*** rwsu has quit IRC | 00:11 | |
*** penick has joined #openstack-ironic | 00:19 | |
openstackgerrit | A change was merged to openstack/ironic: Use standard locale in list_partitions https://review.openstack.org/122414 | 00:25 |
---|---|---|
*** rwsu has joined #openstack-ironic | 00:27 | |
*** penick has quit IRC | 00:34 | |
*** derekh has joined #openstack-ironic | 00:36 | |
*** yjiang5 has quit IRC | 00:38 | |
*** derekh has quit IRC | 00:53 | |
*** Haomeng has joined #openstack-ironic | 01:03 | |
*** Haomeng|2 has quit IRC | 01:04 | |
*** Haomeng has quit IRC | 01:27 | |
*** Haomeng|2 has joined #openstack-ironic | 01:28 | |
*** Haomeng|2 has quit IRC | 01:29 | |
*** marcoemorais has quit IRC | 01:40 | |
*** Haomeng has joined #openstack-ironic | 01:47 | |
*** rwsu has quit IRC | 01:56 | |
*** rloo_afk has quit IRC | 02:04 | |
*** rwsu has joined #openstack-ironic | 02:09 | |
*** rwsu has quit IRC | 02:11 | |
*** Haomeng|2 has joined #openstack-ironic | 02:18 | |
*** Haomeng has quit IRC | 02:18 | |
Haomeng|2 | lazy_prince: fya - neutron /v2.0/ports api can get port id witch filter device_id=<VirtualServer_ID> | 02:25 |
Haomeng|2 | lazy_prince: my env can not boot vm now, so can not help you to very this neutron api call, but it should works | 02:26 |
*** rwsu has joined #openstack-ironic | 02:27 | |
*** harlowja is now known as harlowja_away | 02:29 | |
Haomeng|2 | lazy_prince: http://developer.openstack.org/api-ref-networking-v2.html | 02:34 |
Haomeng|2 | lazy_prince: hope this can help you:) | 02:34 |
*** ramineni has joined #openstack-ironic | 02:53 | |
*** rushiagr_away is now known as rushiagr | 03:23 | |
*** Poornima has joined #openstack-ironic | 03:24 | |
*** penick has joined #openstack-ironic | 03:47 | |
*** rushiagr is now known as rushiagr_away | 03:47 | |
*** pcrews has quit IRC | 03:50 | |
*** penick has quit IRC | 03:50 | |
*** pensu has joined #openstack-ironic | 03:57 | |
*** pcrews has joined #openstack-ironic | 04:01 | |
*** pcrews has quit IRC | 04:09 | |
*** nikunj2512 has joined #openstack-ironic | 04:11 | |
*** rameshg87 has joined #openstack-ironic | 04:15 | |
lazy_prince | Haomeng|2: thanks for the pointers.. i will look into it.. | 04:36 |
*** pensu has quit IRC | 04:55 | |
*** Poornima has quit IRC | 05:06 | |
*** rushiagr_away is now known as rushiagr | 05:06 | |
*** rakesh_hs has joined #openstack-ironic | 05:09 | |
*** Poornima has joined #openstack-ironic | 05:22 | |
*** yuanying has quit IRC | 05:24 | |
*** yuanying has joined #openstack-ironic | 05:25 | |
*** yuanying has quit IRC | 05:29 | |
*** pensu has joined #openstack-ironic | 05:41 | |
Haomeng|2 | lazy_prince: yw | 05:47 |
*** yuanying has joined #openstack-ironic | 05:55 | |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic-python-agent: Preserve environment variables when doing sudo https://review.openstack.org/122613 | 06:00 |
*** krtaylor has quit IRC | 06:09 | |
*** krtaylor has joined #openstack-ironic | 06:09 | |
*** rushiagr is now known as rushiagr_away | 06:11 | |
*** rushiagr_away is now known as rushiagr | 06:16 | |
*** rameshg87 has quit IRC | 06:21 | |
*** yuanying has quit IRC | 06:29 | |
*** k4n0 has joined #openstack-ironic | 06:29 | |
*** rameshg87 has joined #openstack-ironic | 06:39 | |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/120701 | 06:44 |
lazy_prince | Haomeng|2: i have another problem. I have the node uuid and i am trying to find the port registered for that node from ironic. Do you know any apis that can let me do that.. | 06:45 |
Haomeng|2 | lazy_prince: you mean we have ironic node uuid as input to check out the neutron port id? | 06:46 |
Haomeng|2 | Haomeng|2: or ironic port id? | 06:46 |
lazy_prince | Haomeng|2: nope.. ironic port. i want to extract the mac id from it. | 06:47 |
Haomeng|2 | lazy_prince: we have api | 06:47 |
Haomeng|2 | let me check | 06:47 |
Haomeng|2 | lazy_prince: v1/nodes/<ironic_node_id>/ports | 06:48 |
Haomeng|2 | lazy_prince: this api returns ironic port uuid and address | 06:48 |
lazy_prince | Haomeng|2: aha.. let me check that.. | 06:49 |
Haomeng|2 | lazy_prince: sorry, it return port uuid and port mac | 06:49 |
*** dtantsur|afk is now known as dtantsur | 06:51 | |
dtantsur | Morning Ironic, TGIF finally :) | 06:51 |
Haomeng|2 | dtantsur: morning:) | 06:53 |
*** ifarkas has joined #openstack-ironic | 06:53 | |
*** ifarkas has quit IRC | 07:10 | |
*** ifarkas has joined #openstack-ironic | 07:10 | |
*** yuanying has joined #openstack-ironic | 07:25 | |
*** andreykurilin_ has joined #openstack-ironic | 07:35 | |
*** romcheg has joined #openstack-ironic | 07:47 | |
*** ndipanov_gone is now known as ndipanov | 07:49 | |
*** jcoufal has joined #openstack-ironic | 07:52 | |
*** yuanying has quit IRC | 07:59 | |
*** Poornima has quit IRC | 07:59 | |
*** bluex-pl has joined #openstack-ironic | 07:59 | |
*** jistr has joined #openstack-ironic | 08:06 | |
*** rameshg87 has quit IRC | 08:06 | |
*** rameshg87 has joined #openstack-ironic | 08:07 | |
*** derekh has joined #openstack-ironic | 08:11 | |
dtantsur | lsmola, re ironic-conductor[4728]: 2014-09-19 08:01:33.850 4728 TRACE ironic.dhcp.neutron Unauthorized: Unknown auth strategy: None | 08:23 |
dtantsur | what's in $ grep auth etc/ironic/ironic.conf ? | 08:23 |
lsmola | dtantsur: http://paste.openstack.org/show/113193/ | 08:24 |
lsmola | dtantsur: neutron-server[5252]: 2014-09-19 08:01:33.846 5252 WARNING keystonemiddleware.auth_token [-] Authorization failed for token | 08:25 |
lsmola | dtantsur: I would say that it just fails to auth with the token | 08:25 |
dtantsur | lemme have a look at neutron-related code | 08:26 |
*** yongli has joined #openstack-ironic | 08:28 | |
lsmola | dtantsur: this is the log around that event http://paste.openstack.org/show/113194/ | 08:28 |
lsmola | dtantsur: I will try to restart neutron-server and try again | 08:28 |
*** lucasagomes has joined #openstack-ironic | 08:29 | |
lsmola | dtantsur: or maybe just rerun stack-create first | 08:29 |
lsmola | dtantsur: could be that ironic holds the token for too long and it don't get new when it expires? | 08:30 |
dtantsur | lsmola, I'm trying to check this | 08:31 |
lsmola | dtantsur: ok | 08:31 |
dtantsur | lsmola, so could you please dump token arriving at ironic/dhcp/neutron.py ? | 08:31 |
*** MattMan has quit IRC | 08:31 | |
dtantsur | lsmola, and check it manually against keystone | 08:32 |
