| *** erlon has quit IRC | 00:05 | |
| *** financeMann has joined #openstack-kolla | 00:05 | |
| *** ccrouch has quit IRC | 00:09 | |
| *** krtaylor has quit IRC | 00:09 | |
| financeMann | Howdy, tryin to deploy to centos 7 having some dependency issue, any help? http://pastebin.com/FmYfAcC0 | 00:11 |
|---|---|---|
| *** lamt has joined #openstack-kolla | 00:12 | |
| *** lamt has quit IRC | 00:12 | |
| bpattewar | financeMann can you ansible version of your host, does it as per the documentation. Kolla ansible and kolla-kubernetes need different version of ansible | 00:18 |
| *** goldyfruit has quit IRC | 00:18 | |
| bpattewar | I mean kolla-ansible and halcyon-vagrant-kubernetes use different versions | 00:18 |
| financeMann | bpattewar ansible 2.2.1.0 | 00:18 |
| financeMann | i think i need < 2 bpattewar? | 00:19 |
| bpattewar | yes you are right. | 00:21 |
| financeMann | bpattewar whats the command to dongrade? | 00:22 |
| inc0 | no, that's not correct | 00:24 |
| *** salv-orlando has joined #openstack-kolla | 00:24 | |
| inc0 | you need 2+ for newton | 00:24 |
| inc0 | but 2.2 might be a it too new;) | 00:25 |
| inc0 | well... I'm runnning 2.2 so that's not it | 00:25 |
| inc0 | pip freeze | grep cryptography <= run this financeMann plz | 00:26 |
| financeMann | inc0 cryptography==1.7.2 | 00:27 |
| *** penfold has quit IRC | 00:28 | |
| inc0 | financeMann: how about this? python -c "import setuptools; print setuptools.__version__" | 00:29 |
| *** salv-orlando has quit IRC | 00:29 | |
| financeMann | 34.2.0 | 00:29 |
| financeMann | inc0 | 00:29 |
| inc0 | ok so same as me | 00:29 |
| inc0 | hmm | 00:29 |
| financeMann | inc0 ya want access to vnc in? ^_^ | 00:29 |
| inc0 | I'd need a client:)( | 00:30 |
| inc0 | :( | 00:30 |
| inc0 | let me think | 00:30 |
| financeMann | you on a mac? | 00:30 |
| inc0 | now I'm on linux vm on top of windows connected over ssh to cloud vm on which this irc client runs | 00:30 |
| financeMann | ohh gotcha | 00:31 |
| financeMann | well full vnc access to the 64 core system is all rdy for ya ;) | 00:31 |
| financeMann | sitting in my bathroom | 00:31 |
| inc0 | lol | 00:32 |
| financeMann | its got 256gb of ram | 00:32 |
| financeMann | i threw it together to watch lots of cat videos | 00:32 |
| *** jtriley has joined #openstack-kolla | 00:32 | |
| financeMann | i dont want one single cat video to win out so i figured i’d openstack it | 00:32 |
| inc0 | address plz | 00:34 |
| *** hfu has joined #openstack-kolla | 00:36 | |
| *** tovin07 has joined #openstack-kolla | 00:39 | |
| *** tovin07_ has joined #openstack-kolla | 00:40 | |
| *** caowei has joined #openstack-kolla | 00:44 | |
| openstackgerrit | jimmygc proposed openstack/kolla master: Use install-pip macro in most source images https://review.openstack.org/413861 | 00:48 |
| *** krtaylor has joined #openstack-kolla | 00:54 | |
| *** financeMann has quit IRC | 00:58 | |
| *** cuongnv has joined #openstack-kolla | 01:00 | |
| *** bpattewar has quit IRC | 01:00 | |
| *** 07EAAAYBY has joined #openstack-kolla | 01:14 | |
| kfox1111 | sbezverk: just tried a prerendering /slightlly tweaked kolla-kubernetes keepalived on a system. | 01:17 |
| kfox1111 | it works 98%. the nopreempt flag is not working though. | 01:17 |
| sbezverk | kfox1111: maybe there was a change in software versions? | 01:18 |
| *** evilbite has quit IRC | 01:18 | |
| kfox1111 | sbezverk: maybe. been looking through a google search, and there seems to be some confusion on how it works. :/ | 01:19 |
| kfox1111 | or its different in diffrent versions. or its buggy. :/ | 01:19 |
| *** financeMann has joined #openstack-kolla | 01:21 | |
| sbezverk | kfox1111: I should be able to use master/source for just specific set of microservices right? | 01:21 |
| kfox1111 | not sure I follow the question. | 01:26 |
| kfox1111 | looks like nopreempt may want things to be in state BACKKUP in the config.... | 01:26 |
| *** sdake_ has joined #openstack-kolla | 01:27 | |
| kfox1111 | sbezverk: oh. you mean, use a master tagged image, or a source image instead of binary on specific microservcies? if thats the question, yes. | 01:28 |
| *** dave-mccowan has joined #openstack-kolla | 01:28 | |
| *** sdake has quit IRC | 01:30 | |
| financeMann | Whats the default username password on the openstack dashboard login screen http | 01:33 |
| sbezverk | kfox1111: yeah I am suspecting packaging issue for binary 3.0.2 ironic images | 01:34 |
| sbezverk | I want to try master/source for it | 01:35 |
| kfox1111 | sbezverk: ah. yeah. | 01:36 |
| kfox1111 | sbezverk: shoudl be able to change global.kolla.ironic.all.install_type=source | 01:37 |
| financeMann | sbezverk where do I find the dashboard password after i got it deployed? | 01:41 |
| financeMann | i cant figure out how to login | 01:41 |
| sbezverk | kfox1111: will also change tag | 01:41 |
| kfox1111 | sbezverk: can do that the same way. | 01:42 |
| *** eaguilar has quit IRC | 01:44 | |
| sdake_ | inc0 ironic seems to have some problems | 01:46 |
| kfox1111 | sbezverk: I think i fixed the preempt issue. | 01:46 |
| sdake_ | inc0 as in the images themselves | 01:46 |
| kfox1111 | all we need to do is change MASTER to BACKUP. | 01:46 |
| kfox1111 | it then does the right thing. | 01:46 |
| sdake_ | sean-k-mooney had said ironic worked properly | 01:47 |
| sdake_ | alhtough I don't really see that in kolla-ansible | 01:47 |
| sdake_ | atleast not the newton relesae | 01:47 |
| sdake_ | so that would be the kolla repo ;) | 01:47 |
| financeMann | sdake_ I got it deployed! I’m sitting at the dashboard login on the webinterface, how do I determine the default username/pass | 01:52 |
| financeMann | nvm got it :) | 01:58 |
| *** tonanhngo has quit IRC | 02:06 | |
| *** cuongnv_ has joined #openstack-kolla | 02:06 | |
| *** cuongnv has quit IRC | 02:06 | |
| *** duonghq has joined #openstack-kolla | 02:07 | |
| duonghq | inc0, ping | 02:08 |
| *** eaguilar has joined #openstack-kolla | 02:09 | |
| duonghq | sdake_, are you there? | 02:10 |
| *** tonanhngo has joined #openstack-kolla | 02:12 | |
| duonghq | you're streaming all session or just some? | 02:14 |
| *** tonanhngo has quit IRC | 02:17 | |
| sdake_ | shoot | 02:17 |
| sdake_ | i streamed it yeseterday | 02:17 |
| sdake_ | but its over with | 02:17 |
| sdake_ | financeMann grats dude ;) | 02:17 |
| duonghq | how about the PTG sessions? | 02:18 |
| sdake_ | financeMann the default username and apssword are admin (I think) and the password is in /etc/kolla/passwords.yml | 02:18 |
| financeMann | sdake_ thanks! it’s awesome, how do I get the dashboard to be accessible beyond localhost and on my local network? It doesn’t seem to accept outside requests to that network ip | 02:18 |
| financeMann | is that my firewall causing that issue or openstack-kolla? | 02:18 |
| *** unicell has quit IRC | 02:19 | |
| sdake_ | which issue | 02:20 |
| sdake_ | duonghq inc0 is handling ptg sessions I have no idea what he has in mind | 02:20 |
| sdake_ | duonghq i do know ptg rooms dont have projectors | 02:20 |
| financeMann | sdake_ http://192.168.1.113/admin/ works on the local machine itself but from another computer on my network it won’t work | 02:21 |
| sdake_ | duonghq so if anything it would just be an audio feed, and there is no cameras in the room (I speculate) | 02:21 |
| sdake_ | financeMann thats sort of wierd, im not really sure | 02:21 |
| sdake_ | brb | 02:21 |
| *** Pavo has joined #openstack-kolla | 02:21 | |
| duonghq | sdake_, audio is better than nothing | 02:22 |
| financeMann | sdake_ firewall issue, i took care of it | 02:22 |
| sdake_ | duonghq ya - i'd sync up with inc0 , I suggested he publish the info asap, however, he suggested he would do iton the fly | 02:25 |
| sdake_ | I think on the fly is a problem as it odesn't give people a chance to get their systems in place | 02:26 |
| *** salv-orlando has joined #openstack-kolla | 02:26 | |
| duonghq | sdake_, thank you in advanced | 02:26 |
| sdake_ | i didn't do anyting :) | 02:27 |
| sdake_ | by I'd sync up with inc0 i mean you should :) | 02:27 |
| duonghq | ack | 02:27 |
| *** cuongnv_ is now known as cuong | 02:29 | |
| sdake_ | nice my wife bought dr strange | 02:29 |
| *** cuong is now known as cuongnv | 02:29 | |
| * sdake_ watching the idiot tube tonight :) | 02:29 | |
| *** Pavo has quit IRC | 02:29 | |
| duonghq | you mean the Bluray disk or ...? | 02:29 |
| *** salv-orlando has quit IRC | 02:30 | |
| financeMann | sdake_ so I am in this user interface now and I am tryin to figure it all out. Lets say I want to get an ubuntu instance up and running now. Do I just have to create an image and thats basically my ISO and then deploy that image under the instances tab? | 02:32 |
| sdake_ | images are loaded via qsow2 | 02:33 |
| sdake_ | qcow2 | 02:33 |
| financeMann | so I cannot load up an iso via the gui? | 02:33 |
| sdake_ | you can download those from your favorite distro cloud download page | 02:33 |
| sdake_ | iso is not a thing that can be loaded in openstack no | 02:33 |
| sdake_ | gui or otherwise | 02:33 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes master: Helm charts for heat-create-db and heat-manage-db https://review.openstack.org/414496 | 02:34 |
| financeMann | sdake_ if I have a windows ISO or an ubuntu ISO what are the steps I need to take to get an instance up and running from that ISO? | 02:34 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes master: Helm charts for heat-create-db and heat-manage-db https://review.openstack.org/414496 | 02:40 |
| openstackgerrit | Merged openstack/kolla-ansible master: Change python interpreter https://review.openstack.org/434742 | 02:42 |
| *** eaguilar has quit IRC | 02:49 | |
| sbezverk | sdake: should I use 3.0.3 tag for those ironic images? or 4.0.0? | 02:59 |
| sdake_ | for maste rimages? | 03:03 |
| sdake_ | use :master tag | 03:03 |
| sdake_ | if they are stable/newton use 3.0.3 | 03:03 |
| *** cuongnv has quit IRC | 03:05 | |
| sbezverk | sdake_ whici one is master tag 4.0.0? | 03:05 |
| sdake_ | there is no master tag at present, just use master | 03:05 |
| sbezverk | like word master? | 03:06 |
