/srv/irclogs.ubuntu.com/2016/06/03/#juju-dev.txt

perrito666I am saying yay a lot on meetings its a sign of either madness or that I am becoming a teenager schoolgirl00:01
wallyworldredir: chat noe if you want00:02
redirk00:02
wallyworldstandup ho00:02
redirsame bat place?00:02
redirk00:02
* thumper wades through more migration tests00:07
menn0wallyworld, axw: With mongo3.2 I can't use the mongo shell to connect to Juju's DB any more00:16
menn0wallyworld, axw: I suspect it's that the SCRAM-SHA-1 auth mechanism is used by default now00:16
menn0but the 2.6 client doesn't support it00:17
menn0do you guys know a way around it00:17
wallyworldmenn0: right, i have organised for mongo client to be packaged with our juju mongo db00:17
wallyworldthis is not done yet :-(00:17
menn0wallyworld: ok, so at least it's a known issue00:17
wallyworldballoons: ^^^^^ how far away is our mongo packaging fix?00:17
menn0wallyworld: will it be in the juju-mongo3.2-tools package?00:18
wallyworldyes00:18
wallyworldso it is available on any controller00:18
perrito666menn0: install mongodb-org ppa and then use mongodb-org-clients00:18
menn0perrito666: awesome - thanks00:19
perrito666menn0: I might have misplaced some of those dashes00:19
menn0perrito666: i'll figure it out?00:19
perrito666but the names of the packages/ppas are those00:19
=== natefinch-afk is now known as natefinch
davechen1ythumper: this time I think the race build did complete toally00:36
davechen1yand I think I can fix the build hang in the termination worker00:36
anastasiamacdavechen1y: while fixinf races, did u come around any in uniter and hooks?00:40
anastasiamacdavechen1y: i have bugs against 1.25 and m wondering if i need to start from scratch or backport..00:40
davechen1yanastasiamac: hard to say, do you ahve a bug reference00:40
davechen1yat this point, i'd think backporting from 2.0 is near impossible00:40
davechen1ythe codebases have diverged so much00:41
anastasiamacdavechen1y: the one i was going to start with is bug 148671200:42
mupBug #1486712: Race on uniter-hook-execution, prevents to resolve unit. <canonical-bootstack> <hooks> <race-condition> <sts> <sts-needs-review> <juju-core:Triaged> <juju-core 1.25:Triaged> <https://launchpad.net/bugs/1486712>00:42
davechen1yanastasiamac: this is a logical race, not a data race00:44
davechen1yi guess00:44
davechen1y(as there is no output from the race detector000:44
davechen1yin fact, i've no idea why the OP said race in the title00:45
mupBug #1585015 changed: worker/terminationworker: test timeout during CI <tech-debt> <unit-tests> <juju-core:Won't Fix> <https://launchpad.net/bugs/1585015>00:46
mupBug #1585846 changed: worker/terminationworker: test timeout after 20 minutes <timeout> <unit-tests> <juju-core:Won't Fix> <https://launchpad.net/bugs/1585846>00:46
mupBug #1588572 opened: cmd supercommand.go:448 cannot read tools metadata in tools directory: open /var/lib/juju/tools/2.0-beta6-xenial-amd64/downloaded-tools.txt: no such file or directory <juju-core:New> <https://launchpad.net/bugs/1588572>00:52
davechen1yanastasiamac: this doesn't look like a data race, just a straight out bug from not handling an error path and the uniter getting out of sync with what the state thinks it is00:52
davechen1yie, dying when the action will not be retried00:53
axwwallyworld: is it safe to land my instancecfg change yet, or is the beta8 cutover still happening?00:53
davechen1yor maybe it's retrying in a loop00:53
davechen1ybut not making any progress00:53
anastasiamacdavechen1y: k00:55
wallyworldaxw: safe to land, they are taking a previous revision01:02
axwgoodo01:02
wallyworldaxw: haven't looked yet, kust finished talking to people, but here's the acl wip https://github.com/juju/juju/pull/5526/files01:02
axwwallyworld: yeah I just added a comment01:03
davechen1ymenn0: thumper https://github.com/juju/juju/pull/552801:22
davechen1yfix for terminationworker hang01:22
* thumper looks01:22
menn0davechen1y: you already have 2 ship its and thumper hasn't even finished yet :)01:29
davechen1yta01:30
davechen1ylanding it01:30
