| *** lennyb has quit IRC | 00:03 | |
| *** lennyb has joined #openstack-dib | 00:04 | |
| *** pmannidi has quit IRC | 00:41 | |
| *** pmannidi has joined #openstack-dib | 00:46 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: exec_sudo : no decoding https://review.openstack.org/463427 | 01:03 |
|---|---|---|
| ianw | mmedvede: http://dal05.objectstorage.softlayer.net/v1/AUTH_3d8e6ecb-f597-448c-8ec2-164e9f710dd6/pkvmci/diskimage-builder/27/463427/1/check/dib-dsvm-functests-devstack-xenial/3994af3/ | 01:07 |
| ianw | this seems to be failing pretty consistently not being able to download? | 01:07 |
| *** johnsom has quit IRC | 01:13 | |
| *** johnsom has joined #openstack-dib | 01:13 | |
| mmedvede | ianw: not sure how to debug this, not able to reproduce the failure. Wanted to check on our elk for history, but realized the job isn't indexed :( | 01:13 |
| ianw | mmedvede: network issue on the host? | 01:14 |
| ianw | for once, i'm not sure it's dib's fault ;) | 01:14 |
| mmedvede | hehe, that is definitely not dib fault. I suspect it could be networking. I need to check which host it was running on | 01:15 |
| mmedvede | I did a quick wget test on one of other dsvm slaves and it worked every time. So makes it harder to figure out | 01:16 |
| ianw | ahh, i thought i was seeing it pretty regularly | 01:19 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: exec_sudo : no decoding https://review.openstack.org/463427 | 01:19 |
| mmedvede | ianw: should I disable debian-minimal for now? I do not remember if it provides the testing for the broken functionality | 01:20 |
| ianw | mmedvede: yeah ... we just turned it on for ubuntu-minimal i think? | 01:20 |
| ianw | it did just hit for ^^ too | 01:22 |
| ianw | http://dal05.objectstorage.softlayer.net/v1/AUTH_3d8e6ecb-f597-448c-8ec2-164e9f710dd6/pkvmci/diskimage-builder/27/463427/2/check/dib-dsvm-functests-devstack-xenial/d5a6747/ | 01:22 |
| mmedvede | all right. I am able to reproduce the problem by running the debootstrap. Just downloading the package manually works. I can work with that | 01:28 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: exec_sudo : no decoding https://review.openstack.org/463427 | 01:34 |
| *** NobodyCam has quit IRC | 01:58 | |
| *** NobodyCam has joined #openstack-dib | 01:58 | |
| *** khomkrit1499 has joined #openstack-dib | 03:18 | |
| mmedvede | ianw: so I found very strange thing is happening - an older version of debian Pakcages.gz gets downloaded somehow, and of course signature check fails. This is the reason for the powerkvm CI test failure | 03:19 |
| mmedvede | Looks like CDN for http://ftp.us.debian.org/debian/dists/stable/main/binary-ppc64el/Packages.gz did not update (if they even use a CDN) | 03:22 |
| mmedvede | so it appears that that us mirror has an out of date repo, mirroring has failed. I wonder who should be contacted to fix this | 03:25 |
| mmedvede | not related to dib :) | 03:26 |
| *** aparnav has joined #openstack-dib | 03:29 | |
| mmedvede | ianw: I switched to a different deb mirror, seems to work now | 03:44 |
| *** khomkrit1499 has quit IRC | 04:01 | |
| ianw | mmedvede: thanks! yeah, not sure about the mirror issue | 04:04 |
| *** andreas-f has joined #openstack-dib | 04:40 | |
| *** jamielennox is now known as jamielennox|away | 05:12 | |
| *** jamielennox|away is now known as jamielennox | 05:18 | |
| *** jamielennox is now known as jamielennox|away | 05:20 | |
| *** jamielennox|away is now known as jamielennox | 05:23 | |
| *** chhavi has joined #openstack-dib | 05:26 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 05:34 |
| yolanda | hi ianw , are you testing the python 3 issues with sudo? i tried yesterday refactoring the patch, but is still failing | 05:48 |
| ianw | yolanda: yep, see 463427 | 05:48 |
| yolanda | indeed, i was getting errors on different places, depending on the approach i was taking with exec_sudo | 05:49 |
| ianw | yolanda: so, hopefully that is clear. for sanity that function should just take str() input (that way, we can join it for the print, whatever) | 05:50 |
| yolanda | i added the "b" prefix yesterday, but i don't think that it needs to land | 05:50 |
| ianw | if we get bytes from somewhere ... like the output of a subprocess call, etc, then we should make that a str() when we receive it | 05:50 |
| yolanda | let me revert my tries from yesterday so your patch is cleaner | 05:51 |
| ianw | yolanda: if you want, just squash that change into the other one | 05:51 |
