Friday, 2015-11-13

*** diogogmt has joined #openstack-swift00:00
*** jrichli has quit IRC00:01
timburkeclayg: as a guess, we'll try to install dnspython3 on python2 and dnspython on python300:03
claygtimburke: what do you mean "on python3"00:03
claygtimburke: why can't we just install dnspython on the one and only python that swift works with?00:03
notmynameright. dnspython3 is the dnspython project's idea of how to port to py3 (have a separate package). but swift doesn't support py3 anyway?00:04
claygso all we have to do to not have markers is not depend on dnsypython?00:05
claygwouldn't be the first time we inlined a dependency that was doing something stupid00:06
notmynameor perhaps not try to make the distinction now, seeing as it doesn't actually make anything work better? ie get our own house iin order wrt py3 before trying to pull in whatever dependencies have done00:06
notmynamebasically, I'm pretty sure I want to revert
notmynameand I think the pbr inclusion in requirements is just flat out wrong (it's a build-time dependency, not runtime)00:08
claygredbo: when I see something in go like hummingbird.GetRing how am I supposed to know that's in hummingbird/ring.go and not hummingbird/utils.go ?00:08
notmynameclayg: "Namespaces are one honking great idea -- let's do more of those!" ?00:08
claygredbo: like are namespaces not a thing in go - or we just really like grep - or i need better ctags - or swift/go/hummingbird is layed out unconventionally?00:09
clayghurricanerix: ^ what's nadeems handle00:09
claygI feel bad picking on redbo :P00:09
claygnotmyname: not today it isn't?00:09
notmynamenot on now. also past 6 in his TZ, probably gone home00:10
redboall the files in a directory are one package/namespace.  so yeah you pretty much need grep or a smarter IDE.00:11
claygredbo: do you use grep or ctags in vim - or something else that you're liking?00:12
* clayg just wants to be like redbo00:12
redboI use grep00:12
claygfuck yeah00:12
clayglike a man00:12
claygprintf debugging and grep - the way golang is supposed to be00:12
claygbut I seriously don't know why i'm not loggin - it's pretty annoying00:13
claygmy dbench is getting thousands of errors - i tried upping device_limit (or whatever the source named that config option) - but nothing in logs :'(00:13
clayghow about modload imudp in rsyslog.conf, eh?00:15
claygyay permission deined - this'll be much easier now00:16
*** mzhou has quit IRC00:23
pchngclayg: I use Sublime Text, and this plugin for Golang development:
claygpchng: a'ight - so there's that - and it can do the bounce to definition of function stuff?00:23
claygredbo: pchng: maybe godef ->
claygI'll probably just start here ->
pchngclayg: Unfortunately, no, gosublime doesn't. But I don't see why they couldn't add it, a compiled language should make it relatively easier00:25
pchngclayg: What do you use for python?00:26
claygpchng: vim + plugins00:26
pchngI've been using sublime text but wondering if there's something better, i.e. pycharm or whatever00:26
claygpchng: i hear absolutely nothing but good things about pycharm and *constantly* kick myself for not spending time with it00:27
pchngclayg: thanks for the input. well seems like vim-go would be your path of least resistance then :)00:27
claygI looked at it once a few years ago and I griped about it like I do all thick ide's from eclipse to visual stuido - but I'm older and wiser now - so maybe I wouldn't hate it so much00:27
pchngvim-go seems really cool00:27
claygwho knows - it's pathogen'd up in there now - well see - i'll probably hate it00:28
claygi hate everything00:28
pchngclayg: Most of my early development career was spent in Eclipse doing Java. So I've grown a thick skin w.r.t "heavy" IDEs. Eclipse is pretty good for Java, i.e. easily jump to definitions, help with refactoring, etc. (IntelliJ is also good I've heard)00:29
pchngbut for a language like Java you pretty much need that sort of tooling00:29
*** pgbridge has quit IRC00:39
*** m_kazuhiro has joined #openstack-swift00:39
*** Zyric has quit IRC00:46
*** siva_krishnan has quit IRC00:48
hayponotmyname: "what breaks if pbr isn't in requirements.txt?" you cannot use if pbr is too old, because requirements.txt now uses environment markers01:01
haypoclayg: without environment markers, you cannot create tox34: you cannot install swift on python 3, because of dnspython01:01
clayghaypo: sold!01:02
hayponotmyname: another solution if to put pbr requirement in setup.py01:02
haypois to*01:02
hayponotmyname: another solution is to document the requirement on the minimum pbr version (>=1.0 should be enough)01:02
haypoclayg: it's feasible to enhance dnspython to use a single source code for py2 & py3, but it may require a lot of work, and i don't know if dnspython authors like this01:03
haypo(it would avoid the need for env markers, and so swift would work on older pbr version)01:04
haypogood night ;)01:04
notmynamehaypo: right, the pbr in is already a thing01:07
notmynamethats a build-time dependency, not a runtime one (requirements is parsed after
