| babbageclunk | right - that confirms that it's code running inside the container that can't get to the lxd port. | 00:00 |
|---|---|---|
| babbageclunk | dmitri-s: ^ | 00:00 |
| dmitri-s | So my LXD config is incorrect? | 00:01 |
| babbageclunk | I think so | 00:01 |
| dmitri-s | Just tried curl from inside a container and it times out (curl: (7) Failed to connect to 10.208.133.1 port 8443: Connection timed out) | 00:02 |
| kelvinliu | wallyworld: got this PR to add pod addresses to binding addresses for CaaS, could u take a look? thanks! https://github.com/juju/juju/pull/10741 | 00:02 |
| wallyworld | i can, just need to get some stuff done first | 00:03 |
| kelvinliu | np, not urgent at all | 00:04 |
| babbageclunk | dmitri-s: sorry, just trying to work out what config needs to change... | 00:05 |
| dmitri-s | Here is my current lxd config (lxd init --dump) https://github.com/dmitri-s/juju-learn/blob/master/lxd_config | 00:06 |
| babbageclunk | oh thanks | 00:07 |
| babbageclunk | comparing that to mine... | 00:07 |
| babbageclunk | dmitri-s: the only difference I can see is that I have `core.proxy_ignore_hosts: 127.0.0.1,::1,localhost` in the config section. | 00:11 |
| babbageclunk | what does `route -n` show inside the container? | 00:11 |
| dmitri-s | root@pleasant-bonefish:~# route -n | 00:12 |
| dmitri-s | Kernel IP routing table | 00:12 |
| dmitri-s | Destination Gateway Genmask Flags Metric Ref Use Iface | 00:12 |
| dmitri-s | 0.0.0.0 10.208.133.1 0.0.0.0 UG 100 0 0 eth0 | 00:12 |
| dmitri-s | 10.208.133.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 | 00:12 |
| dmitri-s | 10.208.133.1 0.0.0.0 255.255.255.255 UH 100 0 0 eth0 | 00:12 |
| babbageclunk | and obviously you can get to the internet from the container since we can see things installing in the juju log... | 00:13 |
| babbageclunk | (the route output looks the same as mine) | 00:14 |
| dmitri-s | interesting | 00:16 |
| babbageclunk | Sorry, this is about the point I'd ask in #lxcontainers | 00:16 |
| dmitri-s | thanks. I will do that. BTY, was I supposed to post this in discourse.jujucharms.com rather than ask here? | 00:17 |
| dmitri-s | just so I know next time. | 00:17 |
| dmitri-s | Just from the server message "https://jujucharms.com, general chat on https://discourse.jujucharms.com" | 00:17 |
| dmitri-s | seems like it is where it point me | 00:18 |
| babbageclunk | No, this is a good place too, especially for back-and-forth. Although the advantage of discourse is that it's persistent - if people aren't in the same timezone then they won't see irc chat but they can still answer posts. | 00:19 |
| dmitri-s | cool. thanks for your help. | 00:20 |
| babbageclunk | no worries - sorry I couldn't sort it out fully | 00:21 |
| babbageclunk | (or at all) | 00:21 |
| dmitri-s | haha... movement in any direction is progress as I see it. I figured out that "x.x.133.1" is a gateway, so it is a progress. | 00:25 |
| dmitri-s | With your help, nothing less. :) | 00:26 |
| babbageclunk | :) | 00:26 |
| wallyworld | kelvinliu: free now in standup | 00:56 |
| kelvinliu | yep | 00:56 |
| wallyworld | kelvinliu: lgtm with some suggested cleanup to make things more logical now that the behaviour has changed | 02:46 |
| kelvinliu | wallyworld: yeo thx | 02:47 |
| kelvinliu | yep* | 02:47 |
| anastasiamac | wallyworld: for post-lunch digestion, PTAL https://github.com/juju/juju/pull/10746 - 'show-cloud' ask-or-tell | 03:11 |
| wallyworld | ok | 03:12 |
| wallyworld | just getting coffee | 03:12 |
| anastasiamac | oh good! o the timing is right for light reading :D | 03:12 |
| wallyworld | anastasiamac: i asked a question, see if i am being dumb | 03:33 |
| wallyworld | babbageclunk: i pushed a change to address the race, see what you think? | 03:33 |
| anastasiamac | wallyworld: kind of ;) i'll reply on pr | 03:34 |
| wallyworld | anastasiamac: s/kind of/most definitely. lgtm | 03:42 |
| anastasiamac | wallyworld: :) it was existing code anyway that I have just prettied up :) | 03:43 |
| wallyworld | yeah, i didn't see that from the review diff, just mainly saw the comment etc | 03:43 |
| anastasiamac | wallyworld: thank you for the review tho \o/ really appreciate another set of eyes | 03:44 |
| wallyworld | more eyes always good | 03:44 |
| babbageclunk | wallyworld: approved | 03:52 |
| wallyworld | \o/ tyvm | 03:52 |
| wallyworld | babbageclunk: call? | 04:02 |