lsmola | dtantsur: where do I find it? | 08:32 |
*** MattMan has joined #openstack-ironic | 08:32 | |
dtantsur | lsmola, /opt/stack/venvs/openstack/lib/python2.7/site-packages/ironic/dhcp/neutron.py : NeutronDHCPApi.__init__ | 08:33 |
lsmola | dtantsur: well nevertheless, if ironic hold the token, restarting irnocin-conductor should do the trick, rigt? | 08:33 |
dtantsur | lsmola, should do. you can try | 08:33 |
lsmola | dtantsur: ok | 08:33 |
dtantsur | lsmola, please let me know how it goes | 08:33 |
lsmola | dtantsur: sound easier than checking token manually | 08:34 |
*** rakesh_hs has quit IRC | 08:38 | |
lsmola | dtantsur: btw. I am using heat trusts, so might be that the token is propagated from heat to ironic | 08:42 |
lsmola | dtantsur: but no sure about this | 08:42 |
*** igordcard has joined #openstack-ironic | 08:42 | |
dtantsur | lsmola, never heard about heat trusts... | 08:42 |
lsmola | dtantsur: restarting conductor, we will see if there were some stuck token there | 08:43 |
*** teju has joined #openstack-ironic | 08:44 | |
lsmola | dtantsur: https://github.com/openstack/ironic/blob/8fdd7d0d8d4005fa913d453e0b086f9a7ae924d7/ironic/conductor/manager.py#L1074 | 08:44 |
lsmola | dtantsur: it's only use of context.auth_token here | 08:45 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/122423 | 08:45 |
lsmola | dtantsur: hm tha is not master | 08:46 |
dtantsur | what do you mean? | 08:46 |
lsmola | dtantsur: that the link was no the master, but it;s the same in the master | 08:47 |
dtantsur | ah right | 08:48 |
lsmola | dtantsur: hm so restarting ironic-conductor helps | 08:50 |
*** Poornima has joined #openstack-ironic | 08:51 | |
lsmola | dtantsur: so there was some stuck token somewhere | 08:51 |
dtantsur | well... | 08:51 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/120651 | 08:51 |
dtantsur | lsmola, could you please report it with all logs and our observations? this looks like smth serious | 08:51 |
lsmola | dtantsur: just rerunning didn't help, so it was most probably holding the same token | 08:52 |
lsmola | dtantsur: ok | 08:52 |
lsmola | dtantsur: do you know what is the default token expiration date? | 08:52 |
dtantsur | lsmola, seems to me that around a day, but I may be terribly wrong :) | 08:53 |
lsmola | dtantsur: yeah it could be something like that, I was able to deploy yesterday, which was within 24hours | 08:54 |
*** yuanying has joined #openstack-ironic | 08:55 | |
*** rakesh_hs has joined #openstack-ironic | 08:56 | |
lsmola | dtantsur: https://bugs.launchpad.net/ironic/+bug/1371503 | 09:08 |
dtantsur | thanks | 09:09 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Do not set the context twice when forming RPC objects https://review.openstack.org/122401 | 09:15 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Do not use the context parameter on refresh() https://review.openstack.org/122441 | 09:15 |
lsmola | dtantsur: btw. the trusts note was unrelated, so ignore it :-) | 09:18 |
dtantsur | ack | 09:18 |
*** yuanying has quit IRC | 09:29 | |
*** yuanying has joined #openstack-ironic | 09:29 | |
*** yuanying has quit IRC | 09:31 | |
*** ramineni1 has joined #openstack-ironic | 09:33 | |
*** dtantsur is now known as dtantsur|brb | 09:33 | |
*** rakesh_hs has quit IRC | 09:35 | |
*** ramineni has quit IRC | 09:35 | |
*** bluex-pl has quit IRC | 09:37 | |
*** romcheg has quit IRC | 09:40 | |
*** athomas has joined #openstack-ironic | 09:42 | |
rameshg87 | jroll, looks like there is an issue (with a workaround) in agent drivers: https://bugs.launchpad.net/ironic/+bug/1371505 | 09:42 |
*** bluex-pl has joined #openstack-ironic | 09:47 | |
*** romcheg has joined #openstack-ironic | 09:49 | |
*** rakesh_hs has joined #openstack-ironic | 09:59 | |
openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic-python-agent: Cleanup coreos-oem-inject.py https://review.openstack.org/122693 | 10:00 |
*** vinbs has joined #openstack-ironic | 10:03 | |
*** rakesh_hs has quit IRC | 10:21 | |
*** ramineni1 has quit IRC | 10:22 | |
*** ramineni has joined #openstack-ironic | 10:28 | |
*** pelix has joined #openstack-ironic | 10:33 | |
*** bluex-pl has quit IRC | 10:38 | |
*** bluex-pl has joined #openstack-ironic | 10:39 | |
*** nikunj2512 has quit IRC | 10:53 | |
*** andreykurilin_ has quit IRC | 10:56 | |
*** rameshg87 has quit IRC | 11:03 | |
*** bluex-pl has quit IRC | 11:03 | |
*** bluex-pl has joined #openstack-ironic | 11:05 | |
*** ramineni has quit IRC | 11:06 | |
*** rushiagr is now known as rushiagr_away | 11:19 | |
*** bluex-pl has quit IRC | 11:26 | |
lazy_prince | I am trying to provision an instance but it keeps failing while adding dhcp options. Here is the Traceback thrown by ironic http://paste.openstack.org/show/113245/ | 11:29 |
lazy_prince | I am using keystone for auth and the setup was created using devstack. | 11:30 |
*** bluex-pl has joined #openstack-ironic | 11:32 | |
*** bluex-pl has quit IRC | 11:33 | |
*** bluex-pl has joined #openstack-ironic | 11:33 | |
vinbs | lazy_prince, I had faced this issue few days back | 11:37 |
vinbs | lazy_prince, I'm not sure what was the cause of this error | 11:37 |
vinbs | lazy_prince, Just restarting the neutron services and Ironic service resolved the issue in my case | 11:38 |
*** lucasagomes is now known as lucas-hungry | 11:38 | |
vinbs | lazy_prince, my setup was openstack though | 11:38 |
vinbs | lazy_prince, not devstack | 11:38 |
lazy_prince | vinbs: aha.. so i guess then dhcp code is not yet stable in ironic.. or in neutron.. is it..? for me too, it worked couple of times and then I had this error consistent. | 11:42 |
vinbs | lazy_prince, did restarting the neutron services fix the issue for you? | 11:43 |
lazy_prince | vinbs: but i wanted to check with community if its only me or others also are also facing it.. | 11:43 |
lazy_prince | vinbs: going to check that now.. | 11:43 |
lazy_prince | vinbs: do i need to restart all neutron services or just q-svc should do..? | 11:44 |
vinbs | lazy_prince, it was working fine for me too for a long time and ran into this error suddenly for no apparent reason | 11:45 |
vinbs | lazy_prince, I restarted all the neutron services | 11:45 |
vinbs | lazy_prince, I could not pinpoint the problem since I could not reproduce this issue later | 11:47 |
lazy_prince | vinbs: i am facing it very frequently.. not sure why.. | 11:47 |
