=== kadams54 is now known as kadams54-away | ||
=== hazmat is now known as hazaway | ||
=== hazinhell is now known as hazmat | ||
frankban | uiteam I need two reviews for https://codereview.appspot.com/195690043 (quickstart/python) | 14:26 |
---|---|---|
frankban | thanks! | 14:26 |
hatch | on it frankban | 14:36 |
frankban | hatch: thanks | 14:37 |
frankban | hatch: have you seen my email re juju connect stuff? | 14:37 |
hatch | not yet | 14:38 |
hatch | i will | 14:38 |
=== kadams54 is now known as kadams54-away | ||
=== kadams54-away is now known as kadams54 | ||
hatch | frankban: review and qa done | 15:34 |
hatch | looks good | 15:34 |
hatch | frankban: I am interested to know why you wrapped those two methods? | 15:35 |
hatch | frankban: so why did you wrap those functions in a functools.wrap decorator? can you not return a function in python normally? | 16:26 |
frankban | hatch: functools.wrap preserves the original name and documentation of the decorated function | 16:28 |
hatch | frankban: ahh | 16:29 |
frankban | hatch: FYI those values are stored in func.__name__ and func.__doc__ | 16:31 |
hatch | interesting that those aren't included by default | 16:39 |
hatch | any idea why the gui charm needs firefox in its sysdeps? | 17:12 |
* hatch cry's "I just want to run the linter" | 17:16 | |
mbruzek | hatch ping | 17:16 |
hatch | sup | 17:16 |
hatch | uiteam, anyone know why I wouldn't have the correct version of pip for the guicharm linter? | 17:17 |
frankban | hatch: firefox is required by functional tests using selenium | 17:17 |
frankban | hatch: not sure if I understand your second question | 17:17 |
hatch | # Ensure the correct version of pip has been installed | 17:18 |
hatch | tests/.venv/bin/pip --version | grep "pip 1.[4-9]" || exit 1 | 17:18 |
hatch | make: *** [setup] Error 1 | 17:18 |
hatch | ^ frankban | 17:18 |
mbruzek | hatch: I am trying to figure out the charm store uris | 17:18 |
frankban | hatch: what pip version do you have? | 17:18 |
hatch | 1.5.4 | 17:18 |
hatch | pip 1.5.4 from /usr/lib/python2.7/dist-packages (python 2.7) | 17:19 |
mbruzek | hatch: Kapil has a launchpad url lp:~hazmat/charms/bundles/kubernetes/bundle | 17:19 |
* hazmat ears are ringing | 17:19 | |
frankban | hatch: tests/.venv/bin/pip --version | 17:19 |
mbruzek | hatch: and quickstart maps to that juju quickstart bundle:~hazmat/kubernetes/kubernetes | 17:19 |
mbruzek | hello haz | 17:19 |
mbruzek | hello hazmat | 17:20 |
hazmat | greetings | 17:20 |
hatch | frankban: ummm 6.0.7? | 17:20 |
frankban | hatch: really? | 17:20 |
hatch | mbruzek: if you want to deploy a bundle using quickstart or the depployer then you'll have to use the old yaml format | 17:20 |
mbruzek | hatch: trying to figure out how quickstart maps bundle:~hazmat/kubernetes/kubernetes to lp:~hazmat/charms/bundles/kuberentes/bundles | 17:20 |
hatch | pip 6.0.7 from ......../develop-trunk/tests/.venv/local/lib/python2.7/site-packages (python 2.7) | 17:21 |
hatch | mbruzek: it guesses | 17:21 |
hazmat | mbruzek: on lp distro is charms, series is bundle is the nutshell | 17:21 |
mbruzek | hatch trying to figure out the mapping so I can write readmes for other bundles | 17:21 |
frankban | hatch: I am rebuilding the devenv for the char, and checking again | 17:21 |
hatch | mbruzek: all of that is changing soon - makyo_ is working on updating the deployer to support the new syntax | 17:22 |
hazmat | hatch: thats not url syntax is it? i thought it was mostly just syntatic incompatible format changes | 17:23 |
hazmat | in the bundle itself | 17:23 |
mbruzek | hatch So I am writing a readme for how quickstart works now. I just want to know what format will work with our bundle here: p:~kubernetes/charms/trusty/kubernetes-bundle/trunk | 17:23 |
hatch | no the url changed as well | 17:23 |
hatch | mbruzek: if you want to use that you'll have to point to the old yaml file directly | 17:23 |
hatch | the new bundle yaml format is not supported by the deployer | 17:24 |
mbruzek | hatch: I don't want to point to that url, I like the bundle: uri, I just need to know how it works. | 17:24 |
frankban | mbruzek: quickstart still uses the old charmworld for bundles, and the supported syntax is described in quickstart --help | 17:24 |
mbruzek | frankban: thank you. Will read up on that | 17:24 |
hatch | for some reason the deployer wasn't updated before the bundle syntax was | 17:24 |