mupBug #1585424 changed: all: data races in tests are surpressed <race-condition> <tech-debt> <juju-core:Fix Released> <https://launchpad.net/bugs/1585424>01:34
mupBug #1588574 opened: Session already closed in state/presence <intermittent-failure> <juju-core:Triaged> <https://launchpad.net/bugs/1588574>01:34
mupBug #1588575 opened: allwatcher_internal_test has intermittent failure <intermittent-failure> <juju-core:Triaged> <https://launchpad.net/bugs/1588575>01:34
menn0wallyworld: would you be open to the charm upload HTTP endpoint taking an optional "schema" argument01:40
menn0wallyworld: for migrations I need to preserve the original schema01:40
wallyworldmenn0: sure, i think01:40
menn0when charmstore charms are migrated they end up with the wrong charm URL01:40
menn0(local: instead of cs:)01:41
menn0wallyworld: the only concern I can think of is some sort of nefarious behaviour were someone installs a dodgy charm as if it were a charmstore charm01:42
menn0s/installed/uploads01:42
menn0gah01:42
menn0s/installs/uploads/01:43
wallyworldmenn0: can't think at the moment, otp with tim01:43
menn0wallyworld: ok np01:43
axwsinzui: what happened to the DNS records for vapour.ws?01:53
davechen1yhey prople who think we're using mongodb 3.201:55
davechen1ywe're not using mongodb 3.201:55
davechen1yor even mongodb 2.601:55
davechen1ywe're using  Get:20 http://archive.ubuntu.com/ubuntu/ trusty/universe juju-mongodb amd64 2.4.9-0ubuntu3 [6,936 kB]01:55
davechen1ydoes this come as a surprise to anyone ?01:55
axwdavechen1y: I think we only use 3.2 on xenial? not sure tho, need to double check01:55
axwanyone got a cached IP for juju-ci.vapour.ws?01:57
davechen1yaxw: this is the landing bot01:58
axwdavechen1y: ah right, in the tests.01:58
davechen1y% host juju-ci.vapour.ws01:59
davechen1yjuju-ci.vapour.ws has address 54.86.116.23401:59
axwdavechen1y: thanks01:59
davechen1y.ws does appear to be busted at the moment01:59
cmarswallyworld, on it02:00
thumpermenn0: a very dull PR https://github.com/juju/juju/pull/552902:00
wallyworldty02:00
menn0thumper: looking02:01
* thumper heads out to get the dog some exercise02:03
thumperand to leave the house for the day02:03
thumperbbl02:03
menn0thumper:  ship it02:05
natefinchwallyworld: ok, time to actually figure out this bug, now that I have some data that's not just "something went wrong"02:05
wallyworldnatefinch: awesome, otp will catch up soon02:06
natefinchchrome/linux/somebody is getting mad that I'm copying & pasting 3.8megs to pastebin02:07
* natefinch remembers that pastebinit exists02:08
davechen1ynuts, my dns entry for .ws aged out02:38
davechen1ythumper: https://bugs.launchpad.net/juju-core/+bug/158857402:40
mupBug #1588574: Session already closed in state/presence <intermittent-failure> <juju-core:Triaged> <https://launchpad.net/bugs/1588574>02:40
davechen1yi cannot reproduce this locally02:40
davechen1ystress tested for two hours02:40
davechen1ycan you ?02:40
davechen1yironically it just hit me in CI http://juju-ci.vapour.ws:8080/job/github-merge-juju/7972/console02:41
axwwallyworld: am I reviewing https://github.com/juju/juju/pull/5519 or https://github.com/juju/juju/pull/5525?02:53
wallyworldaxw: the one merging into the feature branch02:53
wallyworldcmars: i don't see how the currebt PR fixes the deploy_test compile error?02:54
cmarswallyworld, did i forget to push? checking..02:54
cmarswallyworld, it fixes it because I fucked up the field name in changeset number 2. see http://reviews.vapour.ws/r/4971/diff/2-4/02:56
cmarsso from 0-4 you'll see no net difference but the landing bot failed on #2 i think02:56
wallyworldcmars: ah right, ok. so we retain the field name but serialise as "service-name"02:57
cmarswallyworld, yeah, wireformat needs more time to change02:57
wallyworldcmars: np, lgtm02:58
wallyworldty02:58
cmarswallyworld, thanks!02:58
thumperdavechen1y: I don't think I have hit it locally for quite some time03:00
thumpernot sure where the race03:00
thumperis03:00
davechen1yi'm pushing some debug into mgo.v2 to record the stack of _who_ closed the session that we got handed03:01
davechen1yit's going to be hard to test this if I cannot repro it locally03:01