lazy_prince | vinbs: restarting neutron didn't help. however, i do not see this issue for the other two tenants that i have. | 11:52 |
yuriyz | morning Ironic | 11:55 |
yuriyz | I see Neutron DHCP provider uses only first token https://bugs.launchpad.net/ironic/+bug/1371503 | 11:56 |
*** teju has left #openstack-ironic | 11:59 | |
*** pensu has quit IRC | 12:00 | |
*** bluex-pl has quit IRC | 12:00 | |
lazy_prince | vinbs: now the other two projects also are not able to provision. | 12:03 |
lazy_prince | yuriyz: the bug says if ironic is running for >24hrs.. in my case its been few hrs only.. | 12:04 |
vinbs | lazy_prince, did you restart the ironic service too? | 12:04 |
lazy_prince | vinbs: restarted ironic now and trying now.. | 12:08 |
*** dtantsur|brb is now known as dtantsur | 12:10 | |
dtantsur | yuriyz, hi. what do you mean by 'first' token? | 12:11 |
*** dprince has joined #openstack-ironic | 12:13 | |
dtantsur | ram<TAB> :( | 12:14 |
yuriyz | dtantsur, NeutronDHCPApi instantiated only once, and uses task.context.auth_token from first deploy() call | 12:14 |
dtantsur | As to me, we need vendor-passthru for like 'agent' or 'pxe', not for the whole driver, but I'm not sure devananda and folks will agree... | 12:14 |
dtantsur | yuriyz, ouch | 12:14 |
yuriyz | dtantsur, look at the code on my links | 12:15 |
yuriyz | https://bugs.launchpad.net/ironic/+bug/1371503 | 12:16 |
dtantsur | yuriyz, right, thank you, I didn't get it after the first reading :) | 12:18 |
dtantsur | so, do we need to move 'token' argument to update_dhcp_opts? and create a client there? | 12:19 |
dtantsur | or stop using token there at all? | 12:19 |
* dtantsur needs to have a look how it's done for Glance | 12:20 | |
yuriyz | dtantsur, and I don't known "right" fix yet | 12:20 |
Shrews | morning ironic folk | 12:21 |
dtantsur | Shrews, morning | 12:22 |
Shrews | o/ | 12:22 |
dtantsur | yuriyz, looks like we also pass context + token to GLance service, but it's not cached | 12:22 |
yuriyz | yes | 12:23 |
lazy_prince | so anyone working on the fix for 1371503..? | 12:23 |
yuriyz | if we use token we should make new instance of Neutron client | 12:24 |
dtantsur | I can take it, if yuriyz didn't start | 12:24 |
yuriyz | ok | 12:24 |
dtantsur | yuriyz, that's a problem, as it uses extention manager for creation | 12:25 |
lazy_prince | i would be intrested in the fix. if you are ready with it, i would love to test and certify the fix. | 12:27 |
dtantsur | ack thanks | 12:27 |
*** pensu has joined #openstack-ironic | 12:38 | |
*** dlaube has quit IRC | 12:42 | |
*** dlaube has joined #openstack-ironic | 12:42 | |
*** romcheg1 has joined #openstack-ironic | 12:47 | |
*** romcheg has quit IRC | 12:47 | |
*** vinbs has quit IRC | 12:51 | |
*** romcheg1 has quit IRC | 12:53 | |
dtantsur | O_O How was this neutron code approved?? I hope it was not me >_< | 12:55 |
*** rushiagr_away is now known as rushiagr | 12:56 | |
*** Poornima has quit IRC | 12:59 | |
*** ramineni has joined #openstack-ironic | 12:59 | |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: Do not cache auth token in Neutron DHCP provider https://review.openstack.org/122728 | 13:10 |
dtantsur | lazy_prince, this is a WIP of the fix ^^^ I need more time to finish, because tests there are mainly broken and have to be rewritten | 13:11 |
lazy_prince | dtantsur: i will have a look at it | 13:12 |
toabctl | friendly reminder for another review of https://review.openstack.org/#/c/117704/ and https://review.openstack.org/#/c/120754/ | 13:22 |
*** linggao has joined #openstack-ironic | 13:30 | |
vdrok | afternoon Ironic! | 13:33 |
*** ramineni has quit IRC | 13:33 | |
vdrok | there is a question about locales - it seems that standard locale is required in at least 2 places - https://review.openstack.org/#/c/98102/ and https://review.openstack.org/#/c/122414/ | 13:34 |
vdrok | yuriyz suggestes in this change https://review.openstack.org/#/c/122414/ to add some flag in ironic.common.utils.execute | 13:35 |
Shrews | toabctl: you should poke in #tripleo for those | 13:35 |
vdrok | what are your thoughts on that ironickers? | 13:35 |
toabctl | Shrews: ok.thx | 13:35 |
*** k4n0 has quit IRC | 13:40 | |
Shrews | vdrok: seems reasonable to me to try to localize the locale settings to a single module | 13:43 |
* Shrews hasn't had enough coffee yet to know what that should look like | 13:44 | |
vdrok | Shrews, do you think it is worth of filing a bug, or maybe wait till RC? | 13:45 |
Shrews | vdrok: you may file a bug at any time, it just may not be targeted until after RC | 13:46 |
vdrok | Shrews, ok, thanks for answering :) | 13:47 |
NobodyCam | good morning Ironic | 13:56 |
jroll | gooooooood morning ironic :) | 13:57 |
jroll | lazy_prince: hey, been meaning to follow up with you. we don't use neutron for dhcp, so we haven't run into your problem with the provisioning network etc... but happy to dive in and help you figure it out | 13:57 |
NobodyCam | morning jroll :) | 13:57 |
jroll | ramesh found a bug but he's not online to thank :( | 13:57 |
jroll | heya NobodyCam, tgif! | 13:57 |
NobodyCam | oh ya | 13:58 |
Shrews | NobodyCam: jroll: morning | 13:58 |
jroll | heya Shrews dtantsur vdrok :) | 13:58 |
vdrok | morning jroll! :) | 13:58 |
lazy_prince | jroll: Thanks fo the help but i fixed the issue with DHCP but came across another dhcp related issue https://bugs.launchpad.net/ironic/+bug/1371503. | 13:59 |
jroll | 12:55:48 dtantsur | O_O How was this neutron code approved?? I hope it was not me >_< <-- what's wrong with it? :| | 13:59 |
NobodyCam | morning Shrews | 13:59 |
jroll | lol those tests :/ | 13:59 |
jroll | we need to be using autospec=True, or spec=... | 14:00 |
dtantsur | jroll, in additiona to https://bugs.launchpad.net/ironic/+bug/1371503 it's barely working tests (many using assertCalleOnceWith) | 14:00 |
lazy_prince | jroll: now that you are here, wanted to check the rational behine using a separate provisioning network for ironic-neutron-plugin... | 14:00 |
dtantsur | and morning/afternoon vdrok, jroll, NobodyCam :) | 14:00 |
jroll | lazy_prince: the separate provisioning network is the point... we don't want tenants to be able to access the control plane | 14:01 |
NobodyCam | mornig dtantsur | 14:01 |
vdrok | afternoon dtantsur ! | 14:01 |
lazy_prince | jroll: aha.. i get it.. | 14:01 |