| sbezverk | hm never seen it before | 03:06 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfigure action for rabbitmq https://review.openstack.org/422705 | 03:10 |
| sdake_ | use use the word master | 03:11 |
| *** ccrouch has joined #openstack-kolla | 03:12 | |
| sdake_ | Jeffrey4l ping | 03:19 |
| Jeffrey4l | sdake_, pong. | 03:19 |
| sbezverk | Jeffrey4l: if you have time, please check this PS.https://review.openstack.org/#/c/435045/ | 03:22 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for ironic https://review.openstack.org/423901 | 03:22 |
| Jeffrey4l | sbezverk, np | 03:23 |
| *** ccrouch has quit IRC | 03:24 | |
| Jeffrey4l | sbezverk, done. nit ;) | 03:25 |
| sdake_ | Jeffrey4l shall I backport that? | 03:26 |
| Jeffrey4l | i do not think so. why we need backport that? shouldn't kolla-kuberenetes pike branch depend on kolla-ansible pike? | 03:27 |
| Jeffrey4l | sdake_, ^^ | 03:27 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-ansible master: Fix genconfig for ironic-dnsmasq for kolla-kubernetes https://review.openstack.org/435045 | 03:33 |
| sbezverk | Jeffrey4l: done | 03:33 |
| sdake_ | Jeffrey4l master of kolla-kubernetes does a pip install of kolla-ansible | 03:34 |
| sdake_ | and pike kolla-ansible will not have genconfig at all | 03:34 |
| *** deepbook5broo has joined #openstack-kolla | 03:34 | |
| *** deepbook5broo has left #openstack-kolla | 03:34 | |
| Jeffrey4l | sdake_, what it mean by "pike kolla-ansible do not have genconfig"? | 03:35 |
| Jeffrey4l | it is the same with ocata branch. | 03:35 |
| *** sdake_ has quit IRC | 03:36 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix typo in all.yml https://review.openstack.org/434159 | 03:36 |
| Jeffrey4l | if kolla-kubernetes pike depends on kolla-ansible ocata, then i think we may need merge lots of patchs related to kolla-ansible related to kuberentes configs. | 03:36 |
| Jeffrey4l | sbezverk, roger. | 03:36 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Add nova-compute-ironic microservice chart https://review.openstack.org/435067 | 03:38 |
| *** sdake has joined #openstack-kolla | 03:39 | |
| sbezverk | Jeffrey4l: have you worked on ironic for kolla-ansible? | 03:39 |
| Jeffrey4l | sbezverk, yep | 03:39 |
| Jeffrey4l | i added ironic-inspector recently. | 03:40 |
| sdake | got dc'ed | 03:40 |
| sbezverk | Jeffrey4l: I see some strange things, example I could not find where in kolla ansible starts nova-compute-ironic container | 03:40 |
| Jeffrey4l | sbezverk, in nova's code. | 03:40 |
| sbezverk | Jeffrey4l: can you check this error http://logs.openstack.org/67/435067/5/check/gate-kolla-kubernetes-deploy-ubuntu-binary-2-iscsi-nv/8ff7273/logs/pods/kolla-ironic-conductor-6lmxd-main.txt | 03:41 |
| Jeffrey4l | sbezverk, check https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/nova/handlers/main.yml#L265 | 03:41 |
| Jeffrey4l | sbezverk, seems like there are some bug in ironic's code. | 03:42 |
| sbezverk | Jeffrey4l: how come I do not see any mounts for nova-compute-ironic | 03:42 |
| Jeffrey4l | sbezverk, https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/nova/defaults/main.yml#L127 | 03:43 |
| sdake | Jeffrey4l isbezverk is running the 3.0.2 images in that gate job | 03:43 |
| sbezverk | Jeffrey4l: that error was with 3.0.2 image, now trying with master | 03:43 |
| Jeffrey4l | 3.0.2 may be buggy. i never tested ironic by using 3.0.2 | 03:44 |
| sdake | sean-k-mooney said 3.0.2 was functional | 03:44 |
| sdake | although he said someone els etested it so who knwos :) | 03:44 |
| Jeffrey4l | ok. | 03:44 |
| sbezverk | damn gate cannot get to github :( | 03:44 |
| sdake | if its buggy we should fix it, as we claimed it was rock solid in newton | 03:45 |
| sdake | first step is to test master with sbezverk's code | 03:45 |
| sdake | since it actually gates | 03:46 |
| sdake | and we have no gating for ironic in kolla-ansible | 03:46 |
| *** dave-mccowan has quit IRC | 03:47 | |
| Jeffrey4l | sdake, that bug seems come from ironic's code or one python packages. bump the service version should fix it. | 03:49 |
| *** zhugaoxiao has quit IRC | 03:49 | |
| *** david-lyle has quit IRC | 03:49 | |
| sdake | Jeffrey4l so rebuilding with a new ironic z stream should do the trick? | 03:49 |
| *** zhugaoxiao has joined #openstack-kolla | 03:50 | |
| Jeffrey4l | sdake, yes. | 03:50 |
| *** david-lyle has joined #openstack-kolla | 03:50 | |
| sdake | sbezverk let me try that approach jeffrey4l just outlined | 03:50 |
| sbezverk | sdake: sure | 03:51 |
| sdake | brb, fn network | 03:51 |
| sbezverk | just make sure you build for centos and ubuntu as I have two jobs | 03:51 |
| *** cuongnv has joined #openstack-kolla | 03:52 | |
| *** sdake_ has joined #openstack-kolla | 03:53 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla master: Update running-tests.rst file https://review.openstack.org/435217 | 03:54 |
| *** sdake has quit IRC | 03:55 | |
| *** sudipto_ has joined #openstack-kolla | 03:57 | |
| *** sudipto has joined #openstack-kolla | 03:57 | |
| *** adrian_otto has joined #openstack-kolla | 03:58 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Add nova-compute-ironic microservice chart https://review.openstack.org/435067 | 03:58 |
| sdake_ | Jeffrey4l ironic 6.2.2 (in newton) is the latest version | 03:58 |
| Jeffrey4l | sdake_, hrm in the newton branch i have bump the service version. https://review.openstack.org/433554 | 03:59 |
| sdake_ | Jeffrey4l ok i'll just build the image temporarily as 3.0.3 on dockerhub then | 04:00 |
| sdake_ | when is 3.0.3 scheduled? | 04:00 |
| sdake_ | because of ocata, the whole "match the m1/m2/m3" approach we were using before is AFU | 04:00 |
| sdake_ | Jeffrey4l a long time ago doug hellman had suggested we make that file automatic, as in pull from releases.oo | 04:01 |
| Jeffrey4l | sdake_, no idea, right now. how about release 3.0.3 with kolla ocata release? | 04:01 |
| sdake_ | Jeffrey4l sounds good :) | 04:01 |
| Jeffrey4l | sdake_, yes. i hope so. and i made some change for version-check.py file. | 04:02 |
| sdake_ | i'll temporarily build an image for sbezverk to work with | 04:02 |
| Jeffrey4l | sdake_, check https://review.openstack.org/432300 | 04:02 |
| Jeffrey4l | which is more semi-automatically. | 04:02 |
| Jeffrey4l | independent project is a little hard ;( | 04:02 |
| sdake_ | Jeffrey4l that looks fantastic! | 04:03 |
| sdake_ | Jeffrey4l i'm a bit tired to actually provide a proper python review at present | 04:04 |
| sdake_ | as its 9pm | 04:04 |
| sdake_ | i'll try to review in the morning when i'm in the land of the living | 04:04 |
| Jeffrey4l | you can simple try it. just put this file into kolla ocata project tools folder and simple run it. | 04:04 |
| Jeffrey4l | most version will be updated automatically. | 04:04 |
| Jeffrey4l | np. | 04:04 |
| Jeffrey4l | sbezverk sdake_ , fyi, there are branch registry tarball file now. http://tarballs.openstack.org/kolla/images/ | 04:06 |
| sdake_ | Jeffrey4l i know - i was working on a patchset to bring that into the gate | 04:06 |
| sdake_ | and the 0.5.0 release took over priority wise | 04:06 |
| Jeffrey4l | cool. even though ocata is using openstack master code. ;) | 04:07 |
| sdake_ | that must be fixed | 04:07 |
| sdake_ | now is a good time to do so :) | 04:07 |
| sdake_ | for bot hbinary and source | 04:07 |
| Jeffrey4l | sdake_, check https://review.openstack.org/434650 | 04:08 |
| sdake_ | i guess i can help on the binary side | 04:08 |
| Jeffrey4l | oh wait. still wrong branch. | 04:08 |
| Jeffrey4l | 1 sec | 04:08 |
| sdake_ | Jeffrey4l where do you find all the time to write all this code :) | 04:09 |
| Jeffrey4l | aha, i got more time than others in one day ;) | 04:10 |
| *** sudipto has quit IRC | 04:12 | |
| *** sudipto_ has quit IRC | 04:12 | |
| *** adrian_otto has quit IRC | 04:14 | |
| *** jtriley has quit IRC | 04:20 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible master: Modify the directory path in running-tests.rst https://review.openstack.org/435229 | 04:20 |
| sdake_ | sbezverk 3.0.3 ironic images are on dockerhub pls try those | 04:23 |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible master: Modify the directory path in running-tests.rst https://review.openstack.org/435229 | 04:23 |
| sbezverk | sdake_: I will but the gate is misbehaving | 04:25 |
| sbezverk | 3.0.3 binary or source? | 04:25 |
| *** jrobinson is now known as jrobinson-afk | 04:25 | |
| *** salv-orlando has joined #openstack-kolla | 04:26 | |
| sdake_ | source | 04:28 |
| sdake_ | oh i built binary | 04:28 |
| sdake_ | let me push source | 04:28 |
| *** jtriley has joined #openstack-kolla | 04:29 | |
| sbezverk | I am off, zuul seems to be half broken :( | 04:30 |
| sbezverk | hopefully tomorrow morning it start behaving itself. bye getns | 04:30 |
| sbezverk | gents | 04:30 |
| *** salv-orlando has quit IRC | 04:31 | |
| *** mdorman has quit IRC | 04:32 | |
| *** adrian_otto has joined #openstack-kolla | 04:34 | |
| openstackgerrit | Alicja Kwasniewska proposed openstack/kolla-ansible master: Fixed a typo in tg-agent https://review.openstack.org/435235 | 04:36 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for common https://review.openstack.org/435236 | 04:37 |
| *** sudipto has joined #openstack-kolla | 04:39 | |
| *** sudipto_ has joined #openstack-kolla | 04:39 | |
| *** v1k0d3n has quit IRC | 04:40 | |
| *** mdnadeem has joined #openstack-kolla | 04:40 | |
| sdake_ | qwang your bio is missing your photograph :) | 04:42 |
| sdake_ | qwang re the session your in related to cross foundation collaboration | 04:43 |
| *** sudipto has quit IRC | 04:43 | |
| *** sudipto_ has quit IRC | 04:43 | |
| qwang | sdake_: yeah I want to keep it until the last moment:) | 04:44 |
| sdake_ | qwang i hear that - nothing like procrastination :) | 04:44 |
| *** sudipto_ has joined #openstack-kolla | 04:51 | |
| *** sudipto has joined #openstack-kolla | 04:51 | |
| sdake_ | voting is open for speakers at boston summit - please vote for kolla sessions you think are valuable: https://www.openstack.org/summit/boston-2017/vote-for-speakers/ | 04:51 |
| *** v1k0d3n has joined #openstack-kolla | 04:52 | |
| sdake_ | sbezverk ttyl - also the images are pushed for you - enjoy | 04:52 |
