Monday, 2015-02-09

*** nellysmitt has joined #openstack-swift00:15
*** acoles_away has quit IRC00:19
*** nellysmitt has quit IRC00:19
*** lcurtis has joined #openstack-swift00:23
*** lcurtis has quit IRC00:31
*** acoles_away has joined #openstack-swift00:32
*** acoles_away is now known as acoles00:33
*** ChanServ sets mode: +v acoles00:33
*** oomichi has joined #openstack-swift00:41
*** bkopilov has quit IRC00:48
*** bkopilov has joined #openstack-swift00:51
*** bkopilov has quit IRC01:23
*** bkopilov has joined #openstack-swift01:24
*** goodes has quit IRC01:38
*** serverascode has quit IRC01:38
*** ipolyzos has quit IRC01:39
*** nottrobin has quit IRC01:39
*** briancurtin has quit IRC01:40
*** zhiyan has quit IRC01:40
*** ipolyzos has joined #openstack-swift01:45
*** nottrobin has joined #openstack-swift01:46
*** briancurtin has joined #openstack-swift01:46
*** zhiyan has joined #openstack-swift01:46
*** goodes has joined #openstack-swift01:51
*** serverascode has joined #openstack-swift01:52
*** lcurtis has joined #openstack-swift01:56
*** nellysmitt has joined #openstack-swift02:16
*** lcurtis has quit IRC02:17
*** nellysmitt has quit IRC02:20
*** lcurtis has joined #openstack-swift02:20
*** lcurtis has quit IRC02:59
jamielennoxhi, is someone able to help me debug http://logs.openstack.org/72/150672/4/check/check-swift-dsvm-functional/a36b2d4/console.html#_2015-02-09_01_37_14_75003:19
jamielennoxthe review is https://review.openstack.org/#/c/150672/ which is in devstack and makes it so that the devstack users aren't configured with a default_project_id03:20
jamielennoxwhich means that you have to supply a project_id when doing authentication to get a role in a project03:20
jamielennoxand so somewhere in swift's functional tests there is a mistake in authentication which means that the authentication of the tests is not correctly having the project_id set03:21
notmynamehmm..looks like the functional test config was updated to a v3 endpoint.03:21
jamielennoxnotmyname: where's that set?03:22
notmyname /etc/swift/test.conf03:22
jamielennoxnotmyname: so would adding admin_tenant_name=services to the swift/test/sample.conf do it?03:24
notmynamejamielennox: I've got to go take care of evening family stuff (ie putting kids to bed). acoles is who'd I'd ask about that. he's (probably) just recently landed in SF from LON, so I have no idea if he's awake :-)03:24
jamielennoxnotmyname: no worries - my next step was going to be to step through it anyway, i was just hoping that someone would be around who could do it quicker03:25
jamielennoxor at least knew where to point me03:25
notmynamelook in swift/test/functional/__init__.py03:25
notmynameI'm not sure if that variable is plumbed through or not03:25
notmynamebut in general, that's the config and that's the file it's read in03:26
notmynamesorry to run, but I hope that helps03:26
jamielennoxnotmyname: it's a start - thanks03:26
*** miqui has quit IRC03:41
*** lnxnut has quit IRC04:04
*** lnxnut has joined #openstack-swift04:13
*** nellysmitt has joined #openstack-swift04:17
*** nellysmitt has quit IRC04:21
*** ppai has joined #openstack-swift04:31
*** BaconIsKing has joined #openstack-swift04:32
*** jasondotstar has joined #openstack-swift04:36
*** BaconIsKing has quit IRC04:39
*** silor has joined #openstack-swift04:39
*** lnxnut has quit IRC04:43
*** rmcall has joined #openstack-swift04:45
*** lnxnut has joined #openstack-swift04:49
*** lnxnut has quit IRC04:53
*** BaconIsKing has joined #openstack-swift05:05
*** SkyRocknRoll has joined #openstack-swift05:20
*** SkyRocknRoll has joined #openstack-swift05:20
*** lnxnut has joined #openstack-swift05:50
*** BaconIsKing has quit IRC05:58
*** nellysmitt has joined #openstack-swift06:18
*** ppai has quit IRC06:18
*** nellysmitt has quit IRC06:22
*** pcaruana has quit IRC06:28
*** ppai has joined #openstack-swift06:30
*** lnxnut has quit IRC06:41
*** rmcall has quit IRC07:22
*** chlong has quit IRC07:37
*** ppai has quit IRC07:40
*** ppai has joined #openstack-swift07:52
*** rledisez has joined #openstack-swift08:07
*** egon has quit IRC08:13
*** egon has joined #openstack-swift08:14
*** nellysmitt has joined #openstack-swift08:18
*** nellysmitt has quit IRC08:23
*** nellysmitt has joined #openstack-swift08:23
openstackgerritTushar Gohad proposed openstack/swift: EC: Multi-phase commit conversation for erasure-coded PUTs  https://review.openstack.org/15384308:28