NobodyCam | morning vdrok :) | 14:01 |
vdrok | morning NobodyCam | 14:01 |
vdrok | :) | 14:01 |
*** openstackgerrit has quit IRC | 14:02 | |
jroll | lazy_prince: that and allowing bonding/vlans on the instance were the two goals of that plugin | 14:02 |
lazy_prince | jroll: are you guys also extending the plugin to work with VXLAN or something..? | 14:02 |
jroll | there also may be some port security stuff to block ARP spoofing, etc, but that might just be configs on our switches | 14:03 |
jroll | lazy_prince: I expect to do that eventually, our switches don't support it at the moment :( | 14:03 |
jroll | so right now the tenant networks are flat networks | 14:03 |
NobodyCam | :) | 14:04 |
jroll | I actually am not sure any switches support it right now, to be honest | 14:04 |
lazy_prince | jroll: another thing i wanted to check is how do i specify which server is connected to what port on the switch.. does it need to be part of node extra spec or somewhere else.. | 14:04 |
NobodyCam | coffee ready brb... | 14:04 |
jroll | lazy_prince: yeah, we stuff it in node.extra right now... I expect that to move to port.extra but we haven't done it yet | 14:05 |
jroll | this will all be upstreamed somewhat quickly once K opens | 14:05 |
jroll | (at the latest, it may happen sooner) | 14:05 |
jroll | (but, specs process etc) | 14:05 |
jroll | that was a nice find with that neutron bug, btw | 14:06 |
* jroll looking at dtantsur's patch | 14:06 | |
jroll | I also thought neutron client supported noauth, ugh | 14:07 |
lazy_prince | jroll: Well.. I had to update the plugin code at few places to make it work. but mostly i struggled with database part. but now all issues are resolved. it is working for our switches too.. | 14:08 |
jroll | lazy_prince: nice :D feel free to send pull requests | 14:08 |
jroll | also, I think neutron changed how db migrations work, that might be what you ran into | 14:09 |
lazy_prince | yes | 14:09 |
jroll | if I remember what I heard correctly, they stopped automatically running migrations for extensions or something? | 14:09 |
lazy_prince | ack | 14:09 |
*** dprince has quit IRC | 14:10 | |
*** harlowja_away has quit IRC | 14:10 | |
*** coolsvap has quit IRC | 14:10 | |
*** pcrews has joined #openstack-ironic | 14:10 | |
*** foexle has joined #openstack-ironic | 14:11 | |
*** dprince has joined #openstack-ironic | 14:12 | |
*** coolsvap has joined #openstack-ironic | 14:15 | |
NobodyCam | ahh coffee :) | 14:15 |
*** ChuckC has quit IRC | 14:15 | |
dtantsur | jroll, lazy_prince, lsmola, updated https://review.openstack.org/#/c/122728/ please have a look | 14:18 |
jroll | dtantsur: hmm, what if we just got rid of the singleton, that would also do it yes? is that too much of a perf penalty to load stevedore every time? | 14:25 |
dtantsur | jroll, not sure about penalty (I expect some), but passing it token into the appropriate function looks better to me. Imagine someone would want to cache this factory again (why can't I cache factory?) | 14:26 |
jroll | yeah, suppose that's fair | 14:27 |
*** jcoufal has quit IRC | 14:28 | |
jroll | another approach... would be to cache the neutron client | 14:28 |
jroll | and if the token is expired, re-auth | 14:28 |
jroll | no that's dumb | 14:28 |
jroll | ignore me :) | 14:28 |
jroll | well, maybe it's not | 14:28 |
jroll | idk. | 14:28 |
NobodyCam | lol | 14:28 |
jroll | I tend to think anything that does auth should try to re-auth on a 403 | 14:29 |
jroll | we've been bitten by this with ironic-api's admin token before | 14:29 |
dtantsur | jroll, that's a good idea, but IIRC it's discussed OpenStack-wise somewhere on ML and indeed should be solved globally | 14:29 |
*** rwsu has quit IRC | 14:29 | |
*** romcheg has joined #openstack-ironic | 14:30 | |
jroll | right, in the clients | 14:30 |
jroll | (imo) | 14:30 |
jroll | but why don't we do that | 14:30 |
jroll | rather than passing a token every time | 14:30 |
jroll | either way, you need the admin auth configs. and it will always be an admin token | 14:31 |
dtantsur | jroll, for the sake of audit maybe? | 14:32 |
jroll | I'm not sure I understand | 14:32 |
jroll | if anything, not passing along the token from the context would make auditing easier | 14:33 |
jroll | as you could use different accounts for nova->ironic and ironic->neutron | 14:33 |
dtantsur | jroll, I maybe completely wrong, but a token can be associated with user, so maybe it's possible to track user | 14:33 |
dtantsur | no idea, just guessing | 14:33 |
jroll | not to talk to ironic | 14:33 |
jroll | and I hope your users can't configure dhcp in neutron for arbitrary hosts :) | 14:34 |
lucas-hungry | morning all :) | 14:34 |
lucas-hungry | dtantsur, interesting idea | 14:34 |
*** lucas-hungry is now known as lucasagomes | 14:34 | |
dtantsur | lucas-hungry, morning :) | 14:34 |
jroll | heya, lucas, happy friday :) | 14:34 |
dtantsur | actually afternoon, I guess | 14:34 |
lucasagomes | dtantsur, you have to pass the token ? | 14:34 |
lucasagomes | dtantsur, you already pass task no? the token is there | 14:35 |
NobodyCam | morning lucasagomes :) | 14:35 |
lucasagomes | NobodyCam, morning | 14:35 |
dtantsur | oh, jroll, mind also having a look at the direction where https://review.openstack.org/#/c/122137/ goes? | 14:36 |
dtantsur | lucasagomes, oh my, you're right. | 14:36 |
jroll | dtantsur: sure, meant to do that the other day | 14:36 |
jroll | still early, diffs are a little blurry right now :P | 14:36 |
dtantsur | heh I imagine | 14:36 |
lucasagomes | dtantsur, yeah, so we may not need to change the interfaces | 14:37 |
dtantsur | lucasagomes, right. oh I hate changing it back >_< | 14:38 |
jroll | just: git checkout HEAD~1 $files_you_want_back_to_normal | 14:38 |
jroll | so you don't blow away test fixes | 14:39 |
*** Haomeng has joined #openstack-ironic | 14:39 | |
lucasagomes | dtantsur, :( sorry didn't look the patch before | 14:39 |
*** Haomeng|2 has quit IRC | 14:40 | |
dtantsur | jroll, well, it will help for some files, but not for neutron.py | 14:40 |
dtantsur | lucasagomes, nevermind, I should have been smarter today :) | 14:40 |
jroll | dtantsur: yeah :/ | 14:40 |
dtantsur | also some API do not accept task... | 14:41 |
lucasagomes | dtantsur, no worries it's friday | 14:41 |
jroll | dtantsur: on 122137, direction looks fine, I still wish the tests didn't touch my file system, but that might just be me | 14:45 |