frankban | mbruzek: it will be updated soon to support the new bundle URLs | 17:24 |
hatch | frankban: why do we want to use such an old version of pip? 5 major versions older? | 17:25 |
mbruzek | hatch: frankban: I see the help text now, it seems that quickstart supports 5 formats. Which one would you recommend? I see things are changing, what would be the format that would be recommended that works now and in to the future? | 17:27 |
hatch | I think those two are mutually exclusive :) | 17:27 |
mbruzek | doh | 17:27 |
hatch | juju quickstart bundle:mediawiki/single | 17:28 |
hatch | we want this format to work | 17:28 |
hatch | but currently it doesn't | 17:28 |
frankban | hatch: I think in the future we want "juju quickstart bundle/mediawiki-single" | 17:29 |
frankban | mbruzek: ^^^ | 17:29 |
hatch | oh jeeze that's a new format to me | 17:29 |
mbruzek | hatch: frankban: Our bundle currently lives here: lp:~kubernetes/charms/trusty/kubernetes-bundle/trunk | 17:29 |
hatch | I think I'll just write code - give up on what format is what lol | 17:29 |
frankban | hatch: in the new charm store, the "bundle" string is not a schema, is a series | 17:29 |
mbruzek | It is not yet promulgated. Is that in a wrong spot? Should we change that? | 17:29 |
frankban | hatch: it seems that pip quickly switched from version 1.5.6 to 6.0 :-| | 17:31 |
hatch | lol | 17:31 |
frankban | hatch: so yes that check should be changed, feel free to do that as part of your branch | 17:31 |
hatch | frankban: what was it there for? | 17:31 |
hatch | There must be a reason why we checked at all | 17:32 |
frankban | hatch: I did not add that code, so not sure, perhaps some backward incompatible changes with previous pip versions | 17:32 |
hatch | so think I can just remove the check entirely? | 17:32 |
frankban | hatch: I'd suggest to keep the check, just make it more permissive | 17:34 |
hatch | so just greater than 1.4? | 17:34 |
frankban | hatch: +1 | 17:34 |
hatch | frankban: do you know of a way to structure the grep for that? I am having a heck of a time nesting the conditionals in the makefile | 17:51 |
hatch | ifndef apparently doesn't work for me | 17:52 |
hatch | oh makefiles | 17:52 |
frankban | hatch: grep -E "pip (1.[4-9]|6)" ? | 17:54 |
hatch | negative, tried that :) | 17:55 |
hatch | here is what I am attempting to do in the makefile (and failing) https://gist.github.com/hatched/83544f1c201ee1a71a6e | 17:56 |
hatch | for some reason ifndef is not found? | 17:56 |
frankban | hatch: what I suggested works here | 18:00 |
frankban | % echo "pip 6.3" | grep -E "pip (1.[4-9]|6)" (trunk) ubuntu | 18:00 |
frankban | pip 6.3 | 18:00 |
hatch | yeah it works from the echo but not in practice | 18:01 |
hatch | tests/.venv/bin/pip --version | grep -E '1\.[4-9]\.|6\.[0-9]\.[0-9]' || exit 1 | 18:15 |
hatch | bleh lol | 18:15 |
mbruzek | hatch: I tried to deploy using quickstart using a raw html bundle, but none of the charm store charms showed up. I ran this with --debug and the pastebin is here. http://paste.ubuntu.com/10076915/ Can you help me | 18:17 |
mbruzek | ? | 18:17 |
hatch | looking | 18:18 |
hatch | what was the command you used? | 18:18 |
hatch | btw thanks for writing these docs :) | 18:19 |
hatch | mbruzek: still there? | 18:25 |
mbruzek | yes | 18:25 |
hatch | what was the command you used? | 18:25 |
mbruzek | juju quickstart https://raw.githubusercontent.com/whitmo/bundle-kubernetes/master/bundles.yaml --debug 2>&1 | tee quickstart_problem.txt | 18:25 |
mbruzek | The bundle is legit, it works with deployer and references Charm store uris | 18:26 |
mbruzek | I am trying to figure out why none of the charms are loaded | 18:26 |
hatch | hmm yeah you're right it looks good | 18:29 |
frankban | mbruzek: quickstart states that the bundle deployment request has been accepted | 18:29 |
mbruzek | fankban I can show you the juju status that tells me otherwise | 18:29 |
mbruzek | or the gui if you don't believe status | 18:29 |
* mbruzek *grins* | 18:29 | |
hatch | mbruzek: have you tried deploying those services directly? | 18:29 |
frankban | mbruzek: what's the content of https://15.125.118.3/gui-server-info | 18:29 |
frankban | ? | 18:29 |