*** geaaru has joined #openstack-swift08:34
*** aix has joined #openstack-swift08:40
*** joeljwright1 has joined #openstack-swift08:40
*** oomichi has quit IRC08:42
*** joeljwright has quit IRC08:42
*** mmcardle has joined #openstack-swift08:43
*** kbee has joined #openstack-swift08:48
*** jistr has joined #openstack-swift08:49
*** jordanP has joined #openstack-swift08:54
*** kbee has quit IRC08:55
*** silor has quit IRC09:19
*** cooldharma06 has joined #openstack-swift09:19
cooldharma06hi all09:19
*** aix has quit IRC09:23
*** ppai has quit IRC09:38
*** ppai has joined #openstack-swift09:51
*** jd__ has quit IRC10:06
*** jd__ has joined #openstack-swift10:07
*** cooldharma06 has left #openstack-swift10:17
*** ppai has quit IRC10:18
*** ppai has joined #openstack-swift10:31
*** erlon has joined #openstack-swift10:53
*** remix_tj has quit IRC10:55
*** remix_tj has joined #openstack-swift10:55
*** ppai has quit IRC11:06
*** ppai has joined #openstack-swift11:20
*** jamielennox is now known as jamielennox|away11:56
*** chlong has joined #openstack-swift12:25
*** ppai has quit IRC12:26
*** aix has joined #openstack-swift12:29
*** chlong has quit IRC12:35
*** chlong has joined #openstack-swift12:35
*** ppai has joined #openstack-swift12:39
*** ChanServ sets mode: +v chmouel12:42
*** ppai has quit IRC12:45
*** lnxnut has joined #openstack-swift12:53
*** lnxnut has quit IRC12:58
*** ppai has joined #openstack-swift12:58
*** ppai has quit IRC13:27
*** chlong has quit IRC13:36
*** lcurtis has joined #openstack-swift13:36
*** lcurtis has quit IRC13:45
*** aix has quit IRC13:47
*** lpabon has joined #openstack-swift13:49
*** lpabon has quit IRC13:58
*** david-lyle_afk is now known as david-lyle14:01
*** SkyRocknRoll has quit IRC14:11
*** dencaval has joined #openstack-swift14:14
*** lnxnut has joined #openstack-swift14:17
*** nellysmitt has quit IRC14:19
*** lnxnut has quit IRC14:21
*** BaconIsKing has joined #openstack-swift14:23
*** lcurtis has joined #openstack-swift14:23
*** mahatic has joined #openstack-swift14:24
*** BaconIsKing has quit IRC14:28
lcurtishello all...wondering if there is a general rule of thumb for sizing container db14:30
lcurtislooking like we will need to move it to SSD drives14:30
*** abhirc has joined #openstack-swift14:38
*** reed has joined #openstack-swift14:42
*** sandywalsh has quit IRC14:44
*** sandywalsh has joined #openstack-swift14:45
*** dmsimard_away is now known as dmsimard14:53
*** aix has joined #openstack-swift14:56
*** lnxnut has joined #openstack-swift14:58
*** bkopilov has quit IRC14:58
*** lnxnut_ has joined #openstack-swift15:01
*** lnxnut has quit IRC15:02
notmynamelcurtis: it depends on your use case. as a general guide, I normally tell people to start with 10% of their object capacity available for account+container. so if you have 100TB for objects, have 10TB of flash for account+container15:06
notmynamelcurtis: that's a rough guide. it will probably work fine, but it might be very over or under provisioned based on what your usage patterns are15:06
*** silor has joined #openstack-swift15:08
*** rledisez has quit IRC15:16
*** bsdkurt has quit IRC15:22
lcurtisnotmyname thank you!15:28