davechen1yfwwi, the presence tests are unstable AF, but just not this failure ...03:01
thumperdavechen1y: yeah, presence is kinda crap03:05
thumpermenn0: speaking of mgo, what's the status of the idea of having mgo tell us which assert failed ?03:05
axwwallyworld: https://bugs.launchpad.net/juju-core/+bug/1586089  -- I've thought about the azure instance IDs a bit, I think we should do the same thing that thumper has done for LXD/MAAS03:06
mupBug #1586089: azure arm instance-ids are not ids, cannot find machine in azure <azure-provider> <ci> <ha> <jujuqa> <testing> <juju-core:Triaged> <https://launchpad.net/bugs/1586089>03:06
axwwallyworld: atm they're just "machine-0" (e.g.) because instance IDs never clash across models, there's the invisible resource group qualifying them03:07
wallyworldthat could be a good solution03:07
axwwallyworld: but then you need to know which resource group to look in, so having the suffix of the model UUID would make it easier to identify03:07
wallyworldyep, and it's no worse than other providers03:07
axwwallyworld: another thing that needs to be done before beta9 I'm afraid03:08
wallyworldindeed03:08
davechen1y158857403:08
davechen1y logger.Infof("watcher loop failed: %v", err)03:09
davechen1yseriously ...03:09
natefinchdavechen1y: getting notified of which assert failed would probably have saved me days worth of work on my current bug03:09
davechen1yerrors logged at info03:09
menn0thumper: I haven't raised the idea with niemeyer03:10
menn0thumper: it would be a big change b/c the current API returns a static ErrAborted03:10
menn0thumper: all users of mgo would have to change03:10
thumperugh03:10
thumper:(03:11
wallyworldnatefinch: how is the bug going?03:14
natefinchwallyworld: examining the transaction asserts to try to figure out why they're failing.  Luckily have  some success cases and some failure cases to compare03:17
wallyworldok03:17
natefinchwallyworld: btw, the problem was that I was replacing the machine agent's binary, but it was actually logging an error received from an API call to the controller.. so it was the controller that I really needed to replace the binary on.  It just wasn't clear from the error message in the machine agent log that it was an API call. But once I thought about it, I realized it had to be the controller, since it was code in state.  I was too hung up on03:21
natefinchthe assumption that I needed to patch the code where the error was being logged.03:21
* thumper afk to collect kids from school03:21
wallyworldnatefinch: makes sense. easy in hindsight03:22
davechen1y        defer func() {03:23
davechen1y                // If the session is killed from underneath us, it panics when we03:23
wallyworldnatefinch: also, you'll see you got a bug to fix some fallout from the admin to controller name change03:23
davechen1y                // try to copy it, so deal with that here.03:23
davechen1y                if v := recover(); v != nil {03:23
davechen1y                        err = fmt.Errorf("%v", v)03:23
davechen1y                }03:23
davechen1y        }()03:23
davechen1ywhat the F03:23
natefinchwallyworld: good times :/03:24
natefinchdavechen1y: where's that code?03:25
davechen1ystate/presence03:25
wallyworldaxw: should we change the model global key to "m" (currently "e"). this will be our last chance03:29
natefinchdavechen1y: if only someone had reviewed that code: https://github.com/juju/juju/pull/361#discussion-diff-1526692503:31
davechen1ytouche03:32
axwwallyworld: I think m is taken for machines?03:32
natefinchI think it's natural to hate past you.  Past me screws me all the time.03:32
davechen1yI think this is before michael foord discovered just how bad an idea this was03:32
davechen1ynatefinch: https://github.com/juju/juju/pull/553003:32
davechen1ygive that we don't apply the same logic everywhere else we call Copy03:33
wallyworldaxw: yeah, m#123 is, but not "m" on its own03:33
axwwallyworld: I think having m being wildly different m#123 is a recipe for confusion03:33
wallyworldok03:34
wallyworldmy OCD kicks in when I see "e"03:34
axwwallyworld: "e" is for "em" ? :)03:36
wallyworldlol03:36
thumperhaha03:38