| babbageclunk | oops | 04:02 |
| thumper | babbageclunk: another https://github.com/juju/juju/pull/10747 | 04:47 |
| thumper | or wallyworld ^^ | 04:47 |
| wallyworld | oooh instancepoller, my favourite | 04:47 |
| * thumper EODs | 04:50 | |
| anastasiamac | wallyworld: PTAL https://github.com/juju/juju/pull/10748 - 'show-credential' part | 05:42 |
| wallyworld | ok | 05:42 |
| stickupkid | nammn_de, good catch last night, yes if you hold it differently it breaks, I fixed it so both ways works https://github.com/juju/juju/pull/10749 | 08:56 |
| nammn_de | stickupkid: cool, approved :D | 09:11 |
| nammn_de | rick_h I just scrolled through the systemd code and found out that we do a verify an a basic level. So I don't think we need to add additional checks. So IMO the pr is ready to be reviewed https://github.com/juju/juju/pull/10696 | 09:26 |
| manadart | wallyworld: Is there are a resolution for all apps in a CAAS model getting "already exists" errors for generating secrets? | 09:29 |
| nammn_de | manadart: regarding the bug you mentioned before with the charm version. I did add some additional test, which does not include the bug yet, but still include some other cases. https://github.com/juju/juju/pull/10744 , when we "fix the bug" we could add the other cases as well. | 09:29 |
| nammn_de | Additionally I updated the bug description with my debugging output https://bugs.launchpad.net/juju/+bug/1848149 . I planned to take this on later | 09:29 |
| mup | Bug #1848149: Charm Deployed or Upgrade from Dir Ignores "version" <juju:Triaged> <https://launchpad.net/bugs/1848149> | 09:29 |
| manadart | nammn_de: Ack. | 09:30 |
| manadart | wallyworld: for example: "creating secrets for container: tensorboard: secret "tensorboard-tensorboard-secret" already exists" | 09:30 |
| nammn_de | achilleasa stickupkid : im trying the qa steps from hml https://github.com/juju/juju/pull/10718 there are some things I don't know. Got some time and want to have fun walking me through a little with a little qa? Doesn't have to be now | 09:37 |
| achilleasa | jam: shouldn't this (https://github.com/juju/juju/blob/develop/worker/instancepoller/updater.go#L140-L145) check use the instance status (which can be running) instead of the machine status (AFAICT would be created)? Or are statuses working differently for manual machines? | 09:49 |
| nammn_de | manadart: if you have some free time https://github.com/juju/juju/pull/10691 pr regarding spaces. Sadly I do have some weird network issues and my internal upgrade bootstraps got stuck (qa steps) | 10:22 |
| manadart | nammn_de: Reviewed. Upgrade step needs another look. | 10:47 |
| nammn_de | manadart: Just read it, good to know for me! Thanks | 11:48 |
| achilleasa | nammn_de: for 10691 you probably also need to patch this: https://github.com/juju/juju/blob/develop/cmd/juju/application/upgradecharm.go#L757 | 12:39 |
| nammn_de | manadart: added your suggestions, qa steps worked for me! | 12:40 |
| nammn_de | achilleasa, manadart :" ohh, nvm me than I ll take a look | 12:41 |
| nammn_de | achilleasa: are we fine to use the default if none is set, which should be set in any case with my patch? | 12:44 |
| stickupkid | anyone know where the simple streams url is in the source code? I know we have a model config value, but it's empty, so we must fill it in with something if it's empty. | 12:46 |
| achilleasa | nammn_de: I 'd say leave it for now to be safe in case someone is testing the new client against a controller from a sha before your PR landed | 12:47 |
| nammn_de | achilleasa: not patching this function at all or you mean adding a check if it exists, use it? | 12:48 |
| achilleasa | nammn_de: if the model (it's passed an an arg to that function) defines it use that or use the network.DefaultSpaceName as a fallback | 12:49 |
| nammn_de | achilleasa: thanks. Similiar to that? https://github.com/juju/juju/blob/ef08716950173d3a388d675b5c49c1210137c050/cmd/juju/application/upgradecharm.go#L761-L760 | 12:51 |
| jam | achilleasa: IIRC, MachineStatus is the status of the machine agent, vs the status of the machine itself | 12:53 |
| achilleasa | jam: that's my understanding as well. So can this be ever running? | 12:54 |
| achilleasa | (in running state) | 12:54 |
| jam | achilleasa: I'm really not sure what it is trying to check there. It may just be trying to say "set it to running if it isn't running" | 12:54 |