*** annegentle has joined #openstack-swift15:36
*** bill_az_ has joined #openstack-swift15:51
*** donagh- has joined #openstack-swift15:54
*** aix has quit IRC16:00
*** BaconIsKing has joined #openstack-swift16:15
*** aix has joined #openstack-swift16:29
*** gyee has joined #openstack-swift16:33
*** rmcall has joined #openstack-swift16:35
*** rmcall has quit IRC16:39
*** rmcall has joined #openstack-swift16:40
*** IRTermite has quit IRC16:42
openstackgerritJoel Wright proposed openstack/python-swiftclient: Fix crash when stat'ing objects with non-ascii names  https://review.openstack.org/14784616:47
*** bkopilov has joined #openstack-swift16:48
*** IRTermite has joined #openstack-swift16:56
*** lcurtis has quit IRC16:58
*** cca has joined #openstack-swift17:01
*** Trixboxer has quit IRC17:03
*** jrichli has joined #openstack-swift17:03
*** annegentle has quit IRC17:03
openstackgerritJoel Wright proposed openstack/python-swiftclient: Reduce memory usage for download/delete and add --no-shuffle option to st_download  https://review.openstack.org/13033917:07
*** cca has quit IRC17:08
notmynamegood morning, world17:08
*** cca has joined #openstack-swift17:08
*** jistr has quit IRC17:10
*** EmilienM is now known as EmilienM|afk17:15
*** mahatic has quit IRC17:17
*** mmcardle has quit IRC17:17
*** rook][ has joined #openstack-swift17:20
*** Nadeem has joined #openstack-swift17:26
*** zigo has quit IRC17:29
*** cca has quit IRC17:30
portantenotmyname: morning, how is the water doing there?17:36
*** mahatic has joined #openstack-swift17:36
claygmorning17:38
claygportante: we're fine, it's mostly let up... you mean the rain?17:38
portanteclayg: yes, the media is talking it up big time over here, perhaps to make us feel better some how17:38
*** geaaru has quit IRC17:41
claygportante: I thought we were supposed to be freaked out about measles...17:44
portanteclayg: that was yesterday, I thought, today the left coast is gonna drown, and snowmageddon has hit the right coast17:44
clayglol17:45
claygpeluse: how many devices are you sticking in your ec rings by default?17:50
*** geaaru has joined #openstack-swift17:55
*** dmorita has joined #openstack-swift18:00
*** annegentle has joined #openstack-swift18:04
openstackgerritAlistair Coles proposed openstack/swift: Make container GET call authorize when account not found  https://review.openstack.org/15416418:08
clayghurricanerix_: yeah we can just chat in here!18:14
*** jrichli has quit IRC18:14
*** tdasilva has joined #openstack-swift18:14
*** jrichli has joined #openstack-swift18:14
hurricanerix_clayg: sounds good, thanks.18:14
*** aix has quit IRC18:15
hurricanerix_notmyname: Here is that review I mentioned earlier: https://review.openstack.org/#/c/150149/18:15
*** mahatic has quit IRC18:19
notmynamehurricanerix_: yup.18:25
*** mahatic has joined #openstack-swift18:25
*** silor has quit IRC18:27
*** lpabon has joined #openstack-swift18:28
openstackgerritOpenStack Proposal Bot proposed openstack/python-swiftclient: Updated from global requirements  https://review.openstack.org/8925018:30
*** silor has joined #openstack-swift18:34
*** annegentle has quit IRC18:36
*** klrmn has joined #openstack-swift18:37
klrmngit review -s is telling me "We don't know where your gerrit is. Please manually create18:37
klrmna remote named "gerrit" and try again."18:37
*** _2_Rose has joined #openstack-swift18:40
*** _2_Rose has quit IRC18:41
notmynameklrmn: hmm18:42
notmynameklrmn: newest version of git-review?18:43
klrmnnotmyname: oh, probably not18:43
klrmnnotmyname: my ssh key seems to be correct18:44
*** annegentle has joined #openstack-swift18:46
*** lcurtis has joined #openstack-swift18:46
notmynameklrmn: any luck?18:47
klrmnnotmyname: not yet. was just about to copy-paste the remote from python-swiftclient, which i at least have used more often =)18:48
klrmnnotmyname: ok, got it working18:50
notmynameklrmn: great18:51
lcurtisnotmyname: if my ssd fills up on an account/container node, would swift write to another account/container node?18:51
lcurtisif i added several18:51