thumperperhaps we should rename machine to instance03:38
thumperthen we wouldn't have two m's03:38
wallyworldthumper: i do wish we could solve this. but too hard i think03:40
wallyworld"n" for node maybe03:41
thumpermodel for model03:41
wallyworldglobal keys can only be one letter i think03:41
natefinchI was gonna say, is there a byte shortage?03:41
thumperbollocks03:41
wallyworldin the past i've run into issues03:41
thumperglobal keys are made up from multiple ids03:41
wallyworldthe prefix bit03:41
thumperr#foo#bar#baz03:41
thumperuses split on #03:41
wallyworldright, the r# bit03:42
wallyworldthe r is only one letter03:42
thumperthey just have to be unique in the places it is used03:42
thumperno, can be anything03:42
wallyworldthere's code that assumes it's one letter03:42
thumperit is short by convention03:42
thumperwhere?03:42
wallyworldi'll try and find it03:42
wallyworldi've tried to change it before and it's caused a bug03:42
natefinchrule #1 of code club is: don't encode data in your ids03:43
wallyworldthumper: backingEntityIdForGlobalKey03:43
wallyworldthat would apply to model thugh in this case03:43
wallyworldwouldn't03:44
wallyworldso we could make it "model"03:44
wallyworldthere might be other places too03:44
thumperJFDI and see what  breaks03:44
wallyworldbut the above is one i can find quickly03:44
wallyworldyeah, might do just that03:44
davechen1ymenn0: thumper http://reviews.vapour.ws/r/4977/04:25
menn0davechen1y: lookin04:25
davechen1yta04:27
davechen1ywhy are there three watcher implementations in state04:34
davechen1ywe have state/04:34
davechen1ystate/multiwatcher04:34
davechen1yand state/presence04:34
davechen1yeach with their own watcher model04:34
davechen1ythinggy04:34
menn0davechen1y: review done... not LGTM I'm afraid04:44
menn0davechen1y: the Watcher in state/presence is a completely different thing to the watchers elsewhere04:45
menn0it's unfortunate that the same name was used04:45
natefinchwallyworld: oh, I wanted to talk about the lxc to lxd stuff04:46
wallyworldok04:46
menn0davechen1y: state/watcher implements the low level infrastructure that accepts requests for things to watch, tracks the txn log, and reports when things that are being watched have changed04:47
natefinchwallyworld: if this is the right doc: https://docs.google.com/document/d/1SXBJJ_HHDX4_WvNGVGtLhNajXaioOF_Pn8-EvP1WvaU/edit04:47
natefinchwallyworld: then I'm pretty unclear on which things are tasks Im supposed to do04:47
wallyworld--to lxd04:48
wallyworldlxd in bundles04:48
wallyworldare the first 2 main things04:48
menn0davechen1y: the stuff in state/watcher.go and state/allwatcher.go and state/multiwatcher all rely on what's in state/watcher04:48
natefinchwallyworld: this says to remove --to lxc.. but I think you want it to stick around, warn, but then convert to a --to lxd, correct?04:48
wallyworldwell, we could remove that one. the thing we need to warn about is in bundles04:49
davechen1ymenn0: what a facepalm of tighlyt coupled code04:49
natefinchwallyworld: ok04:49
wallyworldsince we don't want to brek existing bundles04:49
natefinchwallyworld: right04:49
davechen1ymenn0: thanks for the review04:50
wallyworldnatefinch: also stuff like lxd-defaultg-mtu name change04:50
wallyworldinstead of lxc-default-mtu04:50
natefinchAdding "lxd-default-mtu" to mirror "lxc-default-mtu" and pass it through the API stack04:51
davechen1ymenn0: i don't agree with your feedback04:51
natefinchwallyworld: I have no idea what those words mean ^04:51
davechen1ythe session.Copy pattern is everywhere in the code04:51
davechen1yincluding multiple times in the presence package04:52
wallyworldnatefinch: the default mtu is a network setting - "maximum transmission unit"; it defines packet size04:52
natefinchwallyworld: ok, I guess I can grep for it easily enough04:52
davechen1yi think this code was an impleplete solution to that michael foord fixed later04:52
wallyworldnatefinch: the doc says to mirror the setting but i don't see why we can't replace it04:53
wallyworldi'll check with john04:54
natefinchwallyworld: yeah, just renaming it seems fine04:54