notmynameso unless pbr is now being used in some part of the runtime code, it's not needed in requirements01:08
notmynameof course, it's still a build requirement. the sad truth is that requirements.txt, despite its name, is not a full list of requirements. it's only a list of python dependencies that could be fetched from pypi by pip01:09
*** garthb has quit IRC01:10
notmynameeg sqlite3 is a requirement to run swift. but it's not mentioned in requirements.txt01:10
notmynameI'm less worried about the dnspython[3] mentions. actually, I'm sortof ok with that idea, except for the way it's caused a lot of confusion over the past few days01:11
notmynamesince we aren't close to py3 support in swift today, it seems premature to add a requirement for dnspython3. can we add that later?01:12
notmynamethus I get to my conclusion that I'd like to revert that patch01:12
notmynameI'll leave that conversation for tomorrow. I want to make sure there's a consensus instead of a reactionary "everything's broken, undo! undo!" response01:20
timburkenotmyname: we can probably just add dnspython3 when we actually have py3 tests that need it. iirc, this was at my suggestion back in july to avoid the need for a separate requirements.txt and requirements-py3.txt01:21
timburkeeither way, though, eventually we'll need to swallow this pill01:22
*** miurahr has joined #openstack-swift01:23
*** Chavers has quit IRC01:45
portantenotmyname: has anybody considered a swift / elasticsearch integration story?01:50
*** jrichli has joined #openstack-swift01:58
notmynameportante: yes. there's the openstack searchlight project. softlayer sortof has one with their metadata search. and swiftstack is working on something too02:07
portantecool, thanks02:08
*** zhill has quit IRC02:12
*** jerrygb_ has joined #openstack-swift02:16
*** jerrygb has quit IRC02:17
*** dmorita has joined #openstack-swift02:17
*** zhill has joined #openstack-swift02:20
*** openstackgerrit has quit IRC02:22
*** jerrygb_ has quit IRC02:23
*** openstackgerrit has joined #openstack-swift02:24
*** asettle has quit IRC02:25
*** siva_krishnan has joined #openstack-swift02:25
*** haomaiwang has joined #openstack-swift02:25
*** asettle has joined #openstack-swift02:26
*** goodygum has quit IRC02:28
*** goodygum has joined #openstack-swift02:28
*** ctennis has quit IRC02:29
*** jerrygb has joined #openstack-swift02:30
openstackgerritTim Burke proposed openstack/swift: Close EC fragment iterators in the GreenThread that's consuming them
*** ctennis has joined #openstack-swift02:30
claygtimburke: ^ nice that fixes the functest failure?02:31
timburkeclayg: it should. independently check that it doesn't regress the ChunkReadTimeout or whatever, though02:32
claygtimburke: k i think i remember how to do that...02:33
claygoh - didn't the patch have diff?02:33
timburke(also, it couldn't hurt to have other people who were seeing the func test failures try it out. notmyname i think? i know torgomatic_ couldn't reproduce it...)02:33
timburkeclayg: this buddy?
timburkei suppose it couldn't hurt to bump it a bit *past* the threshold, too. 62s? 65?02:34
*** MooingLemur has quit IRC02:34
claygtimburke: looks clean to me02:35
timburkesweet. did you see the func test failures before?02:35
*** MooingLemur has joined #openstack-swift02:36
*** gmmaha has quit IRC02:37
*** StevenK has quit IRC02:37
*** clyps has quit IRC02:37
*** philipw has quit IRC02:37
*** rcernin has quit IRC02:38
*** clyps has joined #openstack-swift02:38
*** StevenK has joined #openstack-swift02:42
*** philipw_ has joined #openstack-swift02:42
*** gmmaha has joined #openstack-swift02:43
*** philipw has joined #openstack-swift02:44
*** philipw has quit IRC02:44
claygtimburke: i didn't go looking for them02:46
claygtimburke: i was lead to believe you didn't notice them obviously unless you default policy is EC02:46
timburkeyup. and even then it's no sure bet02:47
claygsort of a gab in the integration tooling - need to get something fixed up there02:47
claygtimburke: ok we'll i'm getting my default policy setup now - i'll try and get a a/b test with master02:47
timburkei'm not sure that's it...the QA cluster *did* test the original change, and it passed...02:48
timburkestill, good to have more people try it, so thanks. don't worry too much, though; i'm pretty sure notmyname saw it and will be interested enough in fixing it to verify02:49
*** rjaiswal has joined #openstack-swift02:49
*** wbhuber has joined #openstack-swift02:53
*** haomaiwang has quit IRC03:01
*** haomaiwang has joined #openstack-swift03:01
*** haomaiwang has quit IRC03:03
*** haomaiwang has joined #openstack-swift03:06
*** zhill has quit IRC03:08
claygtimburke: so no functest errors on your branch with ec default - trying the other now03:09
openstackgerritJanie Richling proposed openstack/swift: Decrypting Container Listing
jrichliI left a comment describing what was done for this ^^03:22
claygtimburke: yeah werx03:32
notmynametimburke: thanks. I'll check it out later03:32