notmynamelcurtis: yes. if swift can't write to a particular drive, it will chose a handoff node. but note that swift fills available drives evenly* instead of "fill one then move to the next drive"18:52
*** lpabon has quit IRC18:52
notmyname*evenly. this is close enough, but not technically precise18:52
*** geaaru has quit IRC18:52
notmynamelcurtis: but the basic answer is, yes you should have multiple drives and swift will use what's available18:54
*** lpabon has joined #openstack-swift18:54
lcurtisnotmyname the sqlite db on 2 nodes would each have half the object list?18:55
lcurtiscontainer nodes for example18:55
*** mmcardle has joined #openstack-swift19:00
notmynamelcurtis: no. the containers (implemented on-disk as sqlite DBs) are replicated wholly across all replicas. ie 3 times for 3 replicas, and those 3 replicas are placed in the cluster based on the ring19:02
*** zigo has joined #openstack-swift19:03
lcurtisokay...understood19:04
notmynamezigo: this weekend I saw some emails about pyeclib being updated on pypi to publish the version without any net connections in the setup.py19:04
lcurtisthanks much for clarification19:04
notmynamelcurtis: sure. any time19:04
zigonotmyname: Well, I have already uploaded it, and it's waiting in the NEW queue.19:05
zigonotmyname: It's been 2 months already...19:05
zigoNothing more that I can do.19:05
zigohttps://ftp-master.debian.org/new.html19:05
notmynamezigo: ok. thansk19:05
lcurtisis there a way to map which physical server contains which account?19:07
notmynamelcurtis: `swift-get-nodes` will tell you that19:08
*** lpabon has quit IRC19:08
lcurtisperfect thank you19:08
*** EmilienM|afk is now known as EmilienM19:12
*** lpabon has joined #openstack-swift19:16
*** silor1 has joined #openstack-swift19:21
*** silor has quit IRC19:23
*** jordanP has quit IRC19:32
*** sileht has quit IRC19:33
notmynamehttps://github.com/matthewoliver/swift/compare/sharding_trie19:36
*** sileht has joined #openstack-swift19:41
*** openstackgerrit has quit IRC19:41
*** openstackgerrit has joined #openstack-swift19:41
*** bkopilov has quit IRC19:49
*** abhirc has quit IRC19:52
*** annegentle has quit IRC20:03
*** dencaval has quit IRC20:04
*** abhirc has joined #openstack-swift20:05
*** abhirc has quit IRC20:07
*** gyee has quit IRC20:07
*** lpabon has quit IRC20:10
*** silor1 has quit IRC20:19
*** lpabon has joined #openstack-swift20:25
*** Guest46399 has joined #openstack-swift20:37
*** mmcardle has quit IRC20:52
*** gyee has joined #openstack-swift21:14
*** yuanz has quit IRC21:35
*** yuanz has joined #openstack-swift21:36
*** BaconIsKing has quit IRC21:55
*** rook][ has quit IRC21:56
*** oomichi has joined #openstack-swift22:06
*** zaitcev has joined #openstack-swift22:09
*** ChanServ sets mode: +v zaitcev22:09
*** abhirc has joined #openstack-swift22:15
openstackgerritSamuel Merritt proposed openstack/swift: Make GetOrHeadHandler take an optional chunk size.  https://review.openstack.org/15426222:19
openstackgerritSamuel Merritt proposed openstack/swift: First pass at EC GET path  https://review.openstack.org/15426322:19
hurricanerix_I was telling notmyname about this, but I figured I would just share it here and see what people think about it.  I started thinking about modifying staticweb to make hosting single page web applications a little cleaner.  For example using "html5 mode" in AngularJS.  Anyway, I was playing with this, but after doing so I started second guessing if it was really worth anything.  So if you like the idea of it, let me know ;)  https://gith22:39
hurricanerix_ub.com/hurricanerix/swift/compare/staticweb_spa22:39
notmynamehurricanerix_: it got split22:40
notmynamethe link22:40
hurricanerix_whoops22:40
hurricanerix_https://github.com/hurricanerix/swift/compare/staticweb_spa22:40
notmynamemordred: can you help me in here for a moment?22:40
mordrednotmyname: yes!22:45
