti | dansmith: ok, so WMI is the icky version of CIM :-D | 23:20 |
| dansmith | alexpilotti: CIM is pretty icky already, so, yeah.. | 23:20 |
| alexpilotti | dansmith: ok, so you already got my point | 23:20 |
| dansmith | alexpilotti: point me at something in the code that would be hard to do Fake | 23:20 |
| alexpilotti | dansmith: pick a random line (kidding) | 23:20 |
| alexpilotti | dansmith: 1 sec | 23:20 |
| dansmith | alexpilotti: you said it, not me :) | 23:20 |
| *** colinmcnamara has quit IRC | 23:21 | |
| dansmith | https://github.com/openstack/nova/blob/master/nova/virt/hyperv/vmops.py#L71 | 23:21 |
| dansmith | alexpilotti: how about that? | 23:21 |
| dansmith | yes, you need to have a Fake_Msvm_ComputerSystem object, and list_instances() needs to return a few fake ones | 23:22 |
| dansmith | (just like xenapi does) | 23:22 |
| alexpilotti | dansmith: I was looking at the _create_vm code 224 | 23:22 |
| *** ryant has joined #openstack-dev | 23:23 | |
| dansmith | alexpilotti: right, so you have to create something that behaves just like self._conn here | 23:23 |
| alexpilotti | the one you pointed pout can be asily faked | 23:23 |
| *** colinmcnamara has joined #openstack-dev | 23:23 | |
| *** afazekas has quit IRC | 23:23 | |
| alexpilotti | you have to create a fake, that handles method calls, queries, and associators | 23:23 |
| dansmith | alexpilotti: yep | 23:24 |
| alexpilotti | it becomes so complicated that it will require a huge test suite by itself ;-) | 23:24 |
| dansmith | alexpilotti: so maybe xenapi doesn't have the same level of insane complexity in the interface that it's mocking, | 23:24 |
| dansmith | alexpilotti: but that's because they chose a sane interface to begin with :) | 23:24 |
| dansmith | alexpilotti: yes, and there are tests in nova to test even the fake stuff :) | 23:25 |
| alexpilotti | dansmith: I know ;-) | 23:25 |
| alexpilotti | dansmith: let's put things into context | 23:25 |
| *** woorea has joined #openstack-dev | 23:25 | |
| alexpilotti | dansmith: we got called like on the 1st of august, to do everything for the 13th of August | 23:25 |
| dansmith | alexpilotti: so I think that the point is that the hyperv driver side-steps that (admittedly difficult) job because it's "hard" and "weird" and also because most people don't care about it as long as it doesn't get in the way | 23:25 |
| alexpilotti | dansmith: that sincerely, stinked quite a bit | 23:26 |
| dansmith | alexpilotti: no doubt :) | 23:26 |
| alexpilotti | dansmith: my idea was to replace all those direct WMI calls with a good design | 23:26 |
| alexpilotti | dansmith: meaning a class that wraps those lower level calls and that can be easily mocked | 23:27 |
| dansmith | alexpilotti: I think that providing an abstraction layer above the WMI to make it sane, and then providing a faked object with the same abstraction, | 23:27 |
| dansmith | would allow everything you need, plus you could test the faked object to make sure it behaves | 23:27 |
| alexpilotti | dansmith: yep, looks like we are talking about the same thing ;-) | 23:27 |
| dansmith | alexpilotti: and I think what sdague is saying is he wants to see that sooner than later :) | 23:27 |
| *** dolphm has joined #openstack-dev | 23:27 | |
| *** colinmcnamara has quit IRC | 23:28 | |
| alexpilotti | dansmith: we'll have to redo all that code anyway, because ahem it might not be supported in the future | 23:28 |
| jab416171|Cloud | clarkb: no | 23:29 |
| jab416171|Cloud | clarkb: ps does not list screen | 23:29 |
| alexpilotti | dansmith: so my plea :-) is to let us redo all that code in a timely matter without throwing our scheduling to /dev/null | 23:29 |
| *** vipul is now known as vipul|away | 23:29 | |
| dansmith | alexpilotti: a unix reference, from a windows guy? :) | 23:30 |
| sdague | alexpilotti: ok, maybe if you shared your schedule for that, it would be helpful :) | 23:30 |
| alexpilotti | dansmith: I have probably as much *NIX in my veins than Windows ;-) | 23:30 |
| dansmith | alexpilotti: I'm not as dead-set on it as sdague, so that's fine with me | 23:30 |
| dansmith | alexpilotti: as long as it has a timetable, and maybe a blueprint :) | 23:30 |
| alexpilotti | dansmith: sure | 23:31 |
| sdague | my concern is that it wasn't a priority, so now we had a virt driver in the tree that no one except the orig authors could maintain, which I think is bad for the project | 23:31 |