| *** adrian_otto has quit IRC | 04:54 | |
| *** sdake_ has quit IRC | 04:54 | |
| bmace | are the gates all backed up or something.. myself and sbezverk both did rechecks of my review after the odd failure at 5:30 and it didn't seem like either ran | 04:56 |
| *** jrobinson-afk is now known as jrobinson | 04:58 | |
| *** skramaja has joined #openstack-kolla | 04:58 | |
| *** caowei has quit IRC | 05:02 | |
| *** v1k0d3n has quit IRC | 05:05 | |
| *** ssurana has quit IRC | 05:11 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for common https://review.openstack.org/435236 | 05:12 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for common https://review.openstack.org/435236 | 05:15 |
| *** v1k0d3n has joined #openstack-kolla | 05:18 | |
| *** unicell has joined #openstack-kolla | 05:18 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for common https://review.openstack.org/435236 | 05:20 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Optimize reconfiguration for common https://review.openstack.org/435236 | 05:21 |
| *** unicell1 has joined #openstack-kolla | 05:22 | |
| *** unicell has quit IRC | 05:23 | |
| *** ccrouch has joined #openstack-kolla | 05:24 | |
| *** tyrola has quit IRC | 05:24 | |
| *** ccrouch has quit IRC | 05:25 | |
| *** tyrola has joined #openstack-kolla | 05:25 | |
| *** salv-orlando has joined #openstack-kolla | 05:27 | |
| *** salv-orlando has quit IRC | 05:31 | |
| *** qiliang27 has quit IRC | 06:07 | |
| *** qiliang27 has joined #openstack-kolla | 06:07 | |
| *** saneax-_-|AFK is now known as saneax | 06:22 | |
| *** satyar has joined #openstack-kolla | 06:37 | |
| *** yuanying has quit IRC | 06:37 | |
| *** caowei has joined #openstack-kolla | 06:42 | |
| *** zhuzeyu has joined #openstack-kolla | 06:48 | |
| *** jtriley has quit IRC | 06:49 | |
| *** fragatina has quit IRC | 07:19 | |
| *** salv-orlando has joined #openstack-kolla | 07:22 | |
| hrw | ppc64le results: | 07:23 |
| hrw | debian-binary: 132 | 07:23 |
| hrw | debian-source: 184 | 07:23 |
| hrw | ubuntu-binary: 147 | 07:23 |
| hrw | ubuntu-source: 190 | 07:23 |
| hrw | quite nice but needs work | 07:23 |
| *** jrobinson has quit IRC | 07:27 | |
| *** unicell1 has quit IRC | 07:38 | |
| *** lamt has joined #openstack-kolla | 07:55 | |
| *** shardy_afk is now known as shardy | 07:57 | |
| *** yuanying has joined #openstack-kolla | 07:57 | |
| *** bmace has quit IRC | 08:11 | |
| *** bmace has joined #openstack-kolla | 08:11 | |
| *** bmace has joined #openstack-kolla | 08:11 | |
| *** lamt has quit IRC | 08:21 | |
| *** pcaruana has joined #openstack-kolla | 08:22 | |
| *** caowei has quit IRC | 08:31 | |
| *** mgoddard has joined #openstack-kolla | 08:36 | |
| *** gfidente has joined #openstack-kolla | 08:37 | |
| *** egonzalez has joined #openstack-kolla | 08:49 | |
| *** Serlex has joined #openstack-kolla | 08:53 | |
| *** mgoddard has quit IRC | 08:56 | |
| *** pramodrj07 has joined #openstack-kolla | 08:58 | |
| *** caowei has joined #openstack-kolla | 08:59 | |
| *** MasterOfBugs has quit IRC | 09:01 | |
| openstackgerrit | jimmygc proposed openstack/kolla master: Use install-pip macro in most source images https://review.openstack.org/413861 | 09:11 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Map cell0 and sync None cell before nova database upgrade https://review.openstack.org/431508 | 09:18 |
| *** tloeb has joined #openstack-kolla | 09:20 | |
| *** athomas has joined #openstack-kolla | 09:20 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Map cell0 and sync None cell before nova database upgrade https://review.openstack.org/431508 | 09:21 |
| tloeb | hello, we are planning a full shutdown of our openstack cluster (deployed with kolla). Is there something special to take care of? | 09:25 |
| *** ccrouch has joined #openstack-kolla | 09:25 | |
| *** tovin07 has quit IRC | 09:26 | |
| *** ccrouch has quit IRC | 09:27 | |
| *** papacz has joined #openstack-kolla | 09:30 | |
| *** tovin07_ has quit IRC | 09:31 | |
| *** hrw has quit IRC | 09:42 | |
| *** Jeffrey4l has quit IRC | 09:42 | |
| *** hrw has joined #openstack-kolla | 09:44 | |
| *** zioproto has joined #openstack-kolla | 09:45 | |
| egonzalez | tloeb: ensure databases stop properly in order, not kill'em all at the same time | 09:46 |
| *** matrohon has joined #openstack-kolla | 09:48 | |
| *** mgoddard has joined #openstack-kolla | 09:51 | |
| tloeb | @egonzalez ok thanks for the response, when i restart the servers do i have to start the containers manually, will they start automatically or do i have to run kolla-ansible? | 09:51 |
| *** qiliang27 has quit IRC | 09:54 | |
| *** Jeffrey4l has joined #openstack-kolla | 09:55 | |
| *** Jeffrey4l has quit IRC | 09:56 | |
| *** mannamne has joined #openstack-kolla | 09:57 | |
| *** Jeffrey4l has joined #openstack-kolla | 09:57 | |
| *** cuongnv has quit IRC | 10:00 | |
| egonzalez | tloeb: docker by default start containers on boot | 10:00 |
| egonzalez | s/boot/service startup/ | 10:01 |
| tloeb | @egonzalez ok thank you! :) | 10:03 |
| *** haplo37_ has quit IRC | 10:03 | |
| *** haplo37_ has joined #openstack-kolla | 10:03 | |
| *** pbourke has joined #openstack-kolla | 10:04 | |
| *** rmart04 has joined #openstack-kolla | 10:05 | |
| *** duonghq has quit IRC | 10:06 | |
| *** hfu has quit IRC | 10:09 | |
| *** sambetts|afk is now known as sambetts | 10:11 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Add heat-all contaner release note https://review.openstack.org/435346 | 10:14 |
| *** Jeffrey4l has quit IRC | 10:18 | |
| *** Jeffrey4l has joined #openstack-kolla | 10:19 | |
| *** satyar has quit IRC | 10:25 | |
| openstackgerrit | jimmygc proposed openstack/kolla master: Use install-pip macro in most source images https://review.openstack.org/413861 | 10:30 |
| hrw | hm. several debian-binary fail because kolla creates users before installing packages ;( | 10:42 |
| berendt | kernel:NMI watchdog: BUG: soft lockup - CPU#25 stuck for 22s! [runc:[2:INIT]:14637] | 10:43 |
| *** l4yerffeJ has joined #openstack-kolla | 10:43 | |
| berendt | has anbody seen such an issue after deploying cinder_volume services? | 10:43 |
| berendt | we are trying to deploy on centos7 with docker 1.13.1 | 10:44 |
| *** l4yerffeJ has quit IRC | 10:44 | |
| berendt | after the start of cinder_volume on the storage nodes we have a high cpu load on runc and dockerd processes and received the watchdog issues after appr. 20 seconds | 10:44 |
| *** l4yerffeJ has joined #openstack-kolla | 10:45 | |
| egonzalez | berendt: 1.13 is not working fine, if you try to reconfigure, some containers will be removed, also issues with selinux | 10:45 |
| berendt | 1.13 was working pretty good on ubuntu | 10:45 |
| l4yerffeJ | berendt, egonzalez try upgrade. i got failed when ugprade. | 10:50 |
| *** fooliouno has quit IRC | 10:50 | |
| egonzalez | l4yerffeJ: cells? | 10:50 |
| *** l4yerffeJ has quit IRC | 10:51 | |
| *** l4yerffeJ has joined #openstack-kolla | 10:51 | |
| Jeffrey4l | egonzalez, no . i am saying docker 1.13 too. | 10:51 |
| egonzalez | yesterday I had an issue upgrading too, libvirt container was removed instead of restarted with an instance running, with 1.12 | 10:52 |
| berendt | l4yerffeJ: we fail after the deploy.. cinder_volume services starts and then the node dies | 10:52 |
| Jeffrey4l | berendt, starts and node dies? what that mean? got any logs? | 10:53 |
| berendt | i will downgrade to 1.12.6 now and check if the issue can be resolved this way | 10:53 |
| *** skramaja has quit IRC | 10:53 | |
| berendt | kernel:NMI watchdog: BUG: soft lockup - CPU#25 stuck for 22s! [runc:[2:INIT]:14637] | 10:53 |
| berendt | runc/dockerd have 100% cpu time and then we receive this kernel message | 10:53 |
| Jeffrey4l | ops. never says that. docker killed the kernel? | 10:54 |
| berendt | looks like that for me at the moment | 10:54 |
| berendt | it is even not possible to make a safe reboot via ssh :( | 10:54 |
| Jeffrey4l | never saw that. | 10:55 |
| *** athomas has quit IRC | 10:58 | |
| *** athomas has joined #openstack-kolla | 10:58 | |
| *** caowei has quit IRC | 10:59 | |
| hrw | hm. I wrote two separate mails to ML. one about non-x86 support, second about Debian support. Now, which one to reply to write about both... ;D | 11:05 |
| *** sudipto_ has quit IRC | 11:08 | |
| *** sudipto has quit IRC | 11:08 | |
| *** 07EAAAYBY has quit IRC | 11:08 | |
| *** penfold has joined #openstack-kolla | 11:14 | |
| *** sdake has joined #openstack-kolla | 11:18 | |
| sdake | morning peeps | 11:18 |
| *** egonzalez has quit IRC | 11:25 | |
| *** ccrouch has joined #openstack-kolla | 11:26 | |
| *** egonzalez has joined #openstack-kolla | 11:27 | |
| berendt | downgrade to 1.12.6 solved the issue, cinder-volume still not working because of an OSError: [Errno 2] No such file or directory: '/etc/ceph/rbdmap' issue | 11:27 |
| *** ccrouch has quit IRC | 11:28 | |
| *** magicboiz has quit IRC | 11:29 | |
| *** l4yerffeJ has quit IRC | 11:31 | |
| *** l4yerffeJ has joined #openstack-kolla | 11:31 | |
| *** l4yerffeJ has quit IRC | 11:32 | |
| *** l4yerffeJ has joined #openstack-kolla | 11:33 | |
| *** hfu has joined #openstack-kolla | 11:35 | |
| *** l4yerffeJ has quit IRC | 11:37 | |
| *** l4yerffeJ has joined #openstack-kolla | 11:37 | |
| *** rhallisey has joined #openstack-kolla | 11:47 | |
| hrw | uf. looooong mail sent to ML | 11:47 |
| hrw | with informations about non-x86 and debian support | 11:48 |
| hrw | links to all patches, github repo | 11:48 |
| *** sdake has quit IRC | 11:49 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla master: Use install-pip macro in most source images https://review.openstack.org/413861 | 11:49 |
| *** sdake has joined #openstack-kolla | 11:49 | |
| mliima | mornig guys | 11:50 |
| *** salv-orlando has quit IRC | 11:50 | |
| sbezverk | sdake: morning | 11:53 |
| sdake | sbezverk moment settingup my tb display | 11:54 |
| openstackgerrit | Saverio Proto proposed openstack/kolla-ansible master: Specify 'become' to necessary tasks (general roles) https://review.openstack.org/398682 | 12:02 |
| *** jmccarthy has quit IRC | 12:03 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix genconfig for ironic-dnsmasq for kolla-kubernetes https://review.openstack.org/435045 | 12:07 |