*** dmorita has quit IRC22:45
*** dmorita has joined #openstack-swift22:46
notmynamemordred: thanks. I asked in -infra. I need another feature branch in swift. I don't remember how that works, other than "you" have to set it up22:46
mordrednotmyname: yah - it's pretty easy/straightforward - let me go poke people22:47
*** rmcall has quit IRC22:49
notmynamemordred: looks like someone with the proper permissions needs to push a branch. but I don't have those permissions22:53
notmynamemordred: the branch we want is "feature/crypto"22:53
notmynamejeblair: ^22:54
mordrednotmyname: do you want it based on current master?22:55
jeblairnotmyname, mordred: can we run these through ttx?22:55
mordredjeblair: sure!22:56
jeblairnotmyname, mordred: mostly because in openstack, they are effectively global22:56
mordrednod22:56
jeblairthat is, if swift has a feature/crypto branch, it will get tested with keystone feature/crypto branch changes22:56
notmynameah, interesting22:56
jeblairand if they aren't working toward the same goal, that would be counterproductive (but if they are, then it would be very useful)22:56
jeblairso basically just to avoid collisions22:57
*** lpabon has quit IRC22:57
notmynamedo you know if there is crypto, encryption, etc22:57
notmynamethis is for on-disk encryption in swift. keystone is unrelated (but, interestingly, barbican could be)22:57
jeblairnotmyname: i don't know, and i don't know of a good way to check, other than iterate over projects with the api...22:57
jeblairnotmyname: that's why i'm thinking that using ttx as a human registry would be good, or else, maybe come up with something like registering feature branches on a wiki page22:58
jeblair(of course, they are easy to change)22:59
notmynameso you need me to coordinate with him?22:59
jeblairnotmyname: if it's not too much trouble, i think that'd be great.  if this is a blocker and needs to happen now, we can go ahead and do it and fix later if needed23:00
*** reed has quit IRC23:00
notmynameI'd like to get it done today, if possible. we're on day 1 of our hackathon/midcycle, and this is something we're working on23:00
notmynamedo you have a gut instinct on the global use of crypto vs encryption? I'd guess crypto is more likely to already be used23:01
jeblairnotmyname: i think there are very few if any feature branches right now23:01
notmynamethen I'd prefer crypto :-)23:02
jeblairmordred: i remove my objection23:04
jeblairnotmyname: please let ttx know23:04
notmynamejeblair: and he'll do it? or let him know that this has already happened because you have done it? ;-)23:05
jeblairnotmyname: i have not done it.  i think mordred was going to before i objected.  i'm not going to steal the honor from him.  :)23:05
*** jamielennox|away is now known as jamielennox23:05
notmynameheh23:05
*** kota_ has joined #openstack-swift23:06
*** cppforlife_ has quit IRC23:07
mordredi got it23:07
mordrednotmyname: we're set on feature/crypto?23:08
notmynameyes23:08
mordrednotmyname: done23:08
notmynamethanks!23:08
notmynamegot it23:09
notmynameI'll let ttx know23:09
notmynameacoles: http://docs.openstack.org/infra/manual/drivers.html#feature-branches23:09
*** lpabon has joined #openstack-swift23:18
*** cppforlife_ has joined #openstack-swift23:22
*** cppforlife_ has quit IRC23:23
*** occupant has quit IRC23:23
*** occupant has joined #openstack-swift23:24
*** lpabon has quit IRC23:25
*** EmilienM is now known as EmilienM|afk23:28
notmynameacoles: also, peluse has a cheatsheet for doing the master->feature merge.23:33
*** Nadeem has quit IRC23:51
klrmnclayg: is "ERROR: test_sync (test.probe.test_container_sync.TestContainerSync)" indicative of the problem you were describing?23:52
openstackgerritRichard Hawkins proposed openstack/swift: Add support for container TempURL Keys  https://review.openstack.org/15429323:56
hurricanerix_cschwede: ^^^ work in progress, so we can share code/progress.23:58
*** dmsimard is now known as dmsimard_away23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!