wallyworldnatefinch: does this mean the other bug is fixed?04:54
natefinchwallyworld: no, but I wanted to talk to you before tomorrow, so when I get it fixed, I know what to do afterward.04:54
wallyworldyay, ty04:54
wallyworldbet you'll be happy to see the back end of thst bug04:55
natefinchwallyworld: yeah, I looked at the lxd stuff today and realized I was not clear about what I should be doing. Now I am. Thank you.04:55
wallyworldnp04:55
natefinchwallyworld: oh man... this thing was such a cluster just getting it going... like 3 different dumb mistakes stacked on top of one another.  And added on top of needing to deploy openstack each time through another person... not good times.  We lost like an hour or so today when their maas somehow imploded and lost all networking04:56
natefinch(which is one of the times when I had a chance to go looking through the code and the lxd stuff)04:56
natefinchif anyone wants to stare at some mgo transactions and try to figure out why they're failing, there's a nice fat 4meg log here: http://paste.ubuntu.com/16938088/04:58
wallyworldjoy04:58
natefinchsome of the logging is unfortunately duplicated as I was adding in fmt.printlns to try to ensure I wasn't getting screwed by f'd up logging.04:58
natefinchso far I'm not seeing a pattern... the assert that fails is asserting that the current address either doesn't exist or is empty (which probably means it doesn't exist, and the precondition check is returning a zero value for the struct).  Which sounds like it could be bad, except that the exact same kind of assert passes elsewhere04:59
natefinchwallyworld: whelp, I gotta sleep.  Will pick it up in the morning, now that we've figured out all the problems with getting debug builds up, I can sprinkle in smoe more logging, see if I can figure out what's goijng wrong.05:11
wallyworldok, ty05:11
menn0davechen1y: I've replied05:23
davechen1ymenn0: thanks, it's also frustrating that I cannot reproduce this on my machine05:27
davechen1yprotip -- killall mongod while the tests are running; not a big deal05:27
davechen1ynothing panics05:27
davechen1ynothing stops05:27
davechen1ymenn0: session.Copy is nuts05:30
davechen1yall it does is return a copy of the session, pointing to the same backing socket and stuff05:30
davechen1ybut with it's _own_ mutex05:31
davechen1ymenn0: axw thumper https://bugs.launchpad.net/juju-core/+bug/158861405:37
mupBug #1588614: mongodb dying does not cause tests to fail <juju-core:Confirmed> <https://launchpad.net/bugs/1588614>05:37
davechen1ycan you please try to confirm this for me05:37
davechen1yit's a 10 second test05:37
axwdavechen1y: dies straight away for me05:40
axwdavechen1y: the tests exit straight away I mean05:41
davechen1yWOW, this is absolutely the words05:44
davechen1yhttps://github.com/go-mgo/mgo/blob/v2/log.go#L7605:44
davechen1yworst05:44
davechen1yif the race detector is running, add some extra locking to make it happey05:44
davechen1yotherwise -- FUCKING!05:44
davechen1yFUCKIT05:44
davechen1yaxw: can you paste the log05:44
davechen1yseriously, it's been running happily for a few minutes with no bo05:45
davechen1yno db05:45
davechen1yan, no, my mistake05:45
axwdavechen1y: http://paste.ubuntu.com/16940975/05:45
davechen1yi don't get any of that output05:46
davechen1ymy behaviour is entirely different, the tests sit trying to find a working cluster menber for minutes05:47
davechen1ythen give up05:47
axwdavechen1y: you're on master right?05:47
davechen1yyes05:48
axwwallyworld: we may need to remove the ability to bootstrap without first defining a cloud (e.g. bootstrap manual/<IP>), if we're requiring each model to have a cloud name06:30
axwor otherwise come up with a way of generating cloud names06:30
axwit's probably less confusing just to require that they be added first tho06:30
mupBug #1588636 opened: mgo: Panic: Test left sockets in a dirty state (PC=0x46257C) <juju-core:New> <https://launchpad.net/bugs/1588636>06:32
wallyworldhmmm, that was a nice short cut for people, but we could remove it06:33
=== mup_ is now known as mup
=== stokachu_ is now known as stokachu
dimiternfrobware: joining now the HO for top of the hour08:43