| *** rwallner has joined #openstack-kolla | 12:10 | |
| *** haplo37 has quit IRC | 12:13 | |
| *** dave-mccowan has joined #openstack-kolla | 12:13 | |
| *** haplo37 has joined #openstack-kolla | 12:22 | |
| *** shardy is now known as shardy_lunch | 12:23 | |
| *** tvignaud has quit IRC | 12:24 | |
| *** tvignaud has joined #openstack-kolla | 12:24 | |
| openstackgerrit | Saverio Proto proposed openstack/kolla-ansible master: Specify 'become' for only necessary tasks (default roles) https://review.openstack.org/398684 | 12:28 |
| *** sdake has quit IRC | 12:39 | |
| japestinho | morning guys | 12:44 |
| japestinho | I have working multinode kolla-ansible installation, newton based, how to deploy openstack trove for cassandra deploy inside it? | 12:46 |
| *** sdake has joined #openstack-kolla | 12:46 | |
| *** zioproto has quit IRC | 12:46 | |
| *** erlon has joined #openstack-kolla | 12:51 | |
| sdake | kfox1111 around? | 12:52 |
| sdake | sup mliima | 12:53 |
| *** athomas has quit IRC | 12:57 | |
| openstackgerrit | Saverio Proto proposed openstack/kolla-ansible master: Specify 'become' for only necessary tasks (default roles) https://review.openstack.org/398684 | 12:58 |
| *** jpstnh has joined #openstack-kolla | 12:59 | |
| *** sdake_ has joined #openstack-kolla | 12:59 | |
| *** eaguilar has joined #openstack-kolla | 13:02 | |
| *** athomas has joined #openstack-kolla | 13:02 | |
| *** sdake has quit IRC | 13:02 | |
| openstackgerrit | Merged openstack/kolla master: Use correct inventory file for Bifrost https://review.openstack.org/435058 | 13:09 |
| *** satyar has joined #openstack-kolla | 13:12 | |
| *** fguillot has joined #openstack-kolla | 13:14 | |
| *** sdake has joined #openstack-kolla | 13:14 | |
| *** sdake_ has quit IRC | 13:15 | |
| *** sudipto has joined #openstack-kolla | 13:16 | |
| *** sudipto_ has joined #openstack-kolla | 13:16 | |
| matrohon | hi all, | 13:18 |
| *** yuanying has quit IRC | 13:18 | |
| *** rwallner has quit IRC | 13:18 | |
| *** Pavo has joined #openstack-kolla | 13:20 | |
| *** satyar has quit IRC | 13:20 | |
| matrohon | does anybody heard about a bug related to a neutron creating two ports when only one has been asked during "nova boot"? | 13:21 |
| *** rwallner has joined #openstack-kolla | 13:21 | |
| matrohon | several deployments made with kolla (newton and mitaka) hit that bug | 13:22 |
| *** salv-orlando has joined #openstack-kolla | 13:22 | |
| sdake | matrohon i have seen this problem in older versions of openstack | 13:22 |
| rwallner | using stable/newton, i've changed all configuration to use gnocchi as backend for ceilometer. All services seem to be up and running, but 504 gateway timeouts on meter-list, as well as ceilometer still trying to connect to mongodb what i cat /var/lib/docker/volumes/kolla_logs/_data/ceilometer/app.wsgi.log | 13:23 |
| rwallner | any known issues here? | 13:23 |
| matrohon | sdake, do you think about : https://bugs.launchpad.net/kolla/mitaka/+bug/1565105 | 13:24 |
| openstack | Launchpad bug 1565105 in kolla mitaka "while spining vms on large scale some vms get multiple ips and multiple ports get created" [Critical,Fix committed] - Assigned to Steven Dake (sdake) | 13:24 |
| sdake | matrohon yup that is a real issue I saw iti in a real live cluster and got confirmation from an intel dude that it fixed the problem | 13:25 |
| sdake | matrohon what type of scale are you using? | 13:25 |
| sdake | you may have to increase your max pool size | 13:25 |
| matrohon | sdake : really small :( | 13:25 |
| sdake | matrohon how many nodes | 13:26 |
| sdake | and how many vms | 13:26 |
| sdake | and how areoyu launchin gthem | 13:26 |
| *** ccrouch has joined #openstack-kolla | 13:27 | |
| matrohon | I had the issue with kolla 3.0.3 while spawning 20 VMs on a cluster of 3 nodes, launching VMs with concurrent horizon connections | 13:27 |
| sdake | there is no kolla 3.0.3? | 13:28 |
| sdake | however your launching from horizon - horizon communicates very quickly with neutron and exposes this max queue limit bug | 13:28 |
| sdake | its a defect in neutron as far as I'm concerned | 13:28 |
| sdake | however neutron doesn't see it that way - so ymmv :) | 13:29 |
| *** ccrouch has quit IRC | 13:29 | |
| sdake | its definately not a bug in kolla - its a config problem at most | 13:29 |
| matrohon | ok thanks, I'll try to increase the limit, hopefully it will fix the issue | 13:30 |
| sdake | Jeffrey4l ping - stable/newton fails to build nova-novnc proxy - i have a backtrace | 13:30 |
| *** eaguilar has quit IRC | 13:30 | |
| sdake | matrohon that sounds good | 13:31 |
| financeMann | Good morning! Now that I have openstack kolla finally deployed what are the steps I need to follow to get an ISO deployed to an instance? | 13:31 |
| sdake | financeMann you cannot deploy an iso dude | 13:31 |
| sdake | financeMann you can only deploy qcow2 images - this is an openstack fundametal thing - has nothign to do with kolla | 13:31 |
| *** shardy_lunch is now known as shardy | 13:32 | |
| sdake | financeMann fortunately part of what you want is documented here: https://docs.openstack.org/user-guide/dashboard-launch-instances.html | 13:32 |
| rwallner | does ceilometer + gnocchi config still need mongo? not sure why after fresh kolla deployment there is still logs trying to connect to mongo | 13:33 |
| *** Pavo has quit IRC | 13:35 | |
| financeMann | excellent sdake, I understand you cant deploy ISOs I am just saying I am at point A) I have an ISO and want to get to point E) Deployed openstack image from that ISO somehow | 13:39 |
| financeMann | I am assuming thats possibly by deploying the ISO to qcow2 first | 13:40 |
| egonzalez | financeMann: create a VM in KVM with that ISO, install cloud-init, cleanup network stuff and use the qcow2 file created by kvm | 13:40 |
| egonzalez | financeMann: https://docs.openstack.org/image-guide/create-images-manually.html | 13:41 |
| financeMann | egonzalez does qcow2 come with some defaults? | 13:41 |
| egonzalez | qcow2 is a disk format, the rest of the process is to make the image "cloud ready" | 13:41 |
| financeMann | egonzalez does openstack come with any default images like ubuntu 16.04 etc all ready to deploy | 13:44 |
| financeMann | or does a person always have to create these manually up-front | 13:44 |
| *** jheroux has joined #openstack-kolla | 13:44 | |
| egonzalez | financeMann: yeah, there are some on the internet, if is just a test to check if all is running use cirros a 13Mb lightweight image | 13:45 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Helm Ceph Initial Mon https://review.openstack.org/410641 | 13:45 |
| financeMann | egonzalez yea that’d be nice to have that so I can just get a better idea for how to deploy and such | 13:45 |
| financeMann | egonzalez I am assuming get it from here https://docs.openstack.org/image-guide/obtain-images.html | 13:46 |
| rwallner | best place for questions on kolla ceilometer+gnocchi deployment? anyone know. Dont mean to plop questions here if they | 13:47 |
| rwallner | are in the wrong place | 13:47 |
| egonzalez | financeMann: yep | 13:48 |
| des_consolado | the way nova is dealt with in terms of the inventory and the role is a little strange, isn't it? In the inventory everything is under the 'nova' name which is put onto whatever kind of node, control by default - but then in the role the compute, libvirt, ssh services are all hard-coded to be in the compute group | 13:48 |
| financeMann | egonzalez so when I have the img file, how do I upload that to my openstack instance for deployment? | 13:52 |
| egonzalez | financeMann: through horizon or through CLI https://docs.openstack.org/user-guide/common/cli-manage-images.html#create-or-update-an-image-glance | 13:55 |
| rwallner | possibly gnocchi/ceilometer is just broken in newton, but im not seeing the specified error here https://bugs.launchpad.net/kolla/newton/+bug/1635307 | 13:56 |
| openstack | Launchpad bug 1635307 in kolla newton "gnocchi dispatcher fails on stable/netwon" [Critical,Confirmed] - Assigned to Jeffrey Zhang (jeffrey4l) | 13:56 |
| financeMann | egonzalez in horizon image upload I notice it has a specification for selecting Format and ISO is in there | 13:57 |
| egonzalez | financeMann: yep, because some hypervisors support it | 13:57 |
| financeMann | but not kolla egonzalez? | 13:58 |
| sdake | sbezverk hunting down your binary and source of 3.03 I noticed the following - nova-novncproxy is ONLY binary: https://github.com/openstack/kolla/blob/stable/newton/kolla/common/config.py#L276-L279 | 13:58 |
| sdake | Jeffrey4l also building source versions of novncproxy and spicehtml proxy crater the build script in stable/newton | 13:58 |
| *** mgiles has joined #openstack-kolla | 13:59 | |
| *** goldyfruit has joined #openstack-kolla | 14:00 | |
| egonzalez | financeMann: kolla only support KVM for now (xen is comming) | 14:00 |
| sdake | rwallner pretty Jeffrey4l got ceilometer + gnocchi working personally in newton | 14:01 |
| sdake | rwallner he is in apac timezone so he may not be around at this time | 14:01 |
| sdake | rwallner the best forum for asking thatquestion is ask.openstack.org that way the entire openstack community can respond | 14:02 |
| rwallner | thanks sdake, will try and ping him later on. | 14:02 |
| sdake | rwallner if you get an answer please put iit in ask.openstack.org so we don't have to answer it a million times plz :) | 14:03 |
| sdake | or ask it in ask.openstack.org and have Jeffrey4l answer it there | 14:03 |
| rwallner | sounds good | 14:04 |
| *** krtaylor has quit IRC | 14:06 | |
| *** mdnadeem has quit IRC | 14:09 | |
| financeMann | this is so sweet, got my first deploy instance rolling, love it, now my wife can never complain she wants a new faster computer | 14:17 |
| financeMann | and the bathroom will always be nice and warm with the new heater in there | 14:18 |
| financeMann | just need to throw a small solar panel on the roof to sleep at peace at night | 14:18 |
| *** goldyfruit has quit IRC | 14:28 | |
| *** jtriley has joined #openstack-kolla | 14:28 | |
| *** goldyfruit has joined #openstack-kolla | 14:31 | |