| achilleasa | yes, but I would assume that the check would be more like "if my *instance* status is not running and I am a manual machine then set me as running" | 12:55 |
| achilleasa | stickupkid: can you please check if this works on develop for you? go test -check.v -check.f TestAgentSetsToolsVersion (in cmd/jujud/agent) | 15:09 |
| stickupkid | achilleasa, why is that? | 15:10 |
| achilleasa | stickupkid: https://pastebin.canonical.com/p/pf5wXMFmtD/ | 15:11 |
| stickupkid | wtf is that | 15:11 |
| achilleasa | I will retry on current HEAD | 15:12 |
| stickupkid | achilleasa, https://pastebin.canonical.com/p/J86Vg3z53B/ | 15:12 |
| achilleasa | this fails on HEAD for me: "FAIL: machine_test.go:607: MachineLegacyLeasesSuite.TestAgentSetsToolsVersionManageModel" | 15:13 |
| achilleasa | could it be deps? brb | 15:14 |
| achilleasa | nope... still failing | 15:15 |
| nammn_de | manadart: time for a quick review? I added your and achilleasa suggestions https://github.com/juju/juju/pull/10691/files | 15:16 |
| achilleasa | stickupkid: wonder if that test is actually pulling stuff from somewhere... | 15:16 |
| stickupkid | achilleasa, we should have a look at what the test does | 15:16 |
| achilleasa | stickupkid: L598 | 15:18 |
| stickupkid | achilleasa, ho? | 15:20 |
| achilleasa | omw | 15:21 |
| manadart | nammn_de: Needs another change. I have to run. | 15:25 |
| achilleasa | hml: are you on eoan? | 15:30 |
| hml | achilleasa: not yet | 15:30 |
| stickupkid | hml, go test -check.v -check.f TestAgentSetsToolsVersion | 15:30 |
| stickupkid | hml, ho in daily | 15:31 |
| hml | stickupkid: omw | 15:31 |
| achilleasa | rick_h: are you on eoan? | 15:34 |
| achilleasa | also nammn_de are you on eoan? | 15:36 |
| nammn_de | achilleasa: yesss | 15:36 |
| nammn_de | manadart: thanks, damnit | 15:36 |
| achilleasa | can you join daily please? | 15:37 |
| achilleasa | nammn_de: ^ | 15:38 |
| achilleasa | nammn_de: actually can you please run go test -check.v -check.f TestAgentSetsToolsVersion (in cmd/jujud/agent) and let me know if it fails for you? | 15:40 |
| rick_h | achilleasa: yes, I'm on eoan | 15:40 |
| achilleasa | rick_h: can you try the above command ^^ | 15:41 |
| rick_h | achilleasa: the test command? | 15:41 |
| achilleasa | yeap. I think it fails on eoan | 15:41 |
| achilleasa | (only) | 15:42 |
| rick_h | achilleasa: I'll see. I need to setup a dev environment on here. I've not done that yet. | 15:46 |
| rick_h | achilleasa: heads up in case you didn't see it https://github.com/juju/juju/pull/10747#event-2720070282 | 15:46 |
| rick_h | achilleasa: might be worth pulling and making sure that there's no issues in overlap/etc | 15:46 |
| achilleasa | rick_h: thanks, already incorporated that in my PR | 15:47 |
| rick_h | achilleasa: awesome | 15:47 |
| nammn_de | achilleasa: sorry wasnt there, still need me? | 15:58 |
| nammn_de | achilleasa: ah i see not needed anymore | 15:59 |
| nammn_de | manadart: reverted the change | 16:05 |
| manadart | achilleasa: Can you approve that^ one? | 16:06 |
| achilleasa | manadart: one sec | 16:07 |
| manadart | Got hands full with kids here. Not at PC. | 16:07 |
| achilleasa | nammn_de: approved | 16:14 |
| stickupkid | achilleasa, any news? | 16:30 |
| manadart | Need a review for https://github.com/juju/juju/pull/10751. | 16:33 |
| achilleasa | manadart: one for tomorrow: https://github.com/juju/juju/pull/10750 | 16:56 |
| nammn_de | thanks manadart achilleasa | 16:56 |
| mbeierl | Having some trouble getting the apt layer to install a ppa. It seems to be ignoring whatever I put in. Maybe a yaml level problem? Are there any working examples that show a full layer.yaml with adding a ppa? | 17:15 |
| mbeierl | stub: where is the best place to go for more info on the apt-layer? I am obviously missing something. | 17:34 |
| thumper | morning | 19:51 |
| rick_h | morning thumper | 19:58 |
| rick_h | thumper: if you get time can you peek back at nam's branch for the log permissions please? | 21:21 |
| thumper | rick_h: will do | 21:21 |
| rick_h | <3 | 21:22 |
| stub | mbeierl: https://github.com/stub42/layer-apt has the README rendered | 21:57 |
| babbageclunk | can I get a review for this merge? https://github.com/juju/juju/pull/10752 | 22:21 |
| hpidcock | babbageclunk: on it | 22:25 |
| babbageclunk | merci m'sieur | 22:28 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!