jroll | I think it's safe to assume open().write() and os.unlink() are well-tested already | 14:46 |
lucasagomes | jroll, sometimes I wonder that whether we should make tests run in a read-only fs | 14:47 |
lucasagomes | now that we don't generate any pyc etc | 14:47 |
lucasagomes | it may be possible | 14:47 |
lucasagomes | (idk if there's a good use case for that tho) | 14:47 |
jroll | hmm, idk | 14:47 |
dtantsur | lucasagomes, we have quite helpful tests for partitoning on a file | 14:47 |
jroll | I'd like to start using dox | 14:47 |
lucasagomes | dtantsur, yeah, could be in memory tho no? | 14:47 |
dtantsur | jroll, do you suggest to drop 'fullflow' test? | 14:47 |
dtantsur | lucasagomes, if you make gparted work with it :) | 14:48 |
jroll | dtantsur: idk, I think it might be valuable | 14:49 |
lucasagomes | dtantsur, heh, never tried but sounds possible | 14:49 |
lucasagomes | I mean I won't do that, heh I don't have any use case | 14:49 |
lucasagomes | but just sounds interesting :) | 14:49 |
dtantsur | jroll, ok let's leave as it is and see what people say | 14:49 |
jroll | ok | 14:50 |
jroll | well | 14:50 |
jroll | let's make tests pass :P | 14:50 |
dtantsur | jroll, lucasagomes, in the meanwhile pushed new version | 14:50 |
jroll | dtantsur: hmm, why not add a task param to update_port_address | 14:51 |
dtantsur | jroll, I didn't want to change signature too much. Also, if it does not need task, I'd avoid passing in task. I usually prefer passing exactly what functions requires | 14:52 |
jroll | mmmkay, sounds reasonable | 14:52 |
NobodyCam | ok who is going to sue who. http://developer.apple.com/swift/ | 14:54 |
jroll | hmm, I still would prefer to re-auth on 403, but maybe that's an optimization | 14:54 |
jroll | NobodyCam: lol, you're a couple months late on that :P | 14:54 |
NobodyCam | lol | 14:55 |
NobodyCam | :( | 14:55 |
dtantsur | jroll, I think it makes sense, but not in this stage of the release cycle :) | 14:56 |
jroll | yeah | 14:56 |
*** pensu has quit IRC | 14:56 | |
*** athomas has quit IRC | 15:05 | |
*** athomas has joined #openstack-ironic | 15:05 | |
*** ndipanov has quit IRC | 15:08 | |
*** openstackgerrit has joined #openstack-ironic | 15:15 | |
jroll | dtantsur: reviewed 122728 | 15:16 |
dtantsur | jroll, thanks. you suggest just dropping __init__? | 15:17 |
*** andreykurilin_ has joined #openstack-ironic | 15:17 | |
dtantsur | jroll, second, mock_client_init.return_value = None is required, Python complains otherwise IIRC | 15:18 |
dtantsur | lemme check | 15:18 |
jroll | dtantsur: I don't see why not | 15:18 |
jroll | really? | 15:18 |
jroll | good one, python | 15:18 |
dtantsur | jroll, TypeError: __init__() should return None, not 'MagicMock' | 15:19 |
dtantsur | :) | 15:19 |
jroll | oh, right | 15:19 |
jroll | ok | 15:19 |
NobodyCam | brb... quick run to the gas station for a pack of smokes | 15:20 |
dtantsur | jroll, do you have something against mock.sentinel? I use it just for brevity, as nothing is called on it | 15:23 |
jroll | nah, just asking for consistency | 15:23 |
jroll | I'm ok with it | 15:23 |
jroll | asking about consistency, I should say | 15:23 |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: Do not cache auth token in Neutron DHCP provider https://review.openstack.org/122728 | 15:24 |
dtantsur | jroll, ^^ | 15:25 |
jroll | thanks | 15:25 |
* NobodyCam is back | 15:29 | |
*** athomas has quit IRC | 15:29 | |
*** mordred_ has joined #openstack-ironic | 15:31 | |
*** mordred has quit IRC | 15:31 | |
*** aignatov has quit IRC | 15:31 | |
*** viktors|afk has quit IRC | 15:31 | |
*** jeblair has quit IRC | 15:31 | |
*** cinerama has quit IRC | 15:31 | |
*** morgabra has quit IRC | 15:31 | |
*** d0ugal has quit IRC | 15:31 | |
*** mordred_ is now known as mordred | 15:31 | |
NobodyCam | morning mordred :) | 15:31 |
*** viktors has joined #openstack-ironic | 15:31 | |
*** morgabra_ has joined #openstack-ironic | 15:32 | |
*** jeblair has joined #openstack-ironic | 15:32 | |
*** aignatov has joined #openstack-ironic | 15:33 | |
*** cinerama has joined #openstack-ironic | 15:34 | |
*** d0ugal has joined #openstack-ironic | 15:35 | |
*** d0ugal is now known as Guest50648 | 15:35 | |
*** andreykurilin_ has quit IRC | 15:36 | |
dtantsur | finally calling it a day, have a nice weekend! | 15:39 |
*** dtantsur is now known as dtantsur|afk | 15:39 | |
jroll | night dtantsur|afk :) | 15:39 |
NobodyCam | have a good night dtantsur|afk | 15:42 |
*** andreykurilin_ has joined #openstack-ironic | 15:44 | |
*** lynxman_ has joined #openstack-ironic | 15:44 | |
*** JoshNang_ has joined #openstack-ironic | 15:45 | |
*** Mike_D_wk1 has quit IRC | 15:46 | |
*** JoshNang has quit IRC | 15:46 | |
*** lynxman has quit IRC | 15:46 | |
*** ekarlso has quit IRC | 15:46 | |
*** JoshNang_ is now known as JoshNang | 15:46 | |
*** lynxman_ is now known as lynxman | 15:46 | |
*** ekarlso has joined #openstack-ironic | 15:46 | |
*** ChuckC has joined #openstack-ironic | 15:47 | |
*** Mike_D_wk has joined #openstack-ironic | 15:47 | |
*** dprince has quit IRC | 15:47 | |
*** foexle has quit IRC | 15:47 | |
*** romcheg has left #openstack-ironic | 15:49 | |
*** ChuckC_ has joined #openstack-ironic | 15:51 | |
*** ChuckC has quit IRC | 15:51 | |
*** lazy_prince is now known as killer_prince | 15:57 | |
*** ifarkas has quit IRC | 16:18 | |
jroll | final patch for CI with agent_ssh https://review.openstack.org/112693 | 16:20 |
jroll | reviews more than welcome on that | 16:21 |
JayF | jroll: I don't think the IPA job even should be voting to start | 16:23 |
jroll | JayF: it's not :) | 16:23 |
JayF | jroll: until we see it have a pattern of passing tests | 16:23 |
jroll | just posted a new one | 16:23 |
jroll | my commit message might be bad, ugh | 16:23 |
JayF | yeah, your commit message says what I said :P | 16:24 |
jroll | thanks for that | 16:24 |
JayF | I didn't even get to the code | 16:24 |
JayF | gotta know what you're trying to do before I can judge if you did it or not :) | 16:24 |
jroll | right :P | 16:24 |
jroll | forgot my commit message was so explanatory | 16:24 |
jroll | that should be good to go now | 16:24 |
jroll | adam_g: https://review.openstack.org/112693 | 16:24 |