notmynamestarted a test run to go while my kids get in bed ;-)03:35
*** gyee has quit IRC03:36
*** lnxnut has joined #openstack-swift03:41
*** klrmn1 has quit IRC03:54
notmynametimburke: worked OMM03:57
*** haomaiwang has quit IRC04:01
*** badari has quit IRC04:01
*** haomaiwang has joined #openstack-swift04:01
*** siva_krishnan has quit IRC04:01
*** wbhuber has quit IRC04:08
*** haomaiwang has quit IRC04:08
*** wbhuber has joined #openstack-swift04:09
*** haomaiwang has joined #openstack-swift04:10
*** wbhuber has quit IRC04:13
*** jrichli has quit IRC04:14
*** klrmn1 has joined #openstack-swift04:21
*** klrmn2 has joined #openstack-swift04:25
*** klrmn1 has quit IRC04:25
*** tsg has joined #openstack-swift04:28
*** lnxnut has quit IRC04:38
*** mac_ified has quit IRC04:42
*** links has joined #openstack-swift04:52
openstackgerritMerged openstack/swift: Close EC fragment iterators in the GreenThread that's consuming them
*** siva_krishnan has joined #openstack-swift04:59
timburkeclayg: notmyname: thanks!05:00
*** MVenesio has joined #openstack-swift05:00
*** haomaiwang has quit IRC05:01
*** haomaiwang has joined #openstack-swift05:01
kota_timburke: good catch, nice work for patch 24177805:04
patchbotkota_: - Close EC fragment iterators in the GreenThread tha...05:04
timburkekota_: thanks, i was trying to reason about it for like a week05:05
*** MVenesio has quit IRC05:07
*** MVenesio has joined #openstack-swift05:07
*** MVenesio_ has joined #openstack-swift05:13
*** MVenesio has quit IRC05:13
*** MVenesio_ has quit IRC05:16
*** MVenesio has joined #openstack-swift05:17
*** MVenesio has quit IRC05:22
*** MVenesio has joined #openstack-swift05:22
*** MVenesio has quit IRC05:26
*** MVenesio has joined #openstack-swift05:26
*** klrmn2 has quit IRC05:30
*** siva_krishnan has quit IRC05:31
*** siva_krishnan has joined #openstack-swift05:31
*** MVenesio has quit IRC05:31
*** MVenesio has joined #openstack-swift05:31
*** MVenesio has quit IRC05:33
*** MVenesio has joined #openstack-swift05:33
*** MVenesio has quit IRC05:34
*** rjaiswal has quit IRC05:35
*** MVenesio has joined #openstack-swift05:35
*** MVenesio has quit IRC05:38
*** MVenesio has joined #openstack-swift05:39
*** MVenesio has quit IRC05:40
*** MVenesio has joined #openstack-swift05:40
*** matt_____ has quit IRC05:44
*** haomaiwang has quit IRC05:52
*** MVenesio has quit IRC05:55
*** haomaiwang has joined #openstack-swift05:55
*** MVenesio has joined #openstack-swift05:55
*** trifon has joined #openstack-swift05:56
*** haomaiwang has quit IRC06:01
*** haomaiwa_ has joined #openstack-swift06:01
*** MVenesio_ has joined #openstack-swift06:04
*** MVenesio_ has quit IRC06:06
*** MVenesio has quit IRC06:06
*** MVenesio has joined #openstack-swift06:06
*** MVenesio has quit IRC06:09
*** MVenesio has joined #openstack-swift06:10
*** rcernin has joined #openstack-swift06:10
*** tsg has quit IRC06:11
*** haomaiwa_ has quit IRC06:12
*** MVenesio has quit IRC06:13
*** MVenesio has joined #openstack-swift06:14
*** MVenesio has quit IRC06:15
*** MVenesio_ has joined #openstack-swift06:15
*** siva_krishnan has quit IRC06:15
*** MVenesio_ has quit IRC06:17
*** MVenesio has joined #openstack-swift06:17
*** jamielennox is now known as jamielennox|away06:17
*** haomaiwa_ has joined #openstack-swift06:21
*** MVenesio has quit IRC06:22
*** MVenesio has joined #openstack-swift06:23
*** MVenesio has quit IRC06:24
*** MVenesio has joined #openstack-swift06:25
*** rohit_ has quit IRC06:26
*** MVenesio_ has joined #openstack-swift06:26
*** MVenesio has quit IRC06:27
*** rcernin_ has joined #openstack-swift06:27
*** mahatic has joined #openstack-swift06:30
*** MVenesio has joined #openstack-swift06:32
*** MVenesio has quit IRC06:33
*** MVenesio has joined #openstack-swift06:34
*** MVenesio_ has quit IRC06:34
*** haomaiwa_ has quit IRC06:35
*** haomaiwang has joined #openstack-swift06:36
*** rohit_ has joined #openstack-swift06:40
*** rjaiswal has joined #openstack-swift06:40
*** MVenesio_ has joined #openstack-swift06:41
*** MVenesio has quit IRC06:42
*** MVenesio_ has quit IRC06:50
*** MVenesio has joined #openstack-swift06:50
*** MVenesio_ has joined #openstack-swift06:54
*** MVenesio has quit IRC06:57
*** haomaiwang has quit IRC06:58
*** haomaiwa_ has joined #openstack-swift06:59
*** haomaiwa_ has quit IRC07:01
*** haomaiwang has joined #openstack-swift07:01
*** pchng_ has joined #openstack-swift07:03
*** pchng has quit IRC07:06
*** rcernin is now known as Guest1636007:08
*** rcernin_ has quit IRC07:09
*** Guest16360 has quit IRC07:09
*** rcernin_ has joined #openstack-swift07:10
*** MVenesio_ has quit IRC07:10
*** MVenesio has joined #openstack-swift07:10
*** MVenesio has quit IRC07:12
*** MVenesio has joined #openstack-swift07:12
*** rcernin_ is now known as rcernin07:13
*** MVenesio has quit IRC07:14
*** MVenesio has joined #openstack-swift07:14
*** haomaiwang has quit IRC07:18
*** wolsen has quit IRC07:23
*** MVenesio has quit IRC07:23
*** MVenesio has joined #openstack-swift07:23
*** MVenesio has quit IRC07:28
*** MVenesio has joined #openstack-swift07:29
*** MVenesio has quit IRC07:35
*** esker has quit IRC07:46
*** haomaiwang has joined #openstack-swift07:51
*** haomaiwang has quit IRC08:01
*** arnox has joined #openstack-swift08:01
*** haomaiwang has joined #openstack-swift08:01
*** miurahr has quit IRC08:01
*** rledisez has joined #openstack-swift08:14
hayponotmyname: hi. it's not possible to add the env markers on dnspython/dnspython3 later, it's a blocker point to start working on python308:20