| ianw | it passes ocatvia test | 05:51 |
| ianw | also, https://review.openstack.org/463431 is out to give us python3 on the xenial tests | 05:52 |
| yolanda | ianw, i think that yours can even apply before, right? | 05:52 |
| ianw | if it works, i think we run centos/trusty as python2, and xenial as python3. that way we're covered for both | 05:52 |
| ianw | yolanda: maybe? can try rebasing it before if you want | 05:52 |
| yolanda | yep | 05:53 |
| yolanda | let me revert my changes | 05:53 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 05:56 |
| yolanda | agh, messed the cherry-pick | 05:57 |
| ianw | hmm ... are you try to swap them? | 05:58 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 05:59 |
| ianw | i'd "git review -d 463427" then "rebase -i master" and swap them? | 05:59 |
| yolanda | oh i normally just cherry-pick the patch i want and commit again... maybe your idea works better | 06:00 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Use check_output https://review.openstack.org/463471 | 06:00 |
| yolanda | anyway, it's there now | 06:00 |
| ianw | heh, yeah, i'm rebase crazy | 06:00 |
| ianw | "-x" is also awesome to run tox in between | 06:00 |
| yolanda | need to step out for 30 minutes but can continue later | 06:02 |
| yolanda | rebase of yours give some merge conflicts | 06:02 |
| *** isaacb has joined #openstack-dib | 06:22 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 06:25 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 06:27 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 07:40 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 07:46 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 07:51 |
| *** TheJulia has quit IRC | 07:57 | |
| *** TheJulia has joined #openstack-dib | 07:58 | |
| *** jamielennox is now known as jamielennox|away | 08:19 | |
| *** hashar has joined #openstack-dib | 08:22 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 08:34 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 08:38 |
| *** jamielennox|away is now known as jamielennox | 08:51 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 10:05 |
| *** pmannidi has quit IRC | 10:22 | |
| ianw | yolanda: hmmm ... that looping ... ood | 10:40 |
| ianw | odd | 10:40 |
| yolanda | i was hitting the same on my test system actually | 10:42 |
| yolanda | i refactor a bit, so it does not happen | 10:42 |
| ianw | b" " is not right ... that would stop at a space | 10:48 |
| yolanda | but independently on that, this should not cause that loop | 10:52 |
| yolanda | or you mean that the change from b' ' to b" " caused it? | 10:55 |
| ianw | yolanda: it should be b'' | 10:56 |
| ianw | no space | 10:56 |
| ianw | did you inadvertently turn it into b" " when it started failing? | 10:56 |
| yolanda | i guess so | 10:56 |
| ianw | that would explain it :) | 10:56 |
| yolanda | ok let's retry with the right char | 10:57 |
| ianw | i was really wondering how that pattern could fail! | 10:57 |
| ianw | it's saying "stop when you hit EOF" which happens when the program exists | 10:57 |
| ianw | exits | 10:57 |
| yolanda | so yes, i messed that adding an extra space :( | 10:58 |
| yolanda | ianw, don't you think we shall add output to logger.error if exit code from the proc.wait() is not zero? right now if your patch, you could only see the error if enablign debug | 11:00 |
| yolanda | and the output should be relevant if we hit an error in the exec_sudo | 11:00 |
| ianw | yolanda: yes, true, i did remove that check there | 11:04 |
| ianw | yolanda: probably what we should do is copy the way check_output() works | 11:04 |
| ianw | return 0, or raise CommandFailed or whatever it is | 11:04 |
| ianw | CalledProcessError | 11:05 |
| ianw | in fact, if we called it sudo_check_output and made it behave the same ... well that would be pretty neat because anyone who understands check_output() understands it's just that with a sudo infront | 11:08 |
| ianw | though i do want to trace in there ... it's important, so that's not the same | 11:09 |
| ianw | yeah, scratch that | 11:09 |
| yolanda | so you prefer that we replace the popen with the check_output? | 11:10 |
| ianw | yolanda: no, sorry i was thinking we could copy the interface. but it doesn't quite work | 11:11 |
| ianw | i think that the function should check the return value, and raise CalledProcessError if !0 | 11:11 |
| ianw | yolanda: seems pretty nice? -> http://paste.openstack.org/show/609187/ | 11:16 |
| yolanda | and also add the output to logger.error properly | 11:16 |