mbruzek | {"uptime": 3435, "deployer": [{"Status": "started", "Queue": 0, "DeploymentId": 0, "Time": 1423157614}, {"Status": "scheduled", "Queue": 1, "DeploymentId": 1, "Time": 1423159220}, {"Status": "scheduled", "Queue": 2, "DeploymentId": 2, "Time": 1423159241}, {"Status": "scheduled", "Queue": 3, "DeploymentId": 3, "Time": 1423159325}], "apiversion": "go", "sandbox": false, "version": "0.4.2", "debug": false, "apiurl": "wss://10.0.0.166:17070"} | 18:30 |
frankban | mbruzek: so you have pending bundle deployments in the queue, they are executed in order | 18:31 |
mbruzek | frankban: I guess I tried this bundle several times now. | 18:31 |
mbruzek | I can destroy and start over again, but I don't think I would get a different result. | 18:31 |
mbruzek | hatch: I have deployed this bundle through juju-deployer when have the bundle on my local machine. | 18:32 |
frankban | mbruzek: let's check the guiserver logs: they are stored in the gui unit in /var/log/upstart/guiserver.log | 18:32 |
mbruzek | on machine 1? | 18:33 |
frankban | mbruzek: wherever the gui unit is placed: juju ssh juju-gui/0 | 18:33 |
mbruzek | frankban: OK found it, is that safe to pastebin? It does not show my keys or other stuff? | 18:34 |
frankban | mbruzek: I think so, but let's use the canonical pastebin | 18:35 |
mbruzek | frankban: working on it | 18:40 |
frankban | mbruzek: ty | 18:40 |
mbruzek | frankban: had to paste it by hand, no upload feature on pastebin.canonical.com | 18:45 |
mbruzek | https://pastebin.canonical.com/125051/ | 18:45 |
frankban | mbruzek: uhm... I don't see any errors in the logs. it seems that the deployer is still working on your first bundle as usual, except it's not deploying anything | 18:49 |
frankban | mbruzek: you can try, from that machine, "sudo service guiserver restart", then "juju set juju-gui builtin-server-logging=debug" and then try to deploy the bundle again (once) and see if the logs show something | 18:50 |
mbruzek | I see a python exception in queues.py is that nothing? | 18:50 |
frankban | mbruzek: not sure if that's related to the bundle deployment | 18:53 |
frankban | mbruzek: but it's well over my EOD now and I have to go, so could you please try what I described and file a bug against the GUI charm with the info attached? | 18:54 |
mbruzek | will do, thanks for your assistance frankban. | 18:55 |
mbruzek | Have a good day | 18:55 |
mbruzek | night | 18:55 |
frankban | night and ty | 18:55 |
=== kadams54 is now known as kadams54-away | ||
=== kadams54-away is now known as kadams54 | ||
=== kadams54 is now known as kadams54-away | ||
=== kadams54-away is now known as kadams54 | ||
=== kadams54 is now known as kadams54-away | ||
=== kadams54-away is now known as kadams54 | ||
hatch | uiteam do we know which branch I should push my guicharm changes to? lp:charms/trusty/juju-gui or lp:~juju-gui/charms/trusty/juju-gui/trunk | 20:31 |
hatch | and by push I mean propose | 20:31 |
=== kadams54 is now known as kadams54-away | ||
hatch | uiteam lf a review for a smallish diff https://code.launchpad.net/~hatch/charms/trusty/juju-gui/add-uuid-configjs/+merge/248826 | 20:46 |
=== kadams54-away is now known as kadams54 | ||
=== kadams54 is now known as kadams54-away | ||
rick_h_ | hatch: juju-gui/charms/trusty/juju-gui/trunk | 22:22 |
hatch | oh hai! | 22:23 |
hatch | yeah I have it up for review | 22:23 |
hatch | :) | 22:23 |
rick_h_ | ah cool | 22:24 |
hatch | tornado is bonkers | 22:25 |
hatch | just so you know :P | 22:25 |
rick_h_ | :) | 22:25 |
rick_h_ | async python ftw | 22:25 |
rick_h_ | hatch: just think of it like nodejs | 22:25 |
hatch | ehh I'm still thinking an async language might be better :) | 22:25 |
hatch | makyo_ and I are going to have to get schooled by frankban tomorrow :) | 22:26 |
rick_h_ | heh, docs aren't bad | 22:27 |
hatch | nah it's just trying to find the correct spot to put things in the guiserver is the issue | 22:27 |
rick_h_ | gotcha | 22:27 |
hatch | having a yeild in the middle of a function then never returning to it really threw me for a loop | 22:28 |
huwshimi | Morning | 22:32 |
hatch | morning huwshimi | 22:32 |
rick_h_ | morn | 22:32 |
hatch | I fixed some of the rendering performance issues in leankit by modifying their stylesheets :) | 22:32 |
hatch | there are cascading shadows and transparencies so it's no wonder you get screen tearing on slower machines | 22:33 |
hatch | I found this cool site that shows benchmark comparisons for various languages. http://benchmarksgame.alioth.debian.org/u64q/compare.php?lang=dart&lang2=go | 22:43 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!