hayponotmyname: the next milestone is to have a working "tox -e py34", so we must be able to create the venv08:20
*** jamielennox|away is now known as jamielennox08:43
*** rjaiswal has quit IRC08:45
*** jmccarthy has quit IRC08:53
*** ho has quit IRC08:53
*** jmccarthy has joined #openstack-swift09:00
*** haomaiwang has quit IRC09:01
*** haomaiwang has joined #openstack-swift09:01
*** openstackgerrit has quit IRC09:01
*** openstackgerrit has joined #openstack-swift09:01
*** dmorita has quit IRC09:02
*** rohit_ has quit IRC09:06
*** jordanP has joined #openstack-swift09:17
*** joeljwright has joined #openstack-swift09:18
*** ChanServ sets mode: +v joeljwright09:18
*** jistr has joined #openstack-swift09:47
*** haomaiwang has quit IRC10:01
*** haomaiwang has joined #openstack-swift10:01
*** marzif has joined #openstack-swift10:14
*** openstackgerrit has quit IRC10:16
*** openstackgerrit has joined #openstack-swift10:16
*** esker has joined #openstack-swift10:46
*** diogogmt has quit IRC10:51
*** BAKfr has quit IRC10:54
*** BAKfr has joined #openstack-swift10:55
*** acoles_ is now known as acoles10:57
*** resker has joined #openstack-swift11:00
*** haomaiwang has quit IRC11:01
*** haomaiwang has joined #openstack-swift11:01
*** esker has quit IRC11:03
*** resker has quit IRC11:07
*** lpabon has joined #openstack-swift11:29
openstackgerritPeter Lisák proposed openstack/swift: Log error if a local device not identified
openstackgerritMahati Chamarthy proposed openstack/swift: Modify unit tests to include real crypto
*** haomaiwang has quit IRC12:01
*** haomaiwang has joined #openstack-swift12:01
mahaticacoles: jrichili ^ rebased on patchset 13 of 21443812:03
*** haomaiwang has quit IRC12:04
*** marzif has quit IRC12:10
*** kei_yama has quit IRC12:13
*** jcidaho has joined #openstack-swift12:22
*** lpabon has quit IRC12:27
*** MVenesio has joined #openstack-swift12:32
*** MVenesio has quit IRC12:33
*** m_kazuhiro has quit IRC12:34
*** jmccarthy has quit IRC12:35
*** jmccarthy has joined #openstack-swift12:38
*** CR7 has joined #openstack-swift12:45
*** links has quit IRC12:46
*** haomaiwang has joined #openstack-swift13:03
*** jcidaho has quit IRC13:09
*** Guest21878 has quit IRC13:09
*** links has joined #openstack-swift13:13
*** jcidaho has joined #openstack-swift13:14
*** jerrygb has quit IRC13:17
*** mac_ified has joined #openstack-swift13:23
*** marzif has joined #openstack-swift13:26
*** jcidaho has quit IRC13:28
*** tsg has joined #openstack-swift13:29
openstackgerritOpenStack Proposal Bot proposed openstack/swift: Updated from global requirements
*** mac_ified has quit IRC13:43
*** itlinux has joined #openstack-swift13:48
*** blmartin has joined #openstack-swift13:53
*** itlinux has quit IRC13:53
*** haomaiwang has quit IRC14:01
*** haomaiwang has joined #openstack-swift14:01
*** openstackgerrit has quit IRC14:01
*** openstackgerrit has joined #openstack-swift14:01
*** pchng_ is now known as pchng14:10
*** tsg has quit IRC14:20
*** petertr7_away is now known as petertr714:28
*** esker has joined #openstack-swift14:28
*** blmartin has quit IRC14:29
*** rohit_ has joined #openstack-swift14:29
*** itlinux has joined #openstack-swift14:30
*** proteusguy__ has quit IRC14:31
*** esker has quit IRC14:33
*** itlinux has quit IRC14:35
*** badari has joined #openstack-swift14:35
*** proteusguy__ has joined #openstack-swift14:44
*** jcidaho has joined #openstack-swift14:52
*** pdardeau has joined #openstack-swift14:52
*** jerrygb has joined #openstack-swift14:57
*** itlinux has joined #openstack-swift14:59
*** haomaiwang has quit IRC15:01
*** wbhuber has joined #openstack-swift15:01
*** blmartin has joined #openstack-swift15:01
*** haomaiwang has joined #openstack-swift15:01
*** wbhuber_ has joined #openstack-swift15:01
*** CR7 has quit IRC15:03
*** itlinux has quit IRC15:03
*** lcurtis has joined #openstack-swift15:05
*** wbhuber has quit IRC15:05
*** marzif_ has joined #openstack-swift15:08
*** marzif has quit IRC15:12
*** breitz has quit IRC15:18
*** breitz has joined #openstack-swift15:19
*** tsg has joined #openstack-swift15:29
Guest90281does anyone know how to make sure the proxyserver listens on all IPs instead of just loopback ?15:35
*** chsc has joined #openstack-swift15:36
*** jrichli has joined #openstack-swift15:38
*** siva_krishnan has joined #openstack-swift15:41
Guest90281got it .. just edit proxy-server.conf to have instead15:43
*** petertr7 is now known as petertr7_away15:44
*** jistr is now known as jistr|call15:44
*** petertr7_away is now known as petertr715:44
*** jlhinson has joined #openstack-swift15:45
*** CaioBrentano has joined #openstack-swift15:46
notmynamegood morning15:50