| *** hfu has quit IRC | 14:32 | |
| *** tloeb has quit IRC | 14:36 | |
| Jeffrey4l | sup sdake | 14:39 |
| sdake | Jeffrey4l i noticed building nova-novncproxy and nova-spicehtml5 fail on source - is this intended behavior? | 14:41 |
| sdake | in stable/newton | 14:41 |
| sdake | also the docker hub does not contian these images | 14:41 |
| Jeffrey4l | sdake, think so. got any logs? | 14:41 |
| Jeffrey4l | think no*** | 14:41 |
| sdake | i can post them moment | 14:41 |
| Jeffrey4l | i run a test patch recently, which is green. | 14:42 |
| sdake | hrm, fpaste is broken | 14:42 |
| sdake | is there any way to paste to a paste.openstack.org from the cli a file? | 14:43 |
| Jeffrey4l | https://review.openstack.org/417033 all build gate are green. deploy gate failed unrelated. | 14:43 |
| Jeffrey4l | sdake, no idea.. | 14:43 |
| sdake | LTGT :) | 14:43 |
| sdake | moment | 14:43 |
| sdake | LMGT that is | 14:43 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla stable/newton: [TEST][NEWTON][KOLLA]Test ceilometer related project in gate https://review.openstack.org/417283 | 14:47 |
| *** tyrola has quit IRC | 14:48 | |
| *** tyrola has joined #openstack-kolla | 14:49 | |
| financeMann | It’s been a while since I’ve had to setup networks, if I have my local network will I basically want to setup an openstack router that routes traffic from my local network to an openstack network with it’s own subnet? | 14:50 |
| financeMann | I keep getting this error on deploy “No valid host was found. There are not enough hosts available" | 14:50 |
| financeMann | I’m using the cirrOS set as QCOW2 image | 14:51 |
| Serlex | financeMann, look into nova_compute container logs. You probably want to run the same action and follow the log file at the same time | 14:52 |
| Jeffrey4l | rwallner, i leave comments in https://bugs.launchpad.net/kolla/newton/+bug/1635307 ceilometer + gnocchi works. | 14:52 |
| openstack | Launchpad bug 1635307 in kolla newton "gnocchi dispatcher fails on stable/netwon" [Critical,Incomplete] - Assigned to Jeffrey Zhang (jeffrey4l) | 14:52 |
| Serlex | Quick check would be to see if you have host servers online, either via Horizon or CLI | 14:52 |
| sdake | Jeffrey4l http://ix.io/dIe | 14:53 |
| financeMann | Serlex does this setup look good though? http://imgur.com/a/HP8ks | 14:53 |
| Jeffrey4l | sdake, https://bugs.launchpad.net/kolla/newton/+bug/1635307 | 14:53 |
| openstack | Launchpad bug 1635307 in kolla newton "gnocchi dispatcher fails on stable/netwon" [Critical,Incomplete] - Assigned to Jeffrey Zhang (jeffrey4l) | 14:53 |
| Jeffrey4l | sdake, cat: typecript: No such file or directory | 14:53 |
| sdake | ya i see moment | 14:53 |
| rwallner | hmmm, Jeffrey4l i did a fresh install today, but it seems most logs and configs seems OK, besides wsgi.app.log logging that its still trying to connect to mongo, but config is enable_gnocchi: "yes", ceilometer_database_type: "gnocchi", and gnocchi_backend_storage: "{{ 'ceph' if enable_ceph|bool else 'file' }}" where i do have a ceph cluster | 14:54 |
| Serlex | financeMann: follow the container logs while creating an instance, makes life so much easier | 14:55 |
| sdake | Jeffrey4l http://ix.io/dMk | 14:55 |
| financeMann | Serlex is that something I can follow in the dashboards instance log section or do I have to CLI | 14:55 |
| rwallner | _reading your comments_ | 14:55 |
| Serlex | cli | 14:55 |
| Serlex | something like tail -f nova_compute or nova_conductor | 14:55 |
| Jeffrey4l | rwallner, could u paste the gnocchi.conf file out? | 14:57 |
| Jeffrey4l | sdake, looks like you kolla running env got some issue. could u try to run kolla in a new env? | 14:58 |
| Jeffrey4l | sdake, install kolla requirements along with upper-containers.txt file. | 14:58 |
| rwallner | Jeffrey4l http://pastebin.com/AKcGL02x | 14:58 |
| sdake | Jeffrey4l i can try the second but no thte first | 14:58 |
| Jeffrey4l | rwallner, wait. when you are saying wsgi.app.log file, which service generete this file? | 14:59 |
| sdake | Jeffrey4l I see no upper constraints in tox.ini in stable/newton | 14:59 |
| sdake | Jeffrey4l how would I go about installing kolla with upper constraints correctly? | 15:00 |
| Jeffrey4l | hrm, you can add this manually. | 15:00 |
| sdake | Jeffrey4l I did a pip install . on stable/newton | 15:00 |
| Jeffrey4l | sdake, are u using tox to build image? | 15:00 |
| sdake | nope | 15:00 |
| Jeffrey4l | or build.py directly? | 15:00 |
| sdake | I was looking for an exmaple of upper constraints usage | 15:00 |
| sdake | build.py directly | 15:00 |
| Jeffrey4l | ok. run pip install -c <path/to/upper-constrains.txt/file> kolla | 15:01 |
| rwallner | Jeffrey4l, ceilometer seems to be producing it. `ceilometer/app.wsgi.log` | 15:01 |
| rwallner | i've changed mongodb to `no` so its expected, but dont know why its trying to connect. | 15:02 |
| Jeffrey4l | rwallner, ok. then paste the ceilometer.conf file out, please. ;) | 15:02 |
| Jeffrey4l | rwallner, have u eve restart ceilometer after configured ceilometer_database_type to gnocchi? | 15:03 |
| rwallner | http://pastebin.com/zUGYkcFK Jeffrey4l | 15:03 |
| rwallner | yes, this was a fresh kolla deploy with gnocchi / ceilometer changes | 15:04 |
| sdake | Jeffrey4l that fixed it thakns! | 15:05 |
| Jeffrey4l | rwallner, and the wsgi.app.log file? | 15:05 |
| Jeffrey4l | sdake, np ;) | 15:05 |
| sdake | Jeffrey4l question about people pip installing newton, will the upper constraints be applied? | 15:05 |
| Jeffrey4l | sdake, no. and we also can not control this. | 15:06 |
| Jeffrey4l | all openstack project has such kind of issue. | 15:06 |
| Jeffrey4l | too. | 15:06 |
| Jeffrey4l | in requirements.txt file, most critiria is looks like: xx >= x.y.z one day, if x.y.z+1 broken, then that project broken. | 15:07 |
| sdake | Jeffrey4l zug | 15:08 |
| sdake | kolla solves this problem by always using upper constraints | 15:09 |
| sdake | however kolla itself doesn't use upperconstraints during a pip install | 15:09 |
| Jeffrey4l | yep. using upper constrains is more safer. | 15:09 |
| sdake | any way to resolve that in kolla? | 15:09 |
| sdake | sbezverk note: all 3.0.3 images you need are ready: e.g. https://hub.docker.com/r/kolla/centos-source-nova-novncproxy/tags/ | 15:09 |
| Jeffrey4l | hold one day, python packages is distrubute along with the upper constraint.txt file. | 15:10 |
| Jeffrey4l | hope one day* | 15:10 |
| Jeffrey4l | sdake, add doc .. | 15:10 |
| sdake | docs sound good | 15:11 |
| sdake | sayantani01 ^^ | 15:11 |
| Jeffrey4l | pip install -c https://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt kolla | 15:12 |
| Jeffrey4l | this is more safer^^ | 15:12 |
| *** lamt has joined #openstack-kolla | 15:14 | |
| Jeffrey4l | sdake, re the automation of bump openstack service version by using bot. how about add a check/bot in gate, when some project is tagged, run the jobs, and check kolla's code. if the version need to be update, then push a patch automatically. just like how the openstack/requirements bot does. | 15:16 |
| financeMann | Serlex according to the error it makes me feel like it’s an image issue? http://pastebin.com/b73v1Rmp | 15:18 |
| financeMann | "Cannot find OS information - Reason: (No configuration information found for operating system Empty)" | 15:18 |
| NachoDuck | Hey, just wondering if anyone has seen something like this before. Kolla Newton deployment having some issues launching instances. seeing messages similar to "No hosts available" on a 3 node deployment. Looking at nova service list I can see that the services are constantly flicking between up and down status | 15:18 |
| financeMann | that’s the first error its sayin why it cant deploy cirrOS | 15:18 |
| sdake | Jeffrey4l sounds cool - no ide ahow to implement that although infra may have an idea | 15:21 |
| Jeffrey4l | ok. i will sync with them and try to find out how to implement that. | 15:21 |
| sdake | Jeffrey4l or find another core to do it :) | 15:21 |
| sdake | Jeffrey4l you can't do it alone :) | 15:21 |
| Jeffrey4l | the version-check script is almost done. so the left work is add something in project-config. | 15:22 |
| sdake | Jeffrey4l cool you intend to make that voting? | 15:23 |
| sdake | after i t goes through experimental and nonvoting? | 15:23 |
| Serlex | financeMann, thats just a warning | 15:23 |
| *** salv-orlando has quit IRC | 15:23 | |
| financeMann | Serlex I see all sorts of ERROR status for it though i thought | 15:23 |
| Jeffrey4l | this is not a job in kolla project. it will be a check script when some one openstack project is tagged. and finally, this may push a patch for kolla. | 15:24 |
| Serlex | Have you checked if you have hosts (hypervisors) online? | 15:24 |
| sdake | Jeffrey4l i see that sounds cool | 15:24 |
| sdake | Jeffrey4l doug might be able to help figure out how to automate that | 15:24 |
| *** saneax is now known as saneax-_-|AFK | 15:24 | |
| sdake | or know the correct people that do | 15:25 |
| Jeffrey4l | just like this https://review.openstack.org/434205 | 15:25 |
| sdake | since it sound wlike you want to introduce somethign into the release process | 15:25 |
| Jeffrey4l | ya. i will ask for help when implementing this. | 15:25 |
| financeMann | Serlex doesn’t look like it -> http://imgur.com/a/T1j1Y | 15:28 |
| *** ccrouch has joined #openstack-kolla | 15:32 | |
| rwallner | Jeffrey4l sorry was in a meeting, here is the wsgi log http://pastebin.com/CUh54AwS | 15:34 |
| Serlex | that looks good to me financeMann | 15:34 |
| financeMann | oh Serlex i thought it looked like it was missing the hypervisor | 15:35 |
| Jeffrey4l | roger. | 15:35 |
| financeMann | Serlex I haven’t really done a good job of setting up my network and routers, could that cause this issue? I doubt it | 15:35 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla master: Updated from global requirements https://review.openstack.org/431961 | 15:35 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Fix test_build_rpm_setup failure after changed rpm_setup_config https://review.openstack.org/435493 | 15:36 |
| Serlex | Minimum you need is one private network | 15:36 |
| Jeffrey4l | rwallner, seems like configured a default database connection when it is configured. | 15:36 |