dimiternfrobware: i'm in08:45
dimiternjamespage: hey there08:46
frobwaredimitern: otrp08:46
dimiternjamespage: is the ntp charm maintained by the os-charmers?08:47
dimiternfrobware: ok, np08:47
dimiternjamespage: it's the only one not reporting 'Unit is ready' status; see - just 4 lines short of a perfect output: http://paste.ubuntu.com/16942268/ :)08:50
babbageclunkfrobware, dimitern: could you take a look at http://reviews.vapour.ws/r/496708:52
babbageclunkI was hoping to get Menno to look at it but I think he's been too busy.08:52
dimiternbabbageclunk: sure, will take a look08:53
babbageclunkdimitern: thanks!08:54
axwwallyworld: charlotte just reminded me that it's a public holiday here on monday09:06
axwwallyworld: nearly got the guts of my branch done, still tests to write though09:06
perrito666axw: still here?09:35
perrito666axw: I left you a priv message09:37
dimiternbabbageclunk: reviewed09:48
=== \b is now known as benonsoftware
jamespagedimitern, not at the moment11:17
jamespagedimitern, we do have todo's to ensure that all charms that are part of openstack do status11:17
dimiternjamespage: ok11:31
dimiternjamespage: side-note - any plans for neutron to support bindings for the external network as well as the internal (data)?11:32
dimiternjamespage: it looks like the tenant net (and its underlying vlan in maas) also wants to use dhcp and insists on claiming a full /20 (i.e. router's int port uses the .1 address, which is also used by maas..)11:34
frobwaredooferlad: could you try my patch as-is right now with a bonded interface (LACP)?11:46
frobwaredooferlad: http://reviews.vapour.ws/r/4969/11:46
frobwaredooferlad, dimitern: we're screwed11:50
frobwaredooferlad, dimitern: I just got caught out again by /etc/network/interfaces.d/eth0.cfg11:50
frobwaredooferlad, dimitern: guess where my route over eth0 came from... http://pastebin.ubuntu.com/16944207/11:51
* frobware sulks and finds lunch...11:52
mupBug #1588784 opened: juju/state: intermittent test failure in SubnetSuite with mongod 3.2 <mongodb> <unit-tests> <juju-core:New> <https://launchpad.net/bugs/1588784>11:57
babbageclunkfwereade: Here are all the details for that bug, with the standalone reproduction attached.12:05
fwereadebabbageclunk, cool12:05
fwereadebabbageclunk, nice bug report <312:17
frobwaredooferlad, dimitern: the only way I can get stuff to work atm is... https://github.com/frobware/juju/tree/master-lp1588446-hack-hack-hackity-hack12:18
dimiternfrobware: looking12:18
frobwaredimitern: I'll save you the effort: https://github.com/frobware/juju/blob/master-lp1588446-hack-hack-hackity-hack/provider/maas/add-juju-bridge.py#L45412:20
dimiternfrobware: aw, that..12:20
dimiternfwereade: why also remove /e/n/i.d/* ?12:20
dimiternoops12:21
dimiternfrobware: ^^12:21
frobwaredimitern: because eth0 has a running dhclient12:21
frobwaredimitern: from eth0.cfg12:21
frobwaredimitern: so we end up with a route out via eth0 and the nascent br-bond012:21
dimiternfrobware: but, if we replace /e/n/i completely like in my patch?12:21
frobwaredimitern: show me your wares....12:21
dimiternfrobware: https://github.com/juju/juju/pull/5512/files#diff-59aa120b6b815c64439ef2971190a313R6112:22
frobwaredimitern: this is a joke.12:22
frobwaredimitern: we need to stop SPINNING endlessly on this.12:22
dimiternfrobware: nope, that's for containers, sorry..12:22
frobwaredimitern: I was confused...12:23
dimiternfrobware: well, my patch doesn't change where we generate /e/n/i with bridges12:25
dimiternfrobware: but looking at one of the nodes, I can see http://paste.ubuntu.com/16944715/12:25
frobwaredimitern: the problem is we have a  dhclient (unexpected item in bagging area) running against eth012:26
dimiternfrobware: and my patch does rm 50-cloud-init link from /e/systemd/network/12:27
frobwaredimitern: does it kill the dhclient for that iface?12:27
frobwaredimitern: after briding bond0 (parent eth0, eth1) I have http://pastebin.ubuntu.com/16944207/12:28
dimiternfrobware: well, it's not running now on that same node12:28
frobwaredimitern: everything is fine apart from lne 212:28
dimiternfrobware: but haven't checked right after boot12:28