*** jlhinson has quit IRC15:56
hayponotmyname: hi. tsg told me that he played with pip on a fresh VM to try to understand the pbr/pip/env marker issues15:57
notmynamehaypo: AIUI, the "issues" are resolved by reinstalling updated versions of everything. so it's not that it isn't solvable, just frustrating15:58
haypotsg: if i recall correctly, pbr 1.6 doesn't support env markers and so fail at "pip install -r requirements.txt" ?15:58
tsgnotmyname: on my fresh trusty install, the only thing I had to do to fix was to upgrade pip to 1.7.x15:59
hayponotmyname: if i understood correctly, your problem is that developers working on swift have to update pip and/or pbr, otherwise they cannot installl swift anymore, right?15:59
notmynametsg: and probably use pip setuptools instead of distro packaged one?15:59
tsgthe problem is if you upgrade pip with "pip install -U pip", it doesn't uninstall the pip installed by python-pip package15:59
hayponotmyname: i hope that deployers don't use virtualenv, or things like that, but regular linux packages15:59
*** jcidaho has quit IRC15:59
tsgthat's the key .. when you do "pip install -U pip" it installed the new pip to /usr/local/bin and the system still keeps using /usr/bin/pip16:00
tsg(1.5.x on trusty)16:00
haypotsg: i don't understand your /usr/local issue. i'm using regular "pip install -U pip" and /usr/bin/pip* are updated16:00
haypotsg: but i start from a more recent OS, fedora 22 or 23 usually16:00
tsghaypo: I don't know but on Ubuntu they go to /usr/local/bin it appears16:00
*** haomaiwang has quit IRC16:01
tsghaypo: CI uses Ubuntu not Fedora16:01
notmynamehaypo: yeah, that's most likely. deployers build and deploy their own packages or use distro provided ones16:01
haypotsg: CI uses DevStack which handles well pbr & pip16:01
*** haomaiwang has joined #openstack-swift16:01
haypotsg: i'm quite sure that DevStack updates pbr & pip very early. otherwise, bad things happen :)16:01
notmynamehaypo: I understand the dnspython version markers now (why that's needed). I still think pbr in requirements isn't needed16:02
tsgeither way, notmyname: when I upgrade pip to 1.7.x, uninstall python-pip and then reload my shell to clear any old paths etc, the marker errors are not seen16:02
notmynametsg: ok. yeah, that's what I'd expect. thanks for looking into it16:02
hayponotmyname: about pbr, i don't care of having it in requirements.txt. again, we can remove it from requirements.txt, maybe add the requirements to, maybe don't specify the minimum version anywhere (except in the doc?)16:02
notmynameit's already specified in setup.py16:03
*** jistr|call is now known as jistr16:03
hayponotmyname: for, i mean specify the minimum pbr version. as adviced by openstack requirements16:03
haypoputting pbr>=x.y in doesn't cover all use cases because is based on setuptools which is unable to upgrade itself16:04
haypoand if i recall correctly, recent versions of pbr require a recent version of setuptools16:04
haypoall the stuff are linked altogether for our pleasure16:05
notmynamebut since requirements.txt is parsed *after*, specifying it there doesn't actually do anything extra for getting swift installed16:05
*** trifon has quit IRC16:08
*** klrmn1 has joined #openstack-swift16:08
*** rcernin has quit IRC16:08
* notmyname needs to step away for a bit16:09
*** pgbridge has joined #openstack-swift16:11
*** mzhou has joined #openstack-swift16:13
*** links has quit IRC16:13
*** wolsen has joined #openstack-swift16:13
*** marzif_ has quit IRC16:16
timburkehaypo: specifying a minimum pbr version in isn't sufficient, either; it won't be available in time to be useful. i remember talking about it with lifeless months ago in another context:
*** tongli has joined #openstack-swift16:20
haypotimburke: it helps at least users to guess that they need :)16:21
haypotimburke: if they open setup.py16:21
haypotimburke: but yes, it doesn't enough to cover all use cases16:23
notmynametimburke: that friday rant about pbr/whatever was only 4 months ago. I guess we get to start scheduling one every few months?16:28
*** mahatic has quit IRC16:30
jrichlisince I started swift in Feb 2015, the only times that I remember having to do more than just the usual setup command was back then (4 months ago) and now16:33
openstackgerritPeter Lisák proposed openstack/python-swiftclient: Unification of manpages and docstrings
*** klrmn1 has quit IRC16:35
wbhuber_notmyname: planning to update this page to reflect the object storage installation for OS liberty if anything major has been changed:
wbhuber_notmyname: i am planning to update my cluster to swift 2.5.1 from pip16:43
*** diazjf has joined #openstack-swift16:45
*** jistr is now known as jistr|call16:46
*** esker has joined #openstack-swift16:49
notmynamewbhuber_: great! thanks16:51
notmynamebus is coming, commute time16:51