| dansmith | alexpilotti: but it's all tainted with the windows blood! :) | 23:31 |
| clarkb | jab416171|Cloud: what was output at the end of stack.sh? did it look successful? | 23:31 |
| sdague | it's good that people besides citrix can and do fix the xenapi code | 23:31 |
| alexpilotti | dansmith: are you suggesting that my blood is icky? :-D | 23:31 |
| * dansmith whistles innocently | 23:31 | |
| alexpilotti | sdague: I agree 100% | 23:31 |
| *** mrodden has quit IRC | 23:31 | |
| *** dolphm has quit IRC | 23:32 | |
| dansmith | alexpilotti: as someone who actually _has_ modified your driver, I'm less concerned about it | 23:32 |
| alexpilotti | dansmith: you'd be surprised about how neat is Windows under the hood ;-) | 23:32 |
| dansmith | alexpilotti: but the potential for a problem is there | 23:32 |
| alexpilotti | dansmith sdague: sure | 23:32 |
| alexpilotti | dansmith sdague: the testing infrastrucure has some advantages as well | 23:33 |
| sdague | alexpilotti: so I was probably out of line for calling you guys out on an unrelated review, but this did drive a bunch of frustration and grumbling in the #openstack-nova channel the other night, so I wanted to make sure it got addressed in some format | 23:33 |
| dansmith | sdague: the right way, IMHO, is to bring it up on the nova meeting, | 23:33 |
| dansmith | which is apparently going to happen | 23:33 |
| sdague | right, yes, that's a good call | 23:33 |
| alexpilotti | sdague: I already spoke with russellb asking about it | 23:33 |
| sdague | alexpilotti: yes, great | 23:33 |
| alexpilotti | sdague: I'm sorry I didn't attend your discussions in #openstack-nova | 23:34 |
| sdague | alexpilotti: hope I didn't bruise any feelings too badly in the -1. | 23:34 |
| sdague | alexpilotti: it wasn't actually me that was running into the issues | 23:34 |
| alexpilotti | sdague: no way! | 23:34 |
| dansmith | oh, speaking of bruising... | 23:34 |
| russellb | group hug! | 23:34 |
| sdague | but I did share the concern | 23:34 |
| * dansmith slaps sdague across the mouth for outing his CIM background | 23:34 | |
| alexpilotti | sdague: lol | 23:34 |
| alexpilotti | sdague: I just took the occasion for discuss this as openly as possible | 23:35 |
| *** colinmcnamara has joined #openstack-dev | 23:35 | |
| * sdague knows of other buses I can throw dansmith under | 23:35 | |
| *** flaper87 has quit IRC | 23:35 | |
| dansmith | sdague: dude, not cool | 23:35 |
| sdague | alexpilotti: excellent | 23:35 |
| sdague | alexpilotti: yeh, I think this will make a good nova meeting topic | 23:36 |
| russellb | alexpilotti is going to educate us on how it works and how they plan to improve it :- | 23:36 |
| *** colinmcnamara1 has joined #openstack-dev | 23:36 | |
| sdague | sounds great | 23:36 |
| russellb | .... right alexpilotti ?! :) | 23:36 |
| alexpilotti | sdague: I just joinen #openstack-nova as well, pardon me, I just thought you guys were discussing thi stopics here | 23:36 |
| dansmith | alexpilotti: this is a fine place for it | 23:36 |
| russellb | either one | 23:36 |
| sdague | yeh, either one is good | 23:37 |
| sdague | it happened last week some time | 23:37 |
| russellb | sometimes this channel gets noisy, so we end up talking nova stuff over there *shrug* | 23:37 |
| alexpilotti | russellb: sure, I will bring a friend called hypnotoad to help me with it :-D | 23:37 |
| sdague | when the patch with the skip tests went in as a last resort | 23:37 |
| sdague | all hail the hypnotoad.... | 23:37 |
| sdague | ok, it's time for me to head out for the night, catch you all later | 23:38 |
| alexpilotti | sdague: now altogether: all hail hyper-v testing :-D | 23:38 |
| *** colinmcnamara2 has joined #openstack-dev | 23:38 | |
| sdague | :P | 23:39 |
| *** colinmcnamara has quit IRC | 23:39 | |
| sdague | hehehe | 23:39 |
| *** krtaylor has quit IRC | 23:39 | |
| *** vipul|away is now known as vipul | 23:39 | |
| *** donaldh has quit IRC | 23:40 | |
| *** colinmcnamara1 has quit IRC | 23:40 | |
| alexpilotti | sdague: if we got it clear, any chance you can remove that -1 from the review? ;-) | 23:41 |
| openstackgerrit | A change was merged to openstack/nova: Fixes HyperV compute "resume" tests https://review.openstack.org/17090 | 23:41 |
| alexpilotti | dansmith: tx for the review BTW! | 23:42 |
| *** gongysh has joined #openstack-dev | 23:42 | |
| dansmith | alexpilotti: np | 23:42 |
| alexpilotti | dansmith: talking about the tests | 23:43 |