| Serlex | no need for router or external network. You are definitely using kolla-openstack right? | 15:36 |
| financeMann | ya inc0 installed it for me lol | 15:37 |
| financeMann | was having a lot of troubles | 15:37 |
| Serlex | go for default flavour like m1.tiny for cirros image, I see that you are using a custom flavour | 15:38 |
| financeMann | Serlex -> http://imgur.com/a/ZVUco | 15:39 |
| financeMann | hmm i’ll try another cirros image | 15:39 |
| *** pbourke has quit IRC | 15:40 | |
| *** pbourke has joined #openstack-kolla | 15:40 | |
| Serlex | Sure. Try default flavour for now, nothing customized | 15:41 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Get latest versions from openstack/releases repo https://review.openstack.org/432300 | 15:42 |
| *** Pavo has joined #openstack-kolla | 15:42 | |
| *** pbourke has quit IRC | 15:43 | |
| rwallner | Jeffrey4l, thats what's got me, not sure where that config it coming from? | 15:43 |
| Jeffrey4l | rwallner, it should come from ceilometer default value. i will check the code and try to fix it. | 15:44 |
| Jeffrey4l | you are testing 3.0.2 right? rwallner | 15:44 |
| *** crushil has joined #openstack-kolla | 15:46 | |
| openstackgerrit | James McCarthy proposed openstack/kolla master: Add helm_repository_source_install_python_pip block to helm-repository Dockerfile https://review.openstack.org/435499 | 15:47 |
| rwallner | yes 3.0.2 Jeffrey4l | 15:47 |
| Jeffrey4l | roger thanks for feedback ;) | 15:48 |
| *** wwriverrat has quit IRC | 15:48 | |
| *** ccrouch has quit IRC | 15:49 | |
| *** fooliouno has joined #openstack-kolla | 15:49 | |
| *** krtaylor has joined #openstack-kolla | 15:49 | |
| fooliouno | Hello .. reading the kolla-kubernetes mulit-node steup guide, it appears that only a handful of instances are currently supported (such as keystone, horizon and mariadb). Are others such as nova, neutron, glance, cinder etc not supported at this time? | 15:50 |
| *** pbourke has joined #openstack-kolla | 15:51 | |
| fooliouno | Is Ceph a requirement for kolla-kubernetes or can I use the onboard storage of my servers? | 15:51 |
| *** kbyrne has joined #openstack-kolla | 15:52 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Get latest versions from openstack/releases repo https://review.openstack.org/432300 | 15:53 |
| *** pbourke has quit IRC | 15:55 | |
| *** rwallner has quit IRC | 15:55 | |
| *** rwallner has joined #openstack-kolla | 15:58 | |
| financeMann | Serlex I deployed the cow image of centos, same issue -> http://pastebin.com/XPFgk9qi | 16:00 |
| mgoddard | Jeffrey4l: did you mean to recheck https://review.openstack.org/#/c/435361/? It had passed previously but needs reviews | 16:00 |
| Serlex | ok, Can you look into nova_conductor container? | 16:01 |
| Jeffrey4l | mgoddard, there are three job failed. see? | 16:01 |
| Jeffrey4l | recheck will tigger the jobs run again. | 16:01 |
| Jeffrey4l | it is a magic workd ;) | 16:01 |
| Serlex | Those errors don't mean much to me, someone else want to take a look? | 16:02 |
| *** kbyrne has quit IRC | 16:04 | |
| *** sdake has quit IRC | 16:04 | |
| Jeffrey4l | financeMann, disable selinux. | 16:05 |
| Serlex | ahh ha...didnt think of that | 16:05 |
| Jeffrey4l | i encounter this before ;) | 16:05 |
| *** sdake has joined #openstack-kolla | 16:06 | |
| *** kbyrne has joined #openstack-kolla | 16:07 | |
| mgoddard | Jeffrey4l: I see, wasn't looking at the non-voting jobs :) | 16:07 |
| *** jtriley_ has joined #openstack-kolla | 16:07 | |
| mgoddard | Jeffrey4l: does the CI test bifrost? | 16:08 |
| Jeffrey4l | mgoddard, yep. kolla-ansible do not have much voting jobs. But all the jobs are expected to be green. | 16:08 |
| Jeffrey4l | mgoddard, actually no ;( | 16:08 |
| Jeffrey4l | but it is a good habit to make sure every patch is passed the gate. | 16:08 |
| *** pbourke has joined #openstack-kolla | 16:09 | |
| mgoddard | Jeffrey4l: based on my previous experience, P(job succeeds) << 1/Njobs | 16:09 |
| mgoddard | Jeffrey4l: it took 3 rechecks to get the first patch through :) | 16:09 |
| Jeffrey4l | mgoddard, sorry what that means? | 16:09 |
| Jeffrey4l | really? hrm | 16:09 |
| Jeffrey4l | in most the time, kolla's jobs is reliable. | 16:10 |
| Jeffrey4l | but we are facing lots of unpredictable issue. | 16:10 |
| *** jtriley has quit IRC | 16:10 | |
| mgoddard | Jeffrey4l: perhaps it's busy (==unreliable) due to releases | 16:10 |
| financeMann | Jeffrey4l Serlex thanks for your help, fully deployed with console access, so sweet | 16:10 |
| *** Pavo has quit IRC | 16:11 | |
| Jeffrey4l | openstack infra gate failed few hours ago. | 16:11 |
| Jeffrey4l | financeMann, nice. | 16:11 |
| Jeffrey4l | need some selinux expert to fix the issue. i think this should not be a big issue. | 16:11 |
| Jeffrey4l | sdake, have any idea about selinux issue ^^ | 16:12 |
| *** rwallner has quit IRC | 16:16 | |
| financeMann | my brought my wife over to show her the deploy and she said “Why do you have to make everything so complicated" | 16:20 |
| financeMann | told her so her cat videos dont slow down mine | 16:21 |
| *** pcaruana has quit IRC | 16:21 | |
| Jeffrey4l | kolla deployment is complicated? | 16:22 |
| financeMann | she said we should just to go best buy and buy computers lol | 16:22 |
| financeMann | she said the three of us dont need our own datacenter in the bathroom | 16:23 |
| *** salv-orlando has joined #openstack-kolla | 16:24 | |
| *** Pavo has joined #openstack-kolla | 16:24 | |
| *** wwriverrat has joined #openstack-kolla | 16:25 | |
| *** rmart04 has quit IRC | 16:25 | |
| *** salv-orlando has quit IRC | 16:29 | |
| sbezverk | Jeffrey4l: with 3.0.3 images ironic well at least control plane work on kolla-kube too :-) btw thank you for helping to get genconfig changes quickly!! | 16:29 |
| Jeffrey4l | sbezverk, cool. | 16:29 |
| openstackgerrit | Merged openstack/kolla-ansible master: Support kolla-ansible installed in a virtualenv https://review.openstack.org/434917 | 16:29 |
| *** wwriverrat has left #openstack-kolla | 16:31 | |
| *** tonanhngo has joined #openstack-kolla | 16:39 | |
| *** sudipto_ has quit IRC | 16:39 | |
| *** sudipto has quit IRC | 16:39 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Adding Ironic microservices packages https://review.openstack.org/433322 | 16:41 |
| bmace | i know enough about selinux to make the first thing I do after a fresh install disabling it :) | 16:42 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Add nova-compute-ironic microservice chart https://review.openstack.org/435067 | 16:42 |
| sdake | bmace lol | 16:43 |
| sdake | bmace thats about my skill level with selinux too | 16:43 |
| sdake | bmace i'm one step beyond you - I know how to read an audit log and tell what is broken | 16:44 |
| sdake | but not necessarily know how to fix it :) | 16:44 |
| bmace | :) | 16:44 |
| * sdake hates when he knives his hand | 16:45 | |
| sdake | big ass cut on my finger - neosporin ftw | 16:45 |
| *** pramodrj07 has quit IRC | 16:46 | |
| *** MasterOfBugs has joined #openstack-kolla | 16:46 | |
| bmace | try out liquid band-aids? smells terrible but works really well in places that normal band-aids don't work well. | 16:47 |
| *** rwallner has joined #openstack-kolla | 16:47 | |
| *** egonzalez has quit IRC | 16:47 | |
| *** rwallner has quit IRC | 16:48 | |
| *** pbourke has quit IRC | 16:48 | |
| fooliouno | trying my questions again :) reading the kolla-kubernetes multi-node setup guide, it appears that only a handful of instances are currently supported (such as keystone, horizon and mariadb). Are others such as nova, neutron, glance, cinder etc not supported at this time? Is Ceph a requirement for kolla-kubernetes or can I use the local storage of my servers? | 16:50 |
| openstackgerrit | Javier Peña proposed openstack/kolla master: Add work-dir command-line option https://review.openstack.org/435527 | 16:50 |
| *** laki has joined #openstack-kolla | 16:54 | |
| laki | hello | 16:54 |
| laki | has anyone faced such issue ? virSecurityDriverLookup:80 : unsupported configuration: Security driver apparmor not enable | 16:54 |
| *** japestinho_ has joined #openstack-kolla | 17:01 | |
| laki | okay - what is the easiest method to remove whole configuration and rebuild it from 0 ? | 17:03 |
| laki | moving whole config from etc and then kolla-ansible deploy ? | 17:04 |
| *** jpstnh has quit IRC | 17:04 | |
| laki | let's try | 17:05 |
| *** nhlfr has quit IRC | 17:06 | |
| *** nhlfr has joined #openstack-kolla | 17:08 | |
| *** rwallner has joined #openstack-kolla | 17:17 | |
| *** crushil has quit IRC | 17:17 | |
| *** lamt has quit IRC | 17:24 | |
| *** Serlex has quit IRC | 17:25 | |
| *** athomas has quit IRC | 17:26 | |
| *** unicell has joined #openstack-kolla | 17:31 | |
| *** unicell has quit IRC | 17:31 | |
| *** gfidente has quit IRC | 17:32 | |
| *** mgoddard has quit IRC | 17:32 | |
| *** lamt has joined #openstack-kolla | 17:33 | |
| *** zhubingbing has joined #openstack-kolla | 17:34 | |
| *** eaguilar has joined #openstack-kolla | 17:36 | |
| *** salv-orlando has joined #openstack-kolla | 17:40 | |
| *** bpattewar_ has joined #openstack-kolla | 17:46 | |
| *** ssurana has joined #openstack-kolla | 17:48 | |
| openstackgerrit | Javier Peña proposed openstack/kolla master: Add work-dir command-line option https://review.openstack.org/435527 | 17:49 |
| *** matrohon has quit IRC | 17:50 | |
| *** jtriley_ has quit IRC | 17:51 | |
| *** ssurana has quit IRC | 17:52 | |
| *** jtriley has joined #openstack-kolla | 17:54 | |
| *** ssurana has joined #openstack-kolla | 17:57 | |
| *** rwallner has quit IRC | 18:00 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Stagger mariadb / rabbitmq start to avoid cross storage locks https://review.openstack.org/435151 | 18:00 |
| openstackgerrit | Merged openstack/kolla master: Re-enable grafana gpgcheck https://review.openstack.org/434129 | 18:02 |
| *** matrohon has joined #openstack-kolla | 18:02 | |
| *** unicell has joined #openstack-kolla | 18:03 | |
| *** mgoddard has joined #openstack-kolla | 18:06 | |
| *** zhubingbing has quit IRC | 18:10 | |