| ianw | yolanda: well, raising the exception will either be caught, or print a good traceback | 11:17 |
| yolanda | raise subprocess.CalledProcessError(proc.returncode, | 11:18 |
| yolanda | ' '.join(sudo_cmd)) | 11:18 |
| yolanda | that will give errors again on python 3? | 11:18 |
| yolanda | the join like that? | 11:18 |
| yolanda | ah no, because it should have failed earlier | 11:18 |
| ianw | no, because sudo_cmd is a list of "str"s | 11:18 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 11:19 |
| yolanda | yep, that is checked before | 11:20 |
| ianw | yolanda: sorry, tiny tiny nit on the function doc but otherwise LGTM! | 11:23 |
| *** hashar has quit IRC | 11:28 | |
| *** hwoarang has quit IRC | 11:39 | |
| *** hwoarang has joined #openstack-dib | 11:40 | |
| yolanda | i'm also testing Andrea's patch with xfs , and fails with several errors, seems that xfs is not supporting much of the options | 11:50 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: exec_sudo: command should be "str" https://review.openstack.org/463427 | 11:52 |
| *** hwoarang has quit IRC | 12:04 | |
| *** chhavi has quit IRC | 12:05 | |
| *** hashar has joined #openstack-dib | 12:08 | |
| *** hwoarang has joined #openstack-dib | 12:08 | |
| *** hwoarang has quit IRC | 12:28 | |
| *** hwoarang has joined #openstack-dib | 12:32 | |
| *** hwoarang has quit IRC | 12:34 | |
| *** hwoarang has joined #openstack-dib | 12:34 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 12:35 |
| yolanda | ianw, andreas-f , so i posted comments on https://review.openstack.org/#/c/444586/36 . Hitting conflicts between the filesystem definition on the block device, vs the exported FS_TYPE | 12:36 |
| *** hwoarang_ has joined #openstack-dib | 12:49 | |
| *** hwoarang has quit IRC | 12:51 | |
| *** aparnav has quit IRC | 13:25 | |
| *** chhavi has joined #openstack-dib | 13:40 | |
| *** hwoarang_ is now known as hwoarang | 14:11 | |
| *** dmarlin has joined #openstack-dib | 15:06 | |
| *** isaacb has quit IRC | 15:13 | |
| *** abarbaneld has joined #openstack-dib | 15:51 | |
| abarbaneld | I have been reading through the Disk Builder Documentation, I am trying to use the element install-static, I created the directory static and my items are not found in the image, can anyone help? | 15:52 |
| abarbaneld | Command is: disk-image-create -a amd64 -o test -t qcow2 -p mariadb-server,java-1.7.0-openjdk centos7 vm install-static | 15:53 |
| *** dabarban has joined #openstack-dib | 15:59 | |
| *** abarbaneld has quit IRC | 16:00 | |
| *** dabarban is now known as abarbaneld | 16:00 | |
| greghaynes | abarbaneld: that element is only useful if you create the directory static inside of a custom element | 16:01 |
| greghaynes | abarbaneld: so you'd need to create a 'my-element' and inside there a file element-deps with 'install-static' and then a static dir with your files, then add that element to your build | 16:02 |
| abarbaneld | So adding that element to my build would be a matter of passing the element name? | 16:03 |
| greghaynes | yes and export ELEMENTS_PATH=path_to_dir_with_your_element | 16:04 |
| abarbaneld | Great! thanks for the assist greghaynes | 16:05 |
| greghaynes | np | 16:05 |
| greghaynes | that being said, it'd be neat IMO to have the install-static element support setting DIB_INSTALL_STATIC_DIR=/some/dir and it'd automatically grab file from there | 16:06 |
| greghaynes | should be easy to write too if youre interested | 16:06 |
| abarbaneld | Possibly, let me see if I can get this working | 16:07 |
| *** isaacb has joined #openstack-dib | 16:11 | |
| *** hashar has quit IRC | 16:44 | |
| *** isaacb has quit IRC | 17:00 | |
| *** chhavi has quit IRC | 18:18 | |
| *** andreas-f has quit IRC | 19:49 | |
| *** abarbaneld has quit IRC | 20:40 | |
| *** pmannidi has joined #openstack-dib | 21:35 | |
| ianw | yolanda: cool on the python3 jobs, did we test them? | 22:04 |
| ianw | to answer -> http://logs.openstack.org/86/444586/37/check/gate-dib-dsvm-functests-python3-ubuntu-xenial-nv/661a379/console.html | 22:08 |
| ianw | yay! it fails :) | 22:08 |
| *** dmarlin has quit IRC | 22:46 | |
| *** jamielennox is now known as jamielennox|away | 23:07 | |
| *** jamielennox|away is now known as jamielennox | 23:11 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 23:34 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Let exec_sudo raise exceptions https://review.openstack.org/463627 | 23:34 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!