| *** patersonc has joined #cip | 06:16 | |
| *** rajm has joined #cip | 07:08 | |
| rajm | donbrown yes I'm happy with the top level location for initramfs.cpio.gz | 07:21 |
|---|---|---|
| * rajm changes wiki to match that | 07:24 | |
| *** toscalix has joined #cip | 07:57 | |
| wagi | Hi, I am trying to setup b@d, following the instruction on the wiki ('building from scratch'). I am stuck at '5. luchn the Virtual Machine using Vagrant'. I get a "Call to virDomainCreateWithFlags failed: Unable to get index for interface eth0: No such device" even though I changed the bridge to my usb ethernet device. git grep eth0 didn't show anything up so I don't know where this magic knob is. Any pointers? | 08:24 |
| rajm | hmmm looking at output when I 'up' vagrant | 08:27 |
| wagi | It tells me that eth0 is not existing, which is correct | 08:27 |
| rajm | so you're using a usb ethernet connection? | 08:29 |
| wagi | https://paste.fedoraproject.org/paste/2HwHPiGBX7JKfXFd6OHmNw | 08:29 |
| * rajm looks | 08:30 | |
| wagi | Maybe I got this wrong. I though the VM should be attached to an network with the device on it | 08:30 |
| wagi | s/the device/the device to test/ | 08:30 |
| rajm | the device being the board to be tested - ah | 08:31 |
| wagi | yes, that is what I wanted to say :) | 08:31 |
| rajm | though at the point you are at its not trying to connect there yet | 08:32 |
| rajm | the connect to the board happens in the health check | 08:32 |
| wagi | not there yet | 08:33 |
| wagi | Just did the git clone && vagrant up step | 08:33 |
| wagi | and 'vagrant up' is obviously not so happy | 08:33 |
| rajm | I'd remove the bridge you added bring up the vagrant vm and then address the connection within the rule in the health check - mybbb.dat contains an example | 08:34 |
| rajm | Oh an I've run the daily health check ok on the BBB | 08:35 |
| wagi | Stupid question: where can I find mybbb.dat? | 08:37 |
| rajm | when the vagrant is up you vagrant ssh and it's in /home/vagrant | 08:37 |
| rajm | if vagrant up doesn't work for there there may be a problem there :-( | 08:38 |
| wagi | right, will give it a try after I solved vagrant up | 08:38 |
| rajm | yes try with an unmodified Vagrantfile and see if there's still problems | 08:39 |
| wagi | I started with a prestine version | 08:40 |
| *** kristerman has joined #cip | 08:40 | |
| rajm | so even without the bridge you have errors can you pastebin the output | 08:42 |
| toscalix | configured the mailing list to share results. donbrown all yours :-) | 08:54 |
| wagi | https://paste.fedoraproject.org/paste/A-ENWiXEF8ul11jXJiSIlw/ | 08:56 |
| * wagi was fetching tea | 08:56 | |
| wagi | let me try again | 08:57 |
| wagi | it is missing the errno | 08:57 |
| wagi | stderr | 08:57 |
| rajm | thanks wagi | 08:57 |
| wagi | https://paste.fedoraproject.org/paste/wtOUUZMpFjccty-sbyFtQw/ | 08:59 |
| wagi | that is with no modification at all | 08:59 |
| rajm | ok thanks let me investigate | 09:00 |
| wagi | It might something with my host system | 09:02 |
| rajm | I assume your host is running fedora? Is this link https://github.com/vagrant-libvirt/vagrant-libvirt/issues/349 any help? | 09:03 |
| wagi | Yes, I am using F25. The bug description seems to match | 09:07 |
| wagi | Let me try to apply the 'fixes' | 09:08 |
| rajm | let us know how it goes! | 09:08 |
| wagi | Will do. Thanks anyway | 09:10 |
| *** bananadev has joined #cip | 09:14 | |
| *** patersonc has quit IRC | 09:19 | |
| *** patersonc has joined #cip | 09:19 | |
| wagi | rajm: the trick is to start the virt-manager and configure the brige setup there. | 09:43 |
| rajm | \o/ we'll put something in the known issues for this | 09:44 |
| wagi | I am sure you can also do this via virtsh but I didn't bother to find it out | 09:45 |
| wagi | hmm, thinking about it, it should not run libvirt. It should be virtualbox | 09:49 |