| *** salv-orl_ has joined #openstack-kolla | 18:10 | |
| *** sdake_ has joined #openstack-kolla | 18:12 | |
| *** salv-orlando has quit IRC | 18:12 | |
| *** sdake has quit IRC | 18:14 | |
| *** Pavo has quit IRC | 18:19 | |
| *** mannamne has quit IRC | 18:22 | |
| *** eaguilar has quit IRC | 18:27 | |
| *** lamt has quit IRC | 18:28 | |
| *** rwallner has joined #openstack-kolla | 18:28 | |
| *** crushil has joined #openstack-kolla | 18:30 | |
| *** sambetts is now known as sambetts|afk | 18:42 | |
| *** MasterOfBugs has quit IRC | 18:42 | |
| -openstackstatus- NOTICE: Restarting gerrit due to performance problems | 19:03 | |
| *** MasterOfBugs has joined #openstack-kolla | 19:06 | |
| *** sdake_ has quit IRC | 19:10 | |
| *** fguillot has quit IRC | 19:12 | |
| *** mgoddard has quit IRC | 19:12 | |
| *** bjolo_ has joined #openstack-kolla | 19:15 | |
| openstackgerrit | David Moreau Simard proposed openstack/kolla-ansible master: Replace "ara generate" by "ara generate html" https://review.openstack.org/433204 | 19:15 |
| *** aimeeu has quit IRC | 19:18 | |
| *** ccrouch has joined #openstack-kolla | 19:18 | |
| *** aimeeu has joined #openstack-kolla | 19:18 | |
| *** mliima has quit IRC | 19:57 | |
| *** crushil has quit IRC | 20:07 | |
| *** ssurana has quit IRC | 20:11 | |
| kfox1111 | not sure what I hate more these days. ups's, or their lack. :/ | 20:11 |
| *** intlabs is now known as portdirect | 20:13 | |
| portdirect | kfox1111: that statement, imply's the worst sort of Friday :( | 20:14 |
| kfox1111 | indeed. | 20:14 |
| kfox1111 | we're on city power right now because our ups died. but then had a power bump this morning. :/ | 20:15 |
| portdirect | :( | 20:15 |
| kfox1111 | so the current set of activities was related to a lack of ups because of a ups. :) | 20:16 |
| sbezverk | kfox1111: I got ironic api/conductor running.. there are some leftovers from debugging but I would like to merge these two PS soon | 20:16 |
| sbezverk | can you start reviewing | 20:16 |
| kfox1111 | sbezverk: have you thought about testing it yet? | 20:16 |
| *** lamt has joined #openstack-kolla | 20:17 | |
| kfox1111 | sbezverk: gona have to wait a little bit. trying to recover an important fs. :/ | 20:17 |
| sbezverk | testing it done by deploying and creating emty node | 20:17 |
| sbezverk | other type of testing is way more difficult to implement.. it will not be done quickly | 20:17 |
| portdirect | kfox1111: hope your weekend brightens up - look forward to catching you (and you again sbezverk too!) next week :) | 20:18 |
| kfox1111 | sbezverk: in gate, or on one of your test machines? | 20:18 |
| kfox1111 | portdirect: thanks, and +1. :) | 20:18 |
| sbezverk | so my plan is to get api/conductor merged, then address inpector and then discuss if a separate experimental job for ironic is required | 20:18 |
| *** portdirect is now known as portdirect_travl | 20:18 | |
| kfox1111 | will be good to finally meet you guys in person. :) | 20:18 |
| sbezverk | kfox1111: in the agte | 20:19 |
| portdirect_travl | +2 :) | 20:19 |
| sbezverk | gate | 20:19 |
| sbezverk | portdirect_travl: safe travle man!! | 20:19 |
| sbezverk | travel | 20:19 |
| kfox1111 | sbezverk: I see it launched in gate, but not used? maybe I'm misunderstanding. | 20:19 |
| sbezverk | kfox1111: I have it lauched at the end of iscsi job | 20:20 |
| kfox1111 | iscis_workflow.sh? | 20:21 |
| sbezverk | yes | 20:21 |
| kfox1111 | I see it helm installed, but nothing testing it? | 20:21 |
| sbezverk | kfox1111: testing is not done as I mentioend earlier | 20:21 |
| sbezverk | it would require significant effort | 20:21 |
| kfox1111 | ah. ok. | 20:22 |
| sbezverk | and I am not sure if there will be enough interest from community to have a separate experimental job | 20:22 |
| sbezverk | testing ironic, it can be done but there should be a "business" reason to do it | 20:23 |
| *** laki has quit IRC | 20:23 | |
| kfox1111 | buisness reason is clear enough to me. if its not tested, how do we know we don't rbreak it? | 20:23 |
| kfox1111 | just helm installing it doesn't prove that it works at all other then doesn't crash. | 20:23 |
| sbezverk | kfox1111: it starts ;-) | 20:23 |
| sbezverk | and does not crash | 20:23 |
| kfox1111 | there are lots of ways to cause a complicated stack of sofware like openstack to start but be totally unusable. :/ | 20:24 |
| kfox1111 | more ways then to get it to actually work. :) | 20:24 |
| kfox1111 | have you played with pyghmi at all? | 20:24 |
| sbezverk | kfox1111: nope, I checked with ironic and they do have something in their devstack for testing | 20:25 |
| kfox1111 | I've got a snippet of code from when I played with it last... | 20:25 |
| kfox1111 | let me try and dig it up... | 20:25 |
| sbezverk | kfox1111: I would need to use a tested driver, which is the one they use in devstack | 20:27 |
| kfox1111 | http://paste.openstack.org/show/f8dKn1oGbGiENm8RqTr2/ | 20:27 |
| sbezverk | kfox1111: it is just it is not feasable to keep everything in ps until full testing is completed | 20:28 |
| kfox1111 | I'm pretty sure ironci wrote pyghmi | 20:28 |
| sbezverk | it is already huge, and I need to add inspector approx the same size | 20:28 |
| kfox1111 | do you plan on working on some tests then right after merge? | 20:29 |
| kfox1111 | I guess I'm ok splitting it up, but really don't want to have much non tested code for very long. | 20:29 |
| kfox1111 | still waiting for a fix for that 10 minute sleep too. :/ | 20:30 |
| jascott1 | kfox1111 do we have to quote yaml values if they have a . ? | 20:30 |
| kfox1111 | jascott1: no idea. | 20:30 |
| jascott1 | ok I think so but will check | 20:30 |
| sbezverk | kfox1111: from ironic" virtualbmc to translate IPMI requests into libvirt commands" | 20:30 |
| kfox1111 | sbezverk: yeah. thats the one. | 20:31 |
| sbezverk | kfox1111: of course | 20:31 |
| kfox1111 | so if we launch the vm like in that paste, and then hook it to virshbmc, ironic should be able to treat it just like a real physical machine. | 20:31 |
| sbezverk | I mean I will work, I do not plan to drop time I already invested in this thing | 20:31 |
| kfox1111 | ok. cool. just checking. | 20:31 |
| sbezverk | kfox1111: yeah in theory it should work like that but before I try it with my hands I cannot comment ;) | 20:32 |
| *** lamt has quit IRC | 20:33 | |
| kfox1111 | totally get that. :) | 20:33 |
| kfox1111 | I've gotten a version of quintipleo working on one of my clouds once, and I'm pretty sure it was using a nova backed ipmi emulator using pyghmi | 20:34 |
| kfox1111 | so should be very similar. | 20:34 |
| sbezverk | kfox1111: could be. btw I found out that nova 2.0.2 does not work with ironic well | 20:35 |
| *** salv-orl_ has quit IRC | 20:35 | |
| sbezverk | I had to use 3.0.3 for nova and ironic to get it working.. | 20:35 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla master: Updated from global requirements https://review.openstack.org/431961 | 20:35 |
| kfox1111 | hand built? | 20:36 |
| kfox1111 | 3.0.2 didn't work either? | 20:36 |
| jascott1 | kfox1111 seems the answer is no, it doesnt require quotes around strings with . in them, couldnt suss it out of the spec but see a lot of examples without them | 20:37 |
| sbezverk | kfox1111: nope | 20:37 |
| sbezverk | ironic-conductor packaging issue | 20:38 |
| sbezverk | but 3.0.3 worked like a charm | 20:38 |
| *** crushil has joined #openstack-kolla | 20:38 | |
| kfox1111 | weird. ok. | 20:38 |
| kfox1111 | we really need to get to the point where we can test the containers that get released. :/ | 20:39 |
| sbezverk | kfox1111: right, but testing all services kolla provides images for is close to impossible | 20:42 |
| kfox1111 | nah. it just hasn't been prioritized. :/ | 20:43 |
| sbezverk | meeting time, bbiab | 20:44 |
| kfox1111 | k | 20:44 |
| *** mgoddard has joined #openstack-kolla | 20:45 | |
| *** ssurana has joined #openstack-kolla | 20:46 | |
| *** jemcevoy is now known as jemcevoy_ | 20:52 | |
| kfox1111 | jascott1: ping | 20:59 |
| jascott1 | yo | 20:59 |
| kfox1111 | easier to talk this way. :) | 20:59 |
| jascott1 | you dont like getting 1000 emails? :) | 20:59 |
| jascott1 | true but we have the record but I think we have enough now | 20:59 |
| kfox1111 | hehe. no, I'm ok with that. | 20:59 |
| jascott1 | from prev work we throw it to chartutil.LoadRequirements | 21:00 |
| kfox1111 | so, at the point your trying to resolve the values, what do you have access to? | 21:00 |
| jascott1 | you have parent and all its requirements, charts | 21:01 |
| jascott1 | and their values | 21:01 |
| kfox1111 | I was figuring you woudl parse dependencies.yaml at that point then look at the charts/<thingy>/values.yaml directly. | 21:01 |
| jascott1 | probably | 21:01 |
| kfox1111 | but I guess it depends on what step you are doing it at. | 21:02 |
| jascott1 | but I dont want to change order of requirements | 21:02 |
| kfox1111 | do you do it at helm install time, | 21:02 |
| kfox1111 | helm dep up time, | 21:02 |
| kfox1111 | or both? | 21:02 |
| kfox1111 | cause I can think of a few different ways to do it. | 21:02 |
| jascott1 | it would just be install and upgrade | 21:02 |
| jascott1 | so my problem is i dont want to mess with the ordering of deps as loaded by the system | 21:02 |
| jascott1 | i can look into it further to see hows its loading | 21:03 |
| kfox1111 | if it was done partially at dep up time, | 21:03 |
| jascott1 | i say we resolve by that if it makes sense | 21:03 |
| kfox1111 | dep up could generate a childsvalues.yaml in the parent based on the logic we defined in dependencies.yaml | 21:03 |
| jascott1 | hmmm | 21:03 |
| kfox1111 | and then install would only have to merge that preflattened out values in directly. | 21:03 |
| kfox1111 | giving it lowest priority | 21:04 |
| jascott1 | i dont want any more files personally | 21:04 |
| kfox1111 | hmm... k. | 21:04 |
| *** g3ek has quit IRC | 21:04 | |
| jascott1 | yeah I dont see any problems pulling this off other than conflicts within the same requirements | 21:04 |
| jascott1 | which maybe is anti case | 21:04 |
| kfox1111 | then the cleanest way I can think of is to parse dependencies.yaml at install time and look up values files directly, if they aren't already available without merging. | 21:04 |