dimiternfrobware: I suspect that link causes sysd to spawn dhclient, not the one in /e/n/i.d/12:28
frobwaredimitern: if your 50-cloud-init.cfg has 'eth0: dhcp' you're screwed too12:28
dimiternfrobware: I need to try with a bond to see how is it12:29
dimiternfrobware: paste the -before-add-juju-bridge /e/n/i ?12:29
frobwaredimitern: http://pastebin.ubuntu.com/16944804/12:30
frobwaredimitern: line 45 is the killer12:30
babbageclunkfwereade: thx :)12:31
dimiternfrobware: yeah..12:31
frobwaredimitern: we're dooomeed!12:31
dimiternfrobware: is this xenial or trusty?12:31
frobwaredimitern: trusty. i.e. customer centric. :)12:31
dimiternfrobware: no, why? we can just filter out any source stanzas12:31
frobwaredimitern: too late12:31
frobwaredimitern: there's already a running dhclient for that eth012:31
dimiternfrobware: didn't one of dooferlad's fixes for 1.25 make sure dhclient is stopped or something?12:32
frobwaredimitern: hence my hack which downs all interfaces, rm's the crap, ifup's then bridges dynamically12:32
dimiternfrobware: let me try my fix with trusty12:33
frobwaredimitern: the probkem is we have two configs for eth0: static and DHCP12:33
frobwaredimitern: my exemplary setup is a bond0 on eth0 and eth112:34
dimiternfrobware: yeah, I get that, but it's also different for upstart and systemd12:34
frobwaredimitern: ok12:34
* frobware really goes for lunch. really really really12:34
tasdomaswhat determines if a deployed unit gets an ipv4 or ipv6 address?12:42
dimiterntasdomas: on what the provider tells us first12:43
tasdomasdimitern - running juju built from master and it seems pretty random to me (also, pretty anoying, since some charms don't handle ipv6 addresses)12:44
dimiterntasdomas: see bug 1576674 for relevant info12:45
mupBug #1576674: 2.0 beta6: only able to access LXD containers (on maas deployed host) from the maas network <lxd> <maas-provider> <oil> <ssh> <juju-core:In Progress by dimitern> <https://launchpad.net/bugs/1576674>12:45
dimiterntasdomas: oops, sorry, not that one..12:45
dimiterntasdomas: this - bug 157484412:45
mupBug #1574844: juju2 gives ipv6 address for one lxd, rabbit doesn't appreciate it. <conjure> <juju-release-support> <landscape> <lxd-provider> <juju-core:Won't Fix> <rabbitmq-server (Juju Charms Collection):Fix Committed by james-page> <https://launchpad.net/bugs/1574844>12:45
dimiterntasdomas: fwiw, those charms should be fixed to accept both types of addresses12:46
dimiterntasdomas: there are ways to work around this, depending on the provider12:46
tasdomasdimitern, lxd?12:46
dimiterntasdomas: i.e. disabling ipv612:46
tasdomasdimitern, agree that charms should handle this, but random assignment is not a good sign anyway12:47
tasdomasdimitern, ah, yeah12:47
dimiterntasdomas: yeah - easy, sudo dpkg-reconfigure lxd and No any IPv6 related questions12:47
dimiterns/No any/answer No to any/12:47
tasdomasdimitern, thankjs12:48
dimiternfrobware: good news :)13:31
dimiternfrobware: smoser says curtin>379 disables the 50-cloud-init generation (currently in xenial-proposed)13:32
dimiterntesting now13:32
frobwaredimitern: how/where/what?13:36
dimiternfrobware: see on #server@c13:36
frobwaredimitern: though I'm happy to come back to good news. :)13:36
frobwaredimitern: I just disconnected - did I miss anything? last was good news, testing now.13:44
dimiternfrobware: sent you a pm with the log13:45
frobwaredimitern: ok, so waiting on r39013:52
dimiternfrobware: AIUI only for precise?13:52
frobwaredimitern: true, but I was testing that and noticed it failed there too13:53
dimiternfrobware: I've seen it fail for precise, but not trusty before (well, at least not every other deployment)13:54
frobwaredimitern: too much shifting sand. This is _exactly_ why we need better automation around this. And "this" is only bootstrap.13:54
mupBug # changed: 1519095, 1573294, 1582731, 1588135, 158813714:12
mupBug #1568895 changed: Cannot add MAAS-based LXD containers in 2.0beta4 on trusty <ci> <jujuqa> <lxd> <maas-provider> <cloud-images:Confirmed> <juju-core:Invalid> <MAAS:Invalid> <https://launchpad.net/bugs/1568895>14:43