*** Guest50648 is now known as d0ugal | 16:29 | |
*** d0ugal has quit IRC | 16:30 | |
*** d0ugal has joined #openstack-ironic | 16:30 | |
Shrews | All: This bug (https://bugs.launchpad.net/ironic/+bug/1347604) is targetted to RC1. It's actually fixed, but needs a doc update that the author of the bug fix didn't feel comfortable doing. Is anyone working on that? If not, I'll knock that out for us. | 16:33 |
*** rwsu has joined #openstack-ironic | 16:34 | |
jroll | I'm not working on it | 16:35 |
jroll | I know that much :P | 16:35 |
NobodyCam | nor I | 16:35 |
*** wanyen has joined #openstack-ironic | 16:35 | |
Shrews | ok. i speak english goodly, so is unpossible for me to mess this up. | 16:36 |
NobodyCam | lol | 16:36 |
jroll | hehehe | 16:37 |
jroll | lots of open IPA reviews :o | 16:37 |
JayF | I think my docstring one needs updates and re-review | 16:37 |
*** penick has joined #openstack-ironic | 16:38 | |
*** ChuckC_ has quit IRC | 16:41 | |
*** igordcard has quit IRC | 16:41 | |
*** dprince has joined #openstack-ironic | 16:43 | |
Shrews | jroll: the IPA docstrings reviews already have 3 +2s | 16:43 |
* JayF lands things | 16:44 | |
jroll | heh | 16:44 |
jroll | wait | 16:44 |
jroll | ruby had nits, JayF | 16:44 |
JayF | jroll: okay I unlanded the middle one so you can fix :P | 16:45 |
jroll | lol | 16:45 |
*** derekh has quit IRC | 16:46 | |
JayF | jroll: you can review and land this -> https://review.openstack.org/#/c/121433/ | 16:46 |
jroll | ok | 16:46 |
* jroll has a list of ten things to do in his head right now | 16:46 | |
JayF | oh man | 16:46 |
jroll | gimme a few | 16:46 |
JayF | TIL about sudo -E | 16:46 |
JayF | that's amazing | 16:46 |
jroll | JayF: what does $$ do in bash? | 16:47 |
jroll | heh, ikr | 16:47 |
jroll | TMP_BUILD_DIR="/tmp/$SCRIPTNAME.$$" | 16:47 |
JayF | $$ is pid iirc | 16:47 |
JayF | yeah, that is it | 16:47 |
jroll | appears correct | 16:47 |
jroll | thanks | 16:47 |
openstackgerrit | A change was merged to openstack/ironic-python-agent: Add docstrings for agent.py https://review.openstack.org/120660 | 16:51 |
openstackgerrit | Jay Faulkner proposed a change to openstack/ironic-python-agent: Improve docstrings for cmd/agent.py https://review.openstack.org/120662 | 16:53 |
lucasagomes | folks I will call it a day :) | 16:53 |
lucasagomes | have a great night everyone, enjoy the weekend! | 16:53 |
jroll | night lucas :) | 16:53 |
*** lucasagomes is now known as lucas-dinner | 16:54 | |
NobodyCam | night lucas-dinner | 16:54 |
*** pelix has quit IRC | 16:54 | |
NobodyCam | jroll: JayF: I see a lot of comments about $$ such as: Using $$ is a bad idea, because it will usually create a race condition, and allow your shell-script to be subverted by an attacker. | 16:56 |
jroll | NobodyCam: O.o | 16:56 |
jroll | I think it's fine here, it's in a build script | 16:56 |
*** linggao has quit IRC | 16:56 | |
NobodyCam | ack ... just thought I point it out :) | 16:57 |
jroll | yeah :) | 16:57 |
openstackgerrit | Jay Faulkner proposed a change to openstack/ironic-python-agent: Improve/add docstrings for CommandResult classes https://review.openstack.org/120663 | 16:57 |
*** Hefeweizen has quit IRC | 16:59 | |
*** linggao has joined #openstack-ironic | 17:01 | |
openstackgerrit | A change was merged to openstack/ironic-python-agent: Preserve environment variables when doing sudo https://review.openstack.org/122613 | 17:05 |
*** linggao has quit IRC | 17:05 | |
openstackgerrit | A change was merged to openstack/ironic-python-agent: Add support to build iso image https://review.openstack.org/121433 | 17:05 |
*** morgabra_ is now known as morgabra | 17:08 | |
*** morgabra has joined #openstack-ironic | 17:08 | |
*** eghobo has joined #openstack-ironic | 17:14 | |
openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-python-agent: Add docstrings for api/ tree https://review.openstack.org/120661 | 17:18 |
jroll | Shrews: ^ mind looking at that again? | 17:19 |
openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-python-agent: Delete unused method get_agent_mac_address() https://review.openstack.org/121153 | 17:19 |
jroll | and maybe re-+A that one | 17:19 |
Shrews | jroll: sure. give me a little bit | 17:19 |
jroll | yeah, no rush, just a friendly poke :) | 17:19 |
*** lucas-dinner has quit IRC | 17:23 | |
*** MattMan has quit IRC | 17:24 | |
openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Update docs with new dbsync command https://review.openstack.org/122802 | 17:31 |
*** tatyana has joined #openstack-ironic | 17:32 | |
*** jistr has quit IRC | 17:32 | |
*** penick has quit IRC | 17:35 | |
*** yjiang5 has joined #openstack-ironic | 17:35 | |
Shrews | jroll: for the a vs. an stuff, i actually thought you had it correct the first time. Isn't the class name the part that shows up in documentation? | 17:37 |
devananda | morning, all | 17:37 |
NobodyCam | good morning devananda | 17:37 |
jroll | Shrews: it should be an, that was one I overlooked | 17:38 |
jroll | so it says 'an ironic_python_agent....' | 17:38 |
jroll | morning devananda :) | 17:38 |
NobodyCam | just reading mordred's blog post :) | 17:38 |
Shrews | jroll: oh, it doesn't generate as "an BaseCommandResult"? | 17:38 |
jroll | oh | 17:39 |
jroll | yeah, you might be right | 17:39 |
Shrews | NobodyCam: that's a good one :) long, but a good read | 17:39 |
jroll | now I need to look at everything again :( | 17:39 |
jroll | Shrews: can you -1 me for that, please? | 17:39 |
jroll | and I'll take a look later | 17:39 |
Shrews | jroll: joyfully! :-P | 17:39 |
devananda | adam_g: regarding https://review.openstack.org/#/c/122539/1 -- I am wondering if that's a sufficient change | 17:39 |
devananda | adam_g: lib/ironic:588 pip_install diskimage_builder | 17:40 |
jroll | heh | 17:40 |
*** harlowja has joined #openstack-ironic | 17:42 | |
devananda | adam_g: commented on the review | 17:44 |
Shrews | devananda: umm... that is a change to devstack-gate, not devstack itself. it can't be done in the same review | 17:46 |
devananda | Shrews: d'oh | 17:46 |
*** andreykurilin_ has quit IRC | 17:49 | |
*** fyb3r has joined #openstack-ironic | 17:53 | |
adam_g | devananda, yeah, that will end up having it installed from a git checkout before devstack runs | 17:55 |
devananda | adam_g: right. which is good. | 17:55 |
adam_g | okay. i remember there being some discussion here about the devstack changes prior to them merging, but couldnt remember find the logs. | 17:56 |
devananda | i see that dib is a service now | 17:56 |