| *** patersonc has quit IRC | 09:58 | |
| *** patersonc has joined #cip | 09:59 | |
| toscalix | joining | 10:19 |
| *** patersonc has quit IRC | 10:32 | |
| wagi | rajm: it seems to boil down to the fact, that libvirt has a default network which needs to be destroyed before vagrant up runs | 10:32 |
| *** patersonc has joined #cip | 10:33 | |
| wagi | rajm: summary: i had some configuration leftovers from older runs. All needs to go and any default network in libvirt has to remoced as well. After that via the graphic virt-manager the bridged network interface needs to set to your local configuration | 10:38 |
| wagi | virsh net-destroy default | 10:38 |
| wagi | virsh net-undefine default | 10:38 |
| rajm | ok thanks for that feedback! | 10:39 |
| wagi | now I see some install steps from apt-get after running vagrant up | 10:39 |
| wagi | IIRC initiall this project used virtualbox as backend. Is that correct? | 10:40 |
| rajm | I think it's always been vagrant sitting on virtualbox but donbrown may know better than me | 10:44 |
| wagi | I prefer using kvm. virtualbox is not properly intergrated with their out of tree kernel module | 10:45 |
| wagi | and that one is not even working righ now for me | 10:45 |
| rajm | yes we're looking to provide a release supporting kvm but for the initial release we wanted both linux and windows support so initially using vagrant | 10:46 |
| toscalix | wagi: we all linux users want kvm The reason behind virtualbox was to involved the windows users from hitachi, toshiba and siemens | 10:53 |
| toscalix | kvm is obviously in our list | 10:53 |
| toscalix | I doubt we will have time to work on it before ELCE though | 10:54 |
| toscalix | https://gitlab.com/cip-project/cip-testing/testing/issues/31 | 10:54 |
| wagi | Sure, I know the background. <rant> But I find it hard to understand why on earth someone wants to work on Linux and uses Windows for this? Okay in a coperate env it is your base system but why not using a VM with Linux inside? </rant> | 10:56 |
| toscalix | added a comment with you opinion wagi Your colleague Jan had the same opinoin | 10:57 |
| patersonc | wagi: Agreed | 10:57 |
| toscalix | wagi: at Codethinks we provide workshops in which we talk hours about dogfooding :-) | 10:57 |
| toscalix | let me check my slides from ALS. I think I included | 10:58 |
| wagi | toscalix: I know. Sorry about beeing a bit vocal on this | 10:58 |
| toscalix | no no... it is good to have some reinforcement once in a while | 10:59 |
| wagi | I just recenently had to build an SDK for Windows based on Yocto... it is not supported by upstream and it is likely to eat you cat | 10:59 |
| toscalix | yep, I talked about dogfooding and about using windows to create linus systems | 10:59 |
| toscalix | jajajajajajaja | 11:00 |
| wagi | for a 'profissional' product | 11:00 |
| wagi | :) | 11:00 |
| toscalix | if we could talk one day about what we do in customers..... | 11:00 |
| wagi | over a beer I suggest | 11:02 |
| *** bananadev has quit IRC | 11:02 | |
| wagi | btw, I try to setup a b@d for -rt testing | 11:09 |
| wagi | rostedt uses ktest for this | 11:09 |
| wagi | but you need to setup all on your own | 11:09 |
| wagi | tftp, dhcp, ... | 11:09 |
| wagi | FATAL -> Failed to fork. | 11:16 |
| wagi | could it be that I have to increase some limits for the install step? | 11:17 |
| wagi | this happends during the mongodb install step | 11:17 |
| rajm | sorry wagi, I've been lunching... | 11:35 |
| wagi | no worries | 11:36 |
| wagi | I set the memory size to 2gb and added another cpu | 11:36 |
| wagi | it looks promising | 11:37 |
| rajm | running out of memory? how much does your host have ? have you overridden the amount - ah! | 11:37 |
| rajm | yes that's the size in the vagrantfile | 11:37 |
| wagi | I am uhsing libvirt and not virtualbox | 11:37 |