| alexpilotti | dansmith: what about the idea of generating those manally editable stubs? | 23:43 |
| alexpilotti | dansmith: would it be a working idea for you? | 23:43 |
| dansmith | alexpilotti: not sure what you mean | 23:43 |
| alexpilotti | dansmith: I mean now the stubs are serialized in pickled format | 23:44 |
| alexpilotti | dansmith: what about having them in a "clear" json? | 23:44 |
| dansmith | alexpilotti: I think that would likely be better, | 23:44 |
| dansmith | alexpilotti: and I think that for testing the actual implementation model, that's acceptable, | 23:45 |
| dansmith | but I'd like to see a real fake implementation of all the lower layers, | 23:45 |
| alexpilotti | dansmith: this is something we could do relatively quickly (it's already planned in the README.rst you reviewed in the TODO at the end) | 23:45 |
| dansmith | which can be used for testing the upper bits, and then use the pickled (however it is done) approach only for the bits that actually require platform libraries | 23:45 |
| dansmith | alexpilotti: right. It's not "enough", IMHO, but it would be an improvement | 23:46 |
| alexpilotti | dansmith: and we could introduce gradually a fake implementation without rewriting all the code overnight | 23:46 |
| alexpilotti | dansmith: what do you think? | 23:46 |
| dansmith | alexpilotti: yes, I don't think anyone is going to vote against that approach | 23:46 |
| dansmith | alexpilotti: as long as the end goal is a proper implementation | 23:46 |
| alexpilotti | dansmith: do you think that this approach for Grizzly and the full fake implementation for H would be ok? | 23:47 |
| dansmith | alexpilotti: I think you need to ask the folks in the nova meeting about that | 23:47 |
| dansmith | alexpilotti: you seem to be under the mistaken impression that my opinion means anything :D | 23:48 |
| alexpilotti | dansmith: sure, I'd like to know your opinion for now | 23:48 |
| dansmith | heh | 23:48 |
| dansmith | personally, I think that the current approach is better than nothing, and while I don't like it, I'm not overly annoyed by it | 23:48 |
| alexpilotti | dansmith: I'm not going to go to the meeting with: As my friend dansmith agreed on, we are going to, blah blah ;-) | 23:48 |
| dansmith | and I think that the assumption is that we just @skip() the whole thing any time it's not fixable | 23:49 |
| dansmith | alexpilotti: heh, good ;) | 23:49 |
| *** sthaha has joined #openstack-dev | 23:49 | |
| *** sthaha has quit IRC | 23:49 | |
| alexpilotti | dansmith: as you can imaging we are already working around the clock for the HUGE amount of stuff that is going to go into grizzly | 23:50 |
| dansmith | alexpilotti: aren't we all? :) | 23:50 |
| alexpilotti | dansmith: I guess so, otherwise we'd do a different job I guess ;-) | 23:51 |
| alexpilotti | dansmith: what I'm saying is that such a change would require such an amount of dev hours that would be quite… disrupting in teh scheduling | 23:51 |
| *** jcooley is now known as jcooley_afk | 23:51 | |
| alexpilotti | dansmith: which means delaying world domination of OpenStack in the Windows world MUHAHAHA | 23:52 |
| alexpilotti | dansmith: kidding, eh | 23:52 |
| dansmith | alexpilotti: I don't think anyone is worried about that happening :) | 23:52 |
| alexpilotti | dansmith: lol | 23:53 |
| dansmith | alexpilotti: I say, present your case and your schedule at the meeting and I'm sure you'll get acceptance or constructive feedback :) | 23:53 |
| alexpilotti | dansmith: I know 1 or 2 people wchi might be indeed ;-) | 23:53 |
| *** nunosantos has quit IRC | 23:53 | |
| *** jaypipes has quit IRC | 23:53 | |
| alexpilotti | dansmith: sounds good | 23:53 |
| dansmith | alexpilotti: oh, I read that backwards.. heh :) | 23:53 |
| vishy | jgriffith: ping | 23:54 |
| *** colinmcnamara2 has quit IRC | 23:55 | |
| *** jcooley_afk is now known as jcooley | 23:55 | |
| *** colinmcnamara has joined #openstack-dev | 23:56 | |
| jab416171|Cloud | clarkb: yeah, I didn't see any glaring error messages | 23:56 |
| jab416171|Cloud | clarkb: is there some log file I can look at? | 23:56 |
| clarkb | jab416171|Cloud: I think stuff should be sent to syslog | 23:57 |
| jab416171|Cloud | clarkb: when I try to log in, I get "An error occurred authenticating. Please try again later." | 23:59 |
| jab416171|Cloud | clarkb: and there's nothing in /var/log/syslog that's useful | 23:59 |
| *** rafaduran has left #openstack-dev | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!