devananda | but not sure if it defaults to enabled, or what | 17:56 |
devananda | it looks like "no" | 17:56 |
* devananda runs locally | 17:57 | |
adam_g | no, its disabled by default | 17:57 |
adam_g | in which case its not installed, and lib/ironic pip's it | 17:57 |
devananda | enabling ironic will need to turn it on | 17:57 |
devananda | hm | 17:57 |
adam_g | 122539 enables the service in devstack gate | 17:57 |
devananda | right | 17:57 |
devananda | so either we pip install it when the service isn't present, or we have it as a dependent service | 17:58 |
devananda | enabling ironic will enable dib | 17:58 |
adam_g | im okay with keeping devstack as-is and having it installed only when the dib service is explicitly enabled | 17:59 |
devananda | adam_g: the current default is to pip install at run time | 18:00 |
devananda | since it's not listed in files/apts/* I think that's going to pull from the 'net | 18:01 |
devananda | not local caches | 18:01 |
devananda | which is bad for the gate | 18:01 |
*** openstackgerrit has quit IRC | 18:01 | |
*** openstackgerrit has joined #openstack-ironic | 18:03 | |
adam_g | devananda, no | 18:05 |
adam_g | well | 18:05 |
adam_g | right, as-is now | 18:05 |
devananda | right | 18:05 |
adam_g | enabling it as a service in features.yaml will just used the cached git checkout thats created to the zuul_reef before devstack | 18:05 |
devananda | righ | 18:08 |
devananda | also, enabling the dib service will clone tripleo-image-elements and os-*-config repos | 18:08 |
devananda | at the moment, it looks like those are only used if HEAT_CREATE_TEST_IMAGE is True | 18:09 |
devananda | adam_g: so I kinda don't like the idea that if we leave devstack as-is, ironic would run differently locally, by default, than in the gate | 18:11 |
devananda | because it would pip install dib locally but use a git clone in the gate | 18:12 |
adam_g | fair | 18:14 |
devananda | adam_g: better for ironic to refuse to start if dib is not enabled, or for it to enable it? | 18:16 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/114058 | 18:17 |
adam_g | devananda, ill just add a check similar to how to check for neutron | 18:18 |
*** penick has joined #openstack-ironic | 18:20 | |
NobodyCam | anyone know off the tops of their head where the neutron dhcp lease file is kept | 18:21 |
adam_g | $state_dir/dhcp/32fd5e91-e304-404b-a9dd-80626398f273/host | 18:23 |
adam_g | NobodyCam, should be specified as an parameter ot the running dnsmasq process | 18:24 |
devananda | adam_g: see this error before? it's in a recent pxe_ssh-postgres run | 18:24 |
devananda | 2014-09-19 12:26:43.257 31192 WARNING wsme.api [-] Client-side error: Cannot complete the requested action because chassis 18 contains nodes. | 18:24 |
NobodyCam | adam_g: awesome TY | 18:25 |
adam_g | devananda, wtf? no | 18:25 |
devananda | http://logs.openstack.org/23/122423/3/check/check-tempest-dsvm-ironic-pxe_ssh-postgres/04c7358/logs/screen-ir-api.txt.gz?level=INFO | 18:25 |
adam_g | hmm | 18:26 |
adam_g | http://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiQ2Fubm90IGNvbXBsZXRlIHRoZSByZXF1ZXN0ZWQgYWN0aW9uIGJlY2F1c2UgY2hhc3Npc1wiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxNDExMTUxMjI0NDI2fQ== | 18:26 |
Shrews | devananda: some of those can be totally normal based on which tempest test is running | 18:27 |
devananda | Shrews: yes, but not this one | 18:27 |
adam_g | devananda, can you open a bug? ill look at it when i get back | 18:27 |
devananda | adam_g: ack. will gather a few more details | 18:28 |
adam_g | it doesn't look isolated to the postgres jobs either | 18:29 |
adam_g | smells like a race between delete chassis + delete node in test cleanup | 18:29 |
devananda | nope | 18:29 |
devananda | or some test is adding a lot of nodes to a chassis it shouldn't be | 18:29 |
openstackgerrit | Syed Ismail Faizan Barmawer proposed a change to openstack/ironic: pxe_ilo driver will fail on ipmitool error https://review.openstack.org/122819 | 18:29 |
devananda | in the one i firce noticed, the # of nodes keeps increasing | 18:29 |
jroll | sure that isn't the chassis ID? | 18:30 |
NobodyCam | gah --leasefile-ro == Completely suppress use of the lease database file. | 18:30 |
jroll | I read it as "contains 18 nodes" at first | 18:30 |
jroll | devananda: ^ | 18:30 |
devananda | jroll: oh. you're right | 18:30 |
*** tatyana has quit IRC | 18:31 | |
devananda | started < 24hr ago, appearing across any project with a check ironic job | 18:33 |
devananda | :( | 18:33 |
jroll | does that fail the job? | 18:35 |
devananda | yes | 18:37 |
devananda | every time | 18:37 |
devananda | https://bugs.launchpad.net/ironic/+bug/1371755 | 18:37 |
adam_g | hmm | 18:37 |
adam_g | back in a bit, will dig into that when im back | 18:39 |
jroll | <3 broken gates | 18:39 |
* jroll gives up on today | 18:39 | |
jroll | if it's every run, it's probably not a race | 18:40 |
*** rushiagr is now known as rushiagr_away | 18:40 | |
jroll | oh ha | 18:41 |
jroll | https://github.com/openstack/tempest/blob/master/tempest/api/baremetal/admin/base.py#L78 | 18:41 |
jroll | looks like it deletes things in the same order it creates them | 18:41 |
jroll | so setup creates: chassis, nodes | 18:41 |
jroll | then teardown deletes: chassis, nodes | 18:42 |
jroll | it's a dictionary, though | 18:42 |
jroll | but I see tempest landed hash seed = 0 things recently | 18:42 |
jroll | likely related | 18:42 |
jroll | devananda adam_g ^ | 18:43 |
adam_g | jroll, yeah, that was my suspicion. in which case i broke the gate. | 18:44 |
jroll | heh, it happens | 18:44 |
adam_g | may want to propose reverting https://review.openstack.org/#/c/116744 as it may be premature. | 18:44 |
adam_g | i need to run, im late for an appointment | 18:45 |
jroll | ok, I'll put that up | 18:45 |
*** harlowja has quit IRC | 18:46 | |
* devananda notes that two ironic jobs failed on that tempest patch | 18:46 | |
*** harlowja_ has joined #openstack-ironic | 18:46 | |
*** marcoemorais has joined #openstack-ironic | 18:46 | |
jroll | yeah :/ | 18:47 |
devananda | yep. it is definitely caused by that tempest change | 18:48 |
devananda | the only cases of it prior to today were 4 instances that happened on 9/15 on that patch series | 18:48 |