*** arnox has quit IRC16:56
*** CaioBrentano has quit IRC16:58
*** klrmn1 has joined #openstack-swift16:59
*** klrmn1 has quit IRC16:59
*** haomaiwang has quit IRC17:01
*** haomaiwang has joined #openstack-swift17:01
*** diazjf has quit IRC17:04
*** gyee has joined #openstack-swift17:04
*** diazjf has joined #openstack-swift17:06
*** marzif has joined #openstack-swift17:07
*** jistr|call is now known as jistr17:08
*** jlhinson has joined #openstack-swift17:11
*** jistr has quit IRC17:12
*** rledisez has quit IRC17:24
*** CaioBrentano has joined #openstack-swift17:24
openstackgerritTim Burke proposed openstack/swift: Add container and account reverse listings
openstackgerritTim Burke proposed openstack/swift: Have versioned_writes use reverse listings
*** haomaiwang has quit IRC17:30
*** daemontool_ has quit IRC17:32
claygtimburke: what's the over under on me finding another bug in reverse listings with
timburkeclayg: i'm feeling good about it. best version yet :)17:34
timburkeand hey, once we find (and fix) *all* the bugs, we can merge it!17:35
*** diazjf has quit IRC17:37
*** zhill has joined #openstack-swift17:37
*** CaioBrentano has quit IRC17:42
*** daemontool_ has joined #openstack-swift17:42
*** CaioBrentano has joined #openstack-swift17:44
*** diazjf has joined #openstack-swift17:44
openstackgerritJohn Dickinson proposed openstack/swift: remove pbr from requirements.txt. It's not a run-time dependency
notmynamehaypo: ^17:46
*** daemontool_ has quit IRC17:49
*** BAKfr has quit IRC17:51
*** petertr7 is now known as petertr7_away17:51
*** marzif has quit IRC17:54
*** BAKfr has joined #openstack-swift17:56
*** jrichli has quit IRC18:01
*** petertr7_away is now known as petertr718:01
acolesclayg: ack and thanks for comment on patch 135380. i like your idea but be a while before i cam implement it (i'm just here for a short while today and away again next week)18:03
patchbotacoles: - Update container on fast-POST18:03
acolesclayg: also, swift_bytes appended to swift_bytes + POST updates to content-type is also going to take me some time to work around18:04
acolesclayg: i think it can be handled, but will need "more code"18:04
acolesargh, i meant to type " swift_bytes appended to content-type"18:05
*** siva_krishnan has quit IRC18:06
*** jlhinson has quit IRC18:06
*** diazjf has quit IRC18:07
*** klrmn1 has joined #openstack-swift18:08
*** pdardeau has quit IRC18:17
*** joeljwright has quit IRC18:17
acolespeluse: thanks for your reviews, hope to catch up on them in a few days18:22
*** esker has quit IRC18:31
*** jrichli has joined #openstack-swift18:44
*** mfalatic has quit IRC18:47
*** siva_krishnan has joined #openstack-swift18:50
claygacoles: nice work19:00
*** badari_ has joined #openstack-swift19:00
claygacoles: I think jrichli also needs to "fix" the swift_bytes appended to content-type a little bit19:01
acolesclayg: yes. we're working on that too.19:01
*** badari has quit IRC19:03
*** esker has joined #openstack-swift19:03
acolesfun fun fun19:03
*** siva_krishnan has quit IRC19:04
claygacoles: pain pain pain19:04
jrichlistoring infomation is hard ;-)19:04
acolesno pain no gain19:04
*** Guest66106 has joined #openstack-swift19:04
notmynamejrichli: only if you want to get it back later19:04
*** Guest66106 has quit IRC19:05
*** garthb has joined #openstack-swift19:05
jrichlinotmyname: very true!19:05
claygl84 Guest90281 !19:05
Guest90281clayg: I'm sensing that it's a confirmation that what I did is right ... but I'm not able to get the reference - what's l84 ?19:09
*** CaioBrentano has quit IRC19:09
acolesjrichli: if I disable versioned writes I see func test errors drop from 90 to < 20. i.e. a whole bunch of tests start to pass. oh boy.19:10
claygGuest90281: oh you typed in "exit" - I thought you were leaving so I said "later"19:10
acolesso a bunch of those errors are in test setup due to some bad state that other tests have left behind.19:11
jrichliacoles: previous to the patch your are working with: we already had 2 cards related to versioned writes.  you can look at those cards for baseline19:11
peluseacoles: cool, will be back on them here again shortly as well.  are you done with your legal mumbo jumbo?19:11
claygGuest90281: I think it's a typo of l8r - which is a phonetic 1337 speak for later - I think most people find it's easier to ignore me - sorry about that19:11
Guest90281Woah .. I did ? nope .. not leaving .. and thanks for the shortcut19:12
Guest90281and yes I guessed that it should be r instead of 8 - now.19:12
*** petertr7 is now known as petertr7_away19:12
Guest90281instead of 4.19:12
claygoh whoa - no Guest66106 *said* exit!  hard to keep you guys stright19:12
claygyou should change your handle!19:12
acolespeluse: not yet. i got let off today but have to go back next week.19:12
acolespeluse: in our system i may get to act on multiple cases in a 2 week window19:13
Guest90281clayg: I tried to register myself as papercup - but apparently smuxi client has some weird stale handle issue going on - not allowing me to change it back19:13