| rajm | yes we initially used less but had problems at various places | 11:38 |
| wagi | those scripts install quite a lot of packages. even mutt gets installed | 11:39 |
| wagi | well, if the final result works I do not bother :) | 11:39 |
| wagi | rajm: moar memory helped | 11:43 |
| rajm | yes can't remember what dependency drags that in probably via exim - the VM will soon be emailing the results | 11:47 |
| rajm | \o/ - to moar memory | 11:47 |
| wagi | it definietly helped to add moooooarrrr ram but it the install steps just stops. the last output is something on systemd-netword. | 12:07 |
| wagi | https://paste.fedoraproject.org/paste/Z31-DdR1NqeipOaSpKUz0g | 12:08 |
| toscalix | we have a known issues page. rajm should we add something related with the ram? I believe we discussed this in the past but cannot remember | 12:11 |
| toscalix | https://wiki.linuxfoundation.org/civilinfrastructureplatform/ciptestingknownissues | 12:11 |
| wagi | I think install_lava.sh was succesfully processed (that file missed an 'echo success') | 12:12 |
| rajm | the Vagrantfile specifies 2GB of memory but if other hypervisors are being used maybe we need a mention | 12:12 |
| wagi | I'll send a patch as soon I got it working | 12:13 |
| rajm | yes wagi there doesn't appear to be an error will look at the script to check | 12:13 |
| rajm | toscalix there's the adding extra resources in the final section of that page, maybe the prerequisites ought to add that 2GB (at least) is recommended | 12:14 |
| rajm | thanks wagi | 12:14 |
| toscalix | rajm: no mention to the prerequisites in the step 1 page for vagrant | 12:15 |
| toscalix | just for the box | 12:15 |
| wagi | the mybb.dat has been created as well, that means somwhere in the last few lines it gets stuck | 12:15 |
| toscalix | oh wait | 12:16 |
| toscalix | it is mentioned | 12:16 |
| toscalix | but not highlighted like in the box | 12:16 |
| toscalix | changing it | 12:16 |
| rajm | ok | 12:16 |
| rajm | am just checking the prerequisites to mention memory | 12:17 |
| wagi | bah, all looks good. the last line in configure_lava.sh has been executed ('sed') | 12:18 |
| wagi | I am going to add another 'echo END:' | 12:19 |
| rajm | maybe just an exit code not getting back somehow - thanks! | 12:19 |
| wagi | I just don't know what to expect when it succeeds. Maybe having a message telling 'all fine' would help | 12:20 |
| toscalix | creating the ticket with the suggestion so we go over it in the coming days | 12:21 |
| toscalix | https://gitlab.com/cip-project/cip-testing/testing/issues/107 | 12:25 |
| rajm | yes there needs to be more example output in https://wiki.linuxfoundation.org/civilinfrastructureplatform/ciptestingboardatdeskdingledevdeployment#b-d-deployment-method-2-building-vm-from-scratch-using-vagrant-14 | 12:25 |
| rajm | can I add that suggestion to #107 or do we need another ticket? | 12:26 |
| wagi | toscalix: thanks :) | 12:26 |
| toscalix | rajm: I would add it to the description of the same ticket | 12:27 |
| toscalix | including the acceptance criteria | 12:27 |
| toscalix | lava 2017.6 released | 12:29 |
| rajm | will do | 12:29 |
| rajm | thanks for those MR's wagi - I think that #32 should also change the scripts so there's a echo "START: install_lava.sh" sim for the other for consistency with the other scripts? | 13:12 |
| patersonc | Hello bwh, are you around? | 13:57 |
| toscalix | patersonc: I do not think so. On Monday | 14:02 |
| patersonc | Okay. I thought that might be the case! | 14:02 |
| *** rajm has quit IRC | 16:00 | |
| *** kristerman has quit IRC | 16:13 | |
| *** kristerman has joined #cip | 16:35 | |
| *** kristerman has quit IRC | 16:39 | |
| *** toscalix has quit IRC | 17:02 | |
| *** patersonc has quit IRC | 17:13 | |
| *** Jackneilll has joined #cip | 23:17 | |
| *** Jackneill has quit IRC | 23:26 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!