devananda | http://logstash.openstack.org/#eyJzZWFyY2giOiJ0YWdzOmNvbnNvbGUgQU5EIG1lc3NhZ2U6XCJDYW5ub3QgY29tcGxldGUgdGhlIHJlcXVlc3RlZCBhY3Rpb24gYmVjYXVzZSBjaGFzc2lzXCIiLCJmaWVsZHMiOltdLCJvZmZzZXQiOjAsInRpbWVmcmFtZSI6ImN1c3RvbSIsImdyYXBobW9kZSI6ImNvdW50IiwidGltZSI6eyJmcm9tIjoiMjAxNC0wOS0xMlQxODozMjozNSswMDowMCIsInRvIjoiMjAxNC0wOS0xNlQxODozMjozNSswMDowMCIsInVzZXJfaW50ZXJ2YWwiOiIwIn0sInN0YW1wIjoxNDExMTUyNDgwNTM5fQ== | 18:49 |
devananda | jroll: if you're posting the revert, I'm going to step afk for a few. | 18:49 |
jroll | devananda: I just did | 18:50 |
Shrews | w00t: https://review.openstack.org/121726 | 18:50 |
Shrews | so not just ironic with that problem :) | 18:50 |
NobodyCam | :) | 18:51 |
Shrews | fyi, this starts in 8 minutes: https://wiki.openstack.org/wiki/BootstrappingHouse/Mock_Best_Practices | 18:51 |
jroll | lol Shrews | 18:52 |
jroll | fyi devananda adam_g Shrews https://review.openstack.org/122826 | 18:52 |
jroll | that's the revert | 18:52 |
*** pensu has joined #openstack-ironic | 18:52 | |
Shrews | jroll: waht's the connection b/w PYTHONHASHSEED and the failing tests? | 18:53 |
Shrews | s/waht's/what's/ | 18:54 |
Shrews | or do we only know that's when the failures started? | 18:54 |
jroll | Shrews: so, tempest cleanup uses an iteritems() to get things to clean up | 18:55 |
Shrews | ah, ok. nm | 18:55 |
Shrews | that's all you needed to say :) | 18:55 |
jroll | the dict it iterates over is {'chassis': [...], 'node': [...], ...} | 18:55 |
jroll | heh, yeah | 18:55 |
jroll | deleting chassis before nodes | 18:55 |
jroll | etc | 18:55 |
Shrews | i missed your link above | 18:55 |
JayF | jroll: put a comment on 122826 | 19:00 |
jroll | JayF: I'm probably going to abandon that | 19:01 |
jroll | JayF: also, I just reverted a commit, I intended to revert it | 19:02 |
JayF | jroll: yeah just looked weird; I know it's simply a revert | 19:02 |
jroll | yeah :P | 19:05 |
jroll | anyhow, real fix is up: https://review.openstack.org/122832 | 19:05 |
jroll | ^ devananda adam_g Shrews | 19:05 |
*** marcoemorais has quit IRC | 19:40 | |
*** marcoemorais has joined #openstack-ironic | 19:41 | |
*** marcoemorais has quit IRC | 19:41 | |
*** pensu has quit IRC | 19:44 | |
*** dprince has quit IRC | 19:48 | |
*** eghobo has quit IRC | 19:56 | |
NobodyCam | brb | 20:05 |
adam_g | jroll, nice | 20:12 |
-openstackstatus- NOTICE: Gerrit will be offline from 20:30 to 20:50 UTC for project renames | 20:15 | |
*** ChanServ changes topic to "Gerrit will be offline from 20:30 to 20:50 UTC for project renames" | 20:15 | |
*** marcoemorais has joined #openstack-ironic | 20:15 | |
adam_g | sorry about letting that tempest change through, bug 1371755 goes on my permanent record | 20:23 |
*** eghobo has joined #openstack-ironic | 20:25 | |
*** tatyana has joined #openstack-ironic | 20:30 | |
*** marcoemorais has quit IRC | 20:31 | |
NobodyCam | humm review just go away | 20:32 |
*** marcoemorais has joined #openstack-ironic | 20:33 | |
*** andreykurilin_ has joined #openstack-ironic | 20:33 | |
*** ChuckC_ has joined #openstack-ironic | 20:35 | |
*** marcoemorais1 has joined #openstack-ironic | 20:40 | |
*** marcoemorais has quit IRC | 20:40 | |
*** marcoemorais1 has quit IRC | 20:40 | |
*** marcoemorais has joined #openstack-ironic | 20:40 | |
adam_g | NobodyCam, i think -infra is shutting down the gerrit+zuul for maintenance | 20:42 |
NobodyCam | yep I got to reading the romm title | 20:42 |
NobodyCam | :-p | 20:43 |
NobodyCam | lol what they didn't want to make changes at 5 pm on a friday | 20:44 |
NobodyCam | lol | 20:44 |
*** eghobo has quit IRC | 20:44 | |
*** tatyana has quit IRC | 20:48 | |
*** eghobo has joined #openstack-ironic | 20:50 | |
*** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic" | 20:51 | |
-openstackstatus- NOTICE: Gerrit is back online | 20:51 | |
*** openstackgerrit has quit IRC | 20:55 | |
*** openstackgerrit has joined #openstack-ironic | 20:56 | |
*** penick has quit IRC | 20:57 | |
*** penick has joined #openstack-ironic | 21:02 | |
*** marcoemorais has quit IRC | 21:03 | |
*** marcoemorais has joined #openstack-ironic | 21:04 | |
*** marcoemorais has quit IRC | 21:05 | |
*** marcoemorais has joined #openstack-ironic | 21:05 | |
*** marcoemorais has quit IRC | 21:06 | |
*** marcoemorais has joined #openstack-ironic | 21:06 | |
*** fyb3r has left #openstack-ironic | 21:16 | |
*** rushiagr_away is now known as rushiagr | 21:32 | |
jroll | adam_g: https://review.openstack.org/#/c/112693/7/modules/openstack_project/files/jenkins_job_builder/config/devstack-gate.yaml | 21:57 |
jroll | is there somewhere else I need to add the build-ramdisk key? | 21:57 |
adam_g | jroll, no, that looks right | 21:59 |
jroll | adam_g: hrm, it's failing the layout job | 21:59 |
adam_g | jroll, actually you'd need to set it to 1 for the other ironic jobs | 21:59 |
adam_g | pxe_ssh | 21:59 |
*** harlowja_ has quit IRC | 22:00 | |
jroll | mmm yeah | 22:00 |
adam_g | yeah, i think the KeyError is due to the other jobs not specifying it | 22:00 |
jroll | ok | 22:00 |
* jroll tries again | 22:00 | |
*** harlowja has joined #openstack-ironic | 22:00 | |
adam_g | the layout job should be runnable locally | 22:00 |
adam_g | tox -e zuul i think | 22:00 |
jroll | ok | 22:01 |
jroll | does every job in that file need to specify it? | 22:01 |
jroll | or just pxe_ssh etc | 22:01 |
jroll | oh, every devstack-virtual-ironic | 22:01 |
adam_g | yea | 22:01 |
jroll | thanks | 22:02 |
adam_g | AFAIK JJB has no way of setting default values for parameters | 22:02 |
jroll | yeah | 22:02 |
*** penick has quit IRC | 22:02 | |
*** marcoemorais has quit IRC | 22:04 | |
*** marcoemorais has joined #openstack-ironic | 22:04 | |
*** marcoemorais has quit IRC | 22:04 | |
*** marcoemorais has joined #openstack-ironic | 22:05 | |
*** marcoemorais has quit IRC | 22:05 | |
*** marcoemorais has joined #openstack-ironic | 22:05 | |
*** penick has joined #openstack-ironic | 22:08 | |
NobodyCam | brb... walkies time | 22:10 |
jroll | updated https://review.openstack.org/#/c/112693/ | 22:11 |
*** eghobo has quit IRC | 22:18 | |
*** penick has quit IRC | 22:53 | |
openstackgerrit | Syed Ismail Faizan Barmawer proposed a change to openstack/ironic: add_node_capability and rm_node_capability unable to save changes to db https://review.openstack.org/122889 | 23:05 |
*** faizan has joined #openstack-ironic | 23:11 | |
*** andreykurilin_ has quit IRC | 23:37 | |
*** faizan has quit IRC | 23:40 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!