*** badari_ is now known as badari19:13
*** Guest33827 has joined #openstack-swift19:13
*** CaioBrentano has joined #openstack-swift19:16
jrichliacoles: yes, i noticed that there was some sort of state thing going on from previous tests19:16
openstackgerritZack M. Davis proposed openstack/swift: remove remaining simplejson uses
acolesjrichli: ok i think i fugured out whats happening19:26
*** thumpba has joined #openstack-swift19:26
acolesjrichli: it all broke becaue you fixed it :)19:26
*** petertr7_away is now known as petertr719:26
acolesjrichli: try patching the keymaster to use the *same* key for containers and objects i.e. the same for everything (just set key_id="fixed")19:28
jrichliacoles : ok.  you have peaked my curiosity19:29
*** jlhinson has joined #openstack-swift19:29
acolesjrichli: the state cleanup starts to work!19:30
jrichliacoles: this just gets more and more interesting19:30
acolesmy guess is that object copies cause container updates with the wrong content type cipher text - the copy ends up sending the object key encrypted c-type in the container update19:31
acolesand then, when the container is listed (which tests will do to try to cleanup) the decrypter uses the container key and the result is garbage which causes the json/xml processing to barf19:32
*** pdardeau has joined #openstack-swift19:32
acolesjrichli: i am down to FAILED (SKIP=12, errors=7, failures=17)19:32
acolesjrichli: but before you fixed the content-type crypt meta encoding in patchset 14 we never even *tried* to decrypt the container listing, so somehow avoided the barfs.19:33
*** thumpba has quit IRC19:34
acolesjrichli: thats a hypothesis of course ^^ :)19:34
acolesjrichli: or, to summarize in another way, today i achieved nothing :D19:34
jrichliacoles: ok, sounds like your sluthing skills have paid off :-)  but i am still confused because the base64 encoding work-around should have revealed this a long time ago too, right?19:35
*** thumpba has joined #openstack-swift19:35
jrichliacoles: I donno.  I gotta stew on all this after I take some sort of mental break today.  My brain just doesn't want to dig very deep right now.19:37
jrichliacoles: plus, the vm i was working in is no longer responding to mouse movement :-(19:37
acolesjrichli: idk. like i said, its a hypothesis.19:38
acolesjrichli: sounds like you and your vm are in the same place ;)19:38
acolesjrichli: i'm scrapping the email i was writing. i thought i had actually improved things but realise now i just got confused by the test state coupling.19:40
jrichliacoles: oh, ok.  i bet i might have learned something by reading it anyway, but whatever :-)19:42
*** thumpba has quit IRC19:44
*** thumpba has joined #openstack-swift19:45
*** diazjf has joined #openstack-swift19:48
*** esker has quit IRC19:48
*** esker has joined #openstack-swift19:49
acolesjrichli: have a good weekend hope, you get some rest19:52
jrichliacoles: thanks!  have a good time as a juror :-)19:53
*** acoles is now known as acoles_19:53
*** proteusguy__ has quit IRC20:00
*** proteusguy_ has joined #openstack-swift20:04
*** thumpba has quit IRC20:05
*** thumpba has joined #openstack-swift20:07
*** alejandrito has joined #openstack-swift20:16
*** thumpba has quit IRC20:18
*** thumpba has joined #openstack-swift20:19
*** siva_krishnan has joined #openstack-swift20:21
*** thumpba has quit IRC20:21
*** jamielennox is now known as jamielennox|away20:23
*** thumpba has joined #openstack-swift20:26
openstackgerritZack M. Davis proposed openstack/swift: remove Python 2.6-specific logging workaround
*** haomaiwang has joined #openstack-swift20:31
*** Shrews has joined #openstack-swift20:32
ShrewsHello! The infra project, shade, is using swift service to do object uploading (see We've had this code for some time, and I know at one point it *did* work. But our functional tests have been broken (unnoticed), and now we see that the iterator returned from the upload() method is not returning what we expect (a dict20:35
Shrewswith the 'success' attribute).20:35
*** haomaiwang has quit IRC20:35
ShrewsInstead, we get back a string: 'gi_running'20:35
ShrewsDid this change at some point?20:36
ShrewsI can't find *any* examples of using this... so hard to tell what we're doing wrong.20:37
openstackgerritTim Burke proposed openstack/swift: Remove references to now-missing havana docs
*** petertr7 is now known as petertr7_away20:44
*** CaioBrentano has quit IRC20:47
*** thumpba has quit IRC20:49
wbhuber_timburke, notmyname: looks good ^ thx20:53
*** thumpba has joined #openstack-swift20:54
*** wbhuber_ has quit IRC20:55
*** wbhuber has joined #openstack-swift20:55
*** wbhuber_ has joined #openstack-swift20:56
Shrewsoh, ignore my question. just found our bug20:58
timburkeShrews: you're submitting a single upload job there and then iterating over the result, right? won't that mean we're iterating over the keys in the dict?20:58
timburkeah, never mind :)20:58
Shrewstimburke: we were inadvertently trashing the returned generator20:58