mupBug #1584979 changed: better container networking <juju-core:New> <https://launchpad.net/bugs/1584979>14:43
frobwaredimitern: another case for dynamic bridges - https://bugs.launchpad.net/juju-core/+bug/158584715:39
mupBug #1585847: LXD creates all the interfaces as the physical machine has when using MAAS. <lxc> <maas-provider> <network> <spaces> <juju-core:Triaged> <https://launchpad.net/bugs/1585847>15:39
dimiternfrobware: yeah, it was never intended to stay that way :)15:39
dimiternfrobware: nothing present in /e/n/i.d/ on xenial with latest curtin btw15:50
frobware\o/15:51
frobwaredimitern: are you using daily images?15:52
dimiternfrobware: using https://images.maas.io/ephemeral-v2/releases/15:53
frobwaredimitern: ok15:53
mupBug #1588897 opened: Unable to kill or destroy the lxd controller <juju-core:New> <https://launchpad.net/bugs/1588897>16:25
mupBug #1588898 opened: Unable to kill or destroy the lxd controller <juju-core:New> <https://launchpad.net/bugs/1588898>16:25
=== redelmann is now known as rudi|comiendo
=== rudi|comiendo is now known as redelmann
mupBug #1588911 opened: Juju does not support 2.0-beta9 <blocker> <ci> <juju-core:Triaged> <https://launchpad.net/bugs/1588911>17:31
mupBug # changed: 892552, 1287949, 1505504, 1513165, 1514874, 1522544, 1540900, 1557254, 1568122, 1568854, 1568944, 1569361, 1571053, 1572741, 1573136, 1576120, 1576528,18:04
mup1576750, 1577415, 1577609, 1579148, 1580417, 1580418, 1580946, 1580964, 1581074, 1581885, 1581886, 1582620, 1585582, 1585851, 1586217, 1586880, 158689118:04
natefinchmgz_, sinzui, perrito666: https://github.com/juju/version/pull/318:27
natefinch..and now back to debugging this stupid mongo bug18:37
natefinchperrito666: do you have a little time?  I could use some help18:38
natefinchgah, he's idle18:38
=== kwmonroe_ is now known as kwmonroe
mgz_natefinch: looms good to me, I'd expect tag should be alpha only (have we ever defined clearly?), we just need numbers for the buildid18:47
natefinchmgz_: yeah, I don't know. I expect that someone misunderstood \w to mean just alphabet characters, but *shrug*18:53
mgz_natefinch: my one worry of the current change is it's futhering that initial confusion maybe18:54
mgz_and just getting rid of \w is the correct intention18:54
natefinchmgz_: I'm happy to make it alphabeticals and underscore only... or even only a-zA-Z (or even only a-z)18:54
natefinchmgz_: not sure we really every want WoOt_ as a tag18:55
mgz_as far as CI is concerned, all we've ever needed is a-z18:55
natefinchmgz_: I'm happy with that.  I'll make it [a-z]+18:56
mgz_natefinch: sounds good, lets post to -dev or something as well saying what changed so people have a chance to yell18:57
natefinchmgz_: good idea18:58
mgz_natefinch: stamped the pr18:59
mbruzekCan someone advise me how to clean up my controller when I am not able to with juju commands?19:27
mbruzekContext:  https://bugs.launchpad.net/juju-core/+bug/158889819:27
mupBug #1588898: Unable to kill or destroy the lxd controller <juju-core:New> <https://launchpad.net/bugs/1588898>19:27
mbruzekI just want to clean up so juju does not think I have a controller and I can continue to do testing/development.19:31
mbruzekI knew which files to clean up with juju 1.x but 2.x is slightly different.19:32
natefinchmgz_, sinzui: that reminds me, we need to set up CI for github.com/juju/version20:13
natefinchmgz_, sinzui: https://github.com/juju/juju/pull/553320:18
natefinchI gotta step out for a bit... whoever wants that to land is welcome to $$merge$$ it once LGTM'd20:19
mgz_natefinch: okay, I'll merge20:22
rediranyone here understand authorization in apiserver bits?20:46
cmarsredir, like the admin facade? i worked on that a long time ago20:51
cmarsthings have probably changed some20:52
redirlike adding and remoing users cmars20:56
cmarsredir, less familiar with that20:56
redir:) Me too20:58
cmarsredir, i think thumper would know, but he's eow unf20:58
rediryeah20:58
mupBug #1588970 opened: go test .\cmd\jujud\agent hangs on Windows (firewaller) <juju-core:New> <https://launchpad.net/bugs/1588970>21:08

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!