| *** haplo37 has quit IRC | 21:06 | |
| jascott1 | my only problem atm is that we cant rely on order requirements are loaded to determine which internal conlicting path wins | 21:06 |
| jascott1 | is that an issue for us? | 21:06 |
| jascott1 | do we expect to have these conflicts within a requirements? | 21:06 |
| kfox1111 | I could see it happening... | 21:07 |
| jascott1 | oh ok | 21:07 |
| kfox1111 | is just reparsing the files directly a problem? | 21:07 |
| jascott1 | no but what is the order user should expect? | 21:07 |
| kfox1111 | it would have a slight performance hit, but not a huge one | 21:07 |
| jascott1 | nah its not bad, all on the client at run time (install and uprade) | 21:08 |
| kfox1111 | oh. are you saying individual deps, not the entries in the list? | 21:08 |
| jascott1 | right | 21:08 |
| kfox1111 | ah. | 21:08 |
| jascott1 | idk let me look into it | 21:08 |
| jascott1 | maybe it will be clear | 21:08 |
| kfox1111 | well, | 21:08 |
| *** jemcevoy_ is now known as jemcevoy | 21:08 | |
| jascott1 | maybe it can be last one wins within requirements | 21:08 |
| kfox1111 | lets see... | 21:08 |
| kfox1111 | dependencies.yaml is not a requirement for stuff to be in /charts | 21:09 |
| kfox1111 | so there could be some that are pulled in randomly by a user just dropping one in there. | 21:09 |
| kfox1111 | but.... we're putting include-values in dependencies.yaml, so all the ones that may get values included are listed. | 21:09 |
| *** haplo37_ has quit IRC | 21:09 | |
| jascott1 | right | 21:09 |
| kfox1111 | depenedencies.yaml has a list of dicts... | 21:10 |
| kfox1111 | so there is an order there. | 21:10 |
| kfox1111 | maybe we do last one has higest priority there too? | 21:10 |
| jascott1 | right but I think helm doesnt enforce that order but I need to check | 21:10 |
| kfox1111 | the order probably doesn't matter to helm at all. only to this use case. | 21:10 |
| jascott1 | it may not be a big deal | 21:10 |
| kfox1111 | yeah. | 21:11 |
| jascott1 | well I have enough to get started for sure. I have taken that issue and shrunk it into a new proposal thats concise | 21:11 |
| kfox1111 | ok. cool. | 21:12 |
| kfox1111 | man this raid aray is grumpy. :/ | 21:12 |
| jascott1 | sorry to hear, raid | 21:13 |
| jascott1 | https://github.com/kubernetes/helm/issues/1995 | 21:13 |
| kfox1111 | I try to avoid it and just use ceph these days.... not always avoidable though. :/ | 21:13 |
| *** rwallner has quit IRC | 21:14 | |
| kfox1111 | jascott1: looks good. but misses the defined maping of a short to a long? version? its kind of implied in the example though. | 21:15 |
| jascott1 | oh the hidden parent: "." ? | 21:17 |
| *** crushil has quit IRC | 21:18 | |
| *** mgoddard has quit IRC | 21:18 | |
| *** haplo37 has joined #openstack-kolla | 21:18 | |
| jascott1 | since we can mix I dont thinks its required for the conversation but can include in the final docs | 21:18 |
| kfox1111 | yeah. the child: "parent.%s" %string | 21:18 |
| kfox1111 | parent: . thing | 21:18 |
| *** g3ek has joined #openstack-kolla | 21:18 | |
| kfox1111 | actually it looks like the whole parent thing is left out. | 21:18 |
| *** haplo37_ has joined #openstack-kolla | 21:19 | |
| jascott1 | i thought we agreed it was optional | 21:19 |
| kfox1111 | so the example is slightly wrong to me. | 21:19 |
| kfox1111 | it is optional. but for the short string, required. | 21:19 |
| jascott1 | oh missed that point in the discussion | 21:19 |
| kfox1111 | I think the not parent case shuld be supported, but discuraged by default. | 21:19 |
| kfox1111 | as introspection tools have no way of knowing its use case. :/ | 21:20 |
| jascott1 | im opposite there because of the 'must modify the chart' problem | 21:20 |
| kfox1111 | jascott1: see kolla-kubernetes/tools/build_example_yaml.py | 21:20 |
| kfox1111 | we want to provide that functionality, but don't want to have a bunch of random junk show up in it that the user shouldn't ever see. | 21:21 |
| kfox1111 | no, problem is, | 21:21 |
| kfox1111 | they will have to modify the chart anyway to use the feature. | 21:21 |
| *** crushil has joined #openstack-kolla | 21:21 | |
| jascott1 | why? | 21:21 |
| kfox1111 | cause the values that exist today won't match up to the values used by the templates I think, cause that never worked before. | 21:21 |
| jascott1 | as written currently I can reach into any chart without modifying it | 21:22 |
| kfox1111 | but the code comes from macros. and they cant use values today. | 21:22 |
| kfox1111 | so higly likely the no values exist for thme, or maybe even conflicting values. | 21:22 |
| kfox1111 | while possible to mix the two usescases of, direcly launchable chart, and chart that provides a library, I think the vast majority of charts will be either the first or the second type, not both? | 21:24 |
| jascott1 | library chart will not need to be invoked directly, that what you are saying? | 21:24 |
| kfox1111 | and even if both, the latter will quite possibly have values that are ideal for the directly laucnhed chart, as that part works, | 21:24 |
| kfox1111 | but values for the templates wont be correct out of the box/ | 21:25 |
| kfox1111 | kind of, yeah. saying its unlikely to ever be launched standalone, as it won't do anything at all. | 21:25 |
| kfox1111 | but even if it did, the values that are there today, are different then the values you may want to pull in to the parent for the macro's it provides. | 21:26 |
| kfox1111 | cause the macro's and the child objects are two totally different things. | 21:26 |
| kfox1111 | am I making sence? | 21:26 |
| jascott1 | im sure you are but Im not familar enough with how we are doing things I guess | 21:27 |
| jascott1 | let me study the build_yaml and get back, sorry gotta bounce for a couple hours to take care of things before ptg | 21:27 |
| kfox1111 | k. | 21:27 |
| kfox1111 | yeah. maybe talking about it more concretely might help. | 21:27 |
| kfox1111 | the abstract is making it hard to talk about I think. | 21:27 |
| *** Pavo has joined #openstack-kolla | 21:28 | |
| *** l4yerffeJ_ has joined #openstack-kolla | 21:34 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 21:34 | |
| *** Jeffrey4l has quit IRC | 21:34 | |
| *** l4yerffeJ has quit IRC | 21:35 | |
| *** tonanhngo has quit IRC | 21:38 | |
| *** bjolo_ has quit IRC | 21:51 | |
| *** Pavo has quit IRC | 21:53 | |
| *** rwallner has joined #openstack-kolla | 22:02 | |
| *** rwallner has quit IRC | 22:03 | |
| *** Pavo has joined #openstack-kolla | 22:06 | |
| *** jtriley has quit IRC | 22:06 | |
| *** tonanhngo has joined #openstack-kolla | 22:16 | |
| *** dave-mccowan has quit IRC | 22:17 | |
| *** fooliouno has quit IRC | 22:18 | |
| *** crushil has quit IRC | 22:19 | |
| *** rwallner has joined #openstack-kolla | 22:19 | |
| *** rwallner_ has joined #openstack-kolla | 22:21 | |
| *** mannamne has joined #openstack-kolla | 22:23 | |
| *** rwallner has quit IRC | 22:23 | |
| *** jheroux has quit IRC | 22:24 | |
| *** rwallner_ has quit IRC | 22:25 | |
| *** salv-orlando has joined #openstack-kolla | 22:26 | |
| *** Serlex has joined #openstack-kolla | 22:30 | |
| *** sdake has joined #openstack-kolla | 22:30 | |
| sdake | i'd say good morning but it is afternoon :) | 22:35 |
| sdake | high folks | 22:35 |
| sdake | rather hi folks :) | 22:35 |
| bpattewar_ | sdake hi | 22:36 |
| bpattewar_ | sdake I got my dev environment up . Thanks to docker images by bmace. I am looking at https://review.openstack.org/#/c/414825/ these micro services. | 22:37 |
| *** rhallisey has quit IRC | 22:40 | |
| sdake | bpattewar_ nice! | 22:41 |
| sdake | bpattewar_ you can always use the docker images fromt he registry | 22:41 |
| sdake | bpattewar_ for that patch, you can always help out be rebasing it | 22:42 |
| bpattewar_ | sdake will git checkout will work to use it, can you tell me git commands to do it. I pretty new to git. | 22:47 |
| *** matrohon has quit IRC | 22:48 | |
| *** eaguilar has joined #openstack-kolla | 22:52 | |
| *** penfold has quit IRC | 22:54 | |
| sdake | yup | 22:56 |
| sdake | moment | 22:56 |
| sdake | click "download" feature and "cherrypick" option | 22:57 |
| sdake | i have a meeting in 3 minutes for 8-15 minutes | 22:57 |
| sdake | will you be around then? | 22:57 |
| sdake | when you cpy the cherrypick to the clipboard, the cherrypick will fial | 22:57 |
| sdake | this is a normal part of rebasing | 22:57 |
| sdake | its alot easier forme to show you one on one over webex or some other screen sharing tool how to do the job | 22:58 |
| sdake | rather then you experiment with doing it | 22:58 |
| sdake | however up to you :) | 22:58 |
| *** Serlex has quit IRC | 23:00 | |
| bpattewar_ | sdake it will be awesome if you can show me on webex. I will be working 3 more hours. Lets do it after your meeting. | 23:03 |
| sbezverk | sdake: please do not forget to review ironic PSs | 23:06 |
| *** mannamne has quit IRC | 23:25 | |
| kfox1111 | array's finally in good shape... that was a pain. :/ | 23:25 |
| sdake | sbezverk on my list | 23:29 |
| sdake | bpattewar_ lets get rolling | 23:29 |
| bpattewar_ | sdake lets do it. | 23:29 |
| *** ccrouch has quit IRC | 23:30 | |
| *** ccrouch has joined #openstack-kolla | 23:30 | |
| bpattewar_ | sdake how I can join webex | 23:32 |
| *** MasterOfBugs has quit IRC | 23:32 | |
| sdake | bpattewar_ click join | 23:33 |
| sdake | i'm not sur ewhat you see on your screen webex just went through an update | 23:33 |
| sdake | if you want to use a different tool we can do that | 23:33 |
| sdake | i guess i need to see it from the other end after the update | 23:33 |
| *** Pavo has quit IRC | 23:34 | |
| *** Pavo has joined #openstack-kolla | 23:34 | |
| sdake | bpattewar_ can't year you very well | 23:40 |
| sdake | bpattewar_ see PMs | 23:40 |
| *** sbezverk_ has joined #openstack-kolla | 23:45 | |
| *** sbezverk has quit IRC | 23:48 | |
| *** Pavo has quit IRC | 23:49 | |
| *** lamt has joined #openstack-kolla | 23:50 | |
| *** saneax-_-|AFK is now known as saneax | 23:51 | |
| *** lamt has quit IRC | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!