*** jlhinson has quit IRC21:00
*** wbhuber has quit IRC21:00
*** jlhinson has joined #openstack-swift21:00
*** wbhuber_ has quit IRC21:00
*** mzhou_ has joined #openstack-swift21:01
*** diazjf has quit IRC21:01
*** mzhou has quit IRC21:03
*** mzhou_ is now known as mzhou21:03
*** petertr7_away is now known as petertr721:06
*** siva_krishnan has quit IRC21:07
*** siva_krishnan has joined #openstack-swift21:08
*** garthb has quit IRC21:11
*** Guest33827 has left #openstack-swift21:14
notmynametsg: just got back from a meeting. I saw you saying soemting in infra about pyeclib. what's the summary?21:16
tsgnotmyname: I was talking to clarkb and mordred about getting liberasurecode into the CI slaves. One of the questions was which CI variants (precise, trusty, centos6/7, etc) that the package needs to be added for - apparently the gate-swift-python27 jobs are run on precise, trusty and centos6 at the moment, but centos6 and precise will be eliminated by the end of the month.21:20
*** blmartin has quit IRC21:21
notmynametsg: what's needed from our side? and when can we get the updated pyeclib stuff landed?21:21
*** bill_az has joined #openstack-swift21:21
tsgnotmyname: I noticed yesterday that someone on the Debian team had added a libjerasure2 dependency to the liberasurecode1 package:
tsgzigo is working on a new build right now to fix that21:22
tsg(remove that)21:22
tsgnotmyname: I have an active bug to backport the stuff to trusty21:23
openstackLaunchpad bug 1515710 in trusty-backports "Please backport liberasurecode-1.1.0, python-pyeclib-1.1.1 to trusty" [Undecided,New]21:23
notmynametsg: ok. and those needt o get build and deployed to the CI boxes before can land?21:23
tsgnotmyname: the only step after that is to get the package included in
tsgnotmyname: yes - in parallel, I am also talking to clarkb etc to see how the wheel issue can be worked around so we can get the swift reqs patch landed sooner21:24
notmynametsg: thanks for working on it21:25
tsgnotmyname: np21:25
*** diazjf has joined #openstack-swift21:25
*** thumpba has quit IRC21:26
notmynameseems that might be the same thing blocking
zigotsg: notmyname: liberasurecode 1.1.0-2 uploaded with libjerasure2 dependency removed.21:26
tsgzigo: thanks a lot! :)21:27
tsgzigo: let me try a backport and comment on the trusty backport bug so it can move forward21:27
zigoIt will reach Debian mirrors within at least 5 hours (next dak run is in more than 4 hours...)21:27
notmynametsg: zigo: isn't the difference between 1.1.1 and 1.1.0 something about a path? didn't we have an upgrade problem wtih 1.1.0?21:27
tsgnotmyname: this is liberasurecode only21:27
notmynametsg: oh ok21:28
tsgwe didn't have to do a 1.1.1 for liberasurecode21:28
notmynamebut we did21:28
*** thumpba has joined #openstack-swift21:28
notmynamebut we didn't have to?21:28
tsgnotmyname: nope .. it is still 1.1.021:28
notmynameoh ok21:28
tsgnotmyname: zigo simply rebuilt a 1.1.0-2 with the same source, with jerasure2 removed from dependencies ..21:29
notmynametsg: ah, yes, I see now21:30
*** david-lyle has quit IRC21:34
*** wbhuber has joined #openstack-swift21:48
*** wbhuber has quit IRC21:48
*** wbhuber has joined #openstack-swift21:48
*** wbhuber_ has joined #openstack-swift21:55
*** Shrews has left #openstack-swift21:56
*** wbhuber has quit IRC21:57
*** chsc has quit IRC22:01
*** jrichli has quit IRC22:09
*** thumpba has quit IRC22:13
*** thumpba has joined #openstack-swift22:15
*** petertr7 is now known as petertr7_away22:19
*** breitz has quit IRC22:25
*** breitz has joined #openstack-swift22:25
*** alejandrito has quit IRC22:32
*** wbhuber_ is now known as wbhuber22:33
*** thumpba has quit IRC22:36
*** pdardeau has quit IRC22:37
*** thumpba has joined #openstack-swift22:37
*** thumpba has quit IRC22:43
*** Guest90281 is now known as layzee22:44
*** jlhinson has quit IRC22:45
*** thumpba has joined #openstack-swift22:46
*** thumpba has quit IRC22:50
*** thumpba has joined #openstack-swift22:54
*** thumpba has quit IRC22:55
*** siva_krishnan has quit IRC23:00
*** siva_krishnan has joined #openstack-swift23:01
*** mac_ified has joined #openstack-swift23:01
*** badari has quit IRC23:10
*** diazjf has quit IRC23:10
*** diazjf has joined #openstack-swift23:12
*** diazjf has quit IRC23:13
*** breitz has quit IRC23:16
*** siva_krishnan has quit IRC23:24
*** siva_krishnan has joined #openstack-swift23:25
openstackgerritSamuel Merritt proposed openstack/swift: Monkeypatch everything the same
*** siva_krishnan has quit IRC23:28
*** siva_krishnan has joined #openstack-swift23:29
*** zhill has quit IRC23:30
*** jlhinson has joined #openstack-swift23:32
*** okdas has quit IRC23:32
*** okdas has joined #openstack-swift23:33
*** siva_krishnan has quit IRC23:33
*** siva_krishnan has joined #openstack-swift23:33
*** okdas has quit IRC23:33
*** okdas has joined #openstack-swift23:33
*** lcurtis has quit IRC23:34
*** km has joined #openstack-swift23:37
*** km is now known as Guest9898223:37
*** siva_krishnan has quit IRC23:40
*** wbhuber has quit IRC23:51
*** mzhou has quit IRC23:55
*** thumpba has joined #openstack-swift23:56

Generated by 2.14.0 by Marius Gedminas - find it at!