Tuesday, 2011-01-11

*** adiantum has joined #openstack00:01
*** adiantum has quit IRC00:07
*** zul has quit IRC00:08
*** aimon has quit IRC00:17
*** aimon has joined #openstack00:17
*** adiantum has joined #openstack00:19
*** arthurc has quit IRC00:24
*** adiantum has quit IRC00:26
*** adiantum has joined #openstack00:31
*** dubsquared has quit IRC00:33
*** kashyapc has quit IRC00:34
*** adiantum has quit IRC00:39
*** dragondm has quit IRC00:41
*** adiantum has joined #openstack00:44
*** johnpur has quit IRC00:49
*** adiantum has quit IRC00:50
*** kashyapc has joined #openstack00:51
*** adiantum has joined #openstack00:55
*** jdarcy has quit IRC01:00
*** dfg_ has quit IRC01:04
*** rlucio has quit IRC01:05
nelson__Does swift has a "move this object over to this other container" function? I don't see it.01:06
anticwthere is copy01:07
notmynamenelson__: https://programmerthoughts.com/programming/server-side-object-copy-in-openstack-storage/01:09
nelson__well, that's not FAIL.01:11
*** dirakx has joined #openstack01:11
notmynamedid you expect it to be?01:12
nelson__We're gonna need that for uploading files that we don't want to publish immediately (because file upload happens before copyright permissions are retrieved).01:12
*** adiantum has quit IRC01:13
nelson__not having server-side copy would be annoying. I suppose it would just mean more network traffic and more proxies needed.01:13
anticwfor large files you have to pay for a copy01:13
anticwwhy not just not make them public?01:13
nelson__Because we ask for the file on one screen and then the copyright information in another.01:13
notmynamethe bandwidth is all internal, so, at least with RAX's implementation of swift, there would be no bandwidth charge01:13
anticwsame here, but 5G is still a chunk to copy01:14
anticwif it was that large01:14
notmynameya01:14
nelson__In-between those two steps we need to keep the file private.01:14
nelson__I know we can do that with swift ... I'm just trying to dig further into the details.01:14
notmynamehow busy are those upload forms in requests per second?01:15
nelson__20-30.01:15
notmynamerough estimate01:15
notmynameok01:15
notmynamewould you be using your own auth system or using a provided one (ie swauth)?01:17
*** adiantum has joined #openstack01:18
notmynameit may be better to use ACLs in your auth system to control the publicness01:18
nelson__yes, it might.01:19
nelson__we'll have to see how the existing code works.01:19
notmynameI want to talk to gholt tomorrow about this use case. we saw something last week (while I wasn't in the office) that may be similar01:20
*** mray has quit IRC01:22
*** schisamo has quit IRC01:23
nelson__Might be simplest just to adjust the ACL for that file.01:37
nelson__We don't actually NEED to change the container; just to make sure that it's not publicly readable.01:37
*** joearnold has quit IRC01:39
*** dirakx has quit IRC01:43
*** jdurgin has quit IRC01:48
nelson__Are the items returned by get_container sorted? I don't see that in the documentation for hte function.01:49
notmynamelistings are sorted by name (container or object for their respective listings)01:49
nelson__Oh good, that's a useful property.01:51
notmynamehttp://bazaar.launchpad.net/%7Ehudson-openstack/swift/trunk/annotate/head%3A/swift/common/db.py#L94401:51
notmynameespecially look at line 99101:52
notmyname(ORDER BY)01:52
notmynamethe account listings (to show container) is ordered in the same way (see later in the same file)01:56
nelson__yep; it should be documented as part of the API, though. Otherwise it's just "yeah, we put this into the code, but we might take it out."01:57
nelson__knowwhaddimean?01:57
notmynameheh, if we take out ordering of results, everything breaks (marker, end_marker, prefix, path, delimiter, ...)01:58
notmynamenelson__: the latest cloud files dev guide (http://docs.rackspacecloud.com/files/api/v1/cf-devguide-20110104.pdf) which will be used as the basis of official swift api docs, mentions it02:00
nelson__Oh cool. I'll take a look at that.02:00
notmynameignore the CDN parts, cause that's part of the cloud files product and not a part of swift02:01
*** maplebed has quit IRC02:03
*** adiantum has quit IRC02:08
*** adiantum has joined #openstack02:15
creihtnelson__: and btw, the acls are on the container only02:16
nelson__ah! Okay, we will need to move from one container to another, then.02:17
nelson__No biggie.02:17
creihtor upload to a staging server that then gets put into swift when copyright verified02:17
*** adiantum has quit IRC02:20
*** adiantum has joined #openstack02:25
*** rossij has joined #openstack02:36
*** skrusty has quit IRC02:45
*** hadrian has quit IRC02:48
gholtMaking a write-only container should work fine, I'd think. A staging server would just be something that'd break. ;)02:50
gholtThe auth implementations don't currently allow public writes; but it'd be easy enough to alter them to do that. If you're okay with people uploading who knows what and you getting charged.02:52
creihtheh02:53
*** skrusty has joined #openstack02:58
*** zykes- has quit IRC03:04
*** zykes- has joined #openstack03:04
*** adiantum has quit IRC03:06
*** adiantum has joined #openstack03:11
*** crazed has quit IRC03:11
*** zykes- has quit IRC03:17
*** zykes- has joined #openstack03:17
*** adiantum has quit IRC03:17
*** gaveen has joined #openstack03:26
*** gaveen has joined #openstack03:26
*** kainam- has joined #openstack03:26
*** adiantum has joined #openstack03:30
*** mark_ has joined #openstack03:30
*** _0x44_ has joined #openstack03:30
*** karmabot` has joined #openstack03:33
*** miclorb__ has joined #openstack03:34
jeremybgholt: well you're not charged if you have unmetered network and your own cluster (i think those would both be true for nelson__)03:34
*** jbaker_ has joined #openstack03:34
*** rossij has quit IRC03:34
*** kainam has quit IRC03:34
*** miclorb has quit IRC03:34
*** openstackhudson has quit IRC03:34
*** jimbaker has quit IRC03:34
*** mark has quit IRC03:34
*** karmabot has quit IRC03:34
*** _0x44 has quit IRC03:34
*** glenc has quit IRC03:34
*** jtimberman has quit IRC03:34
*** JordanRinke has quit IRC03:34
*** brainproxy has quit IRC03:34
*** _cerberus_ has quit IRC03:34
*** klumpie has quit IRC03:34
*** cory__ has quit IRC03:34
*** smoser has quit IRC03:34
*** Dweezahr has quit IRC03:34
*** letterj has quit IRC03:34
*** sunech has quit IRC03:34
jeremybhuh, never realized ma rk was here before03:35
creihtwoot03:36
jeremybhe would certainly know!03:36
creihtswauth is now merged into swift trunk03:36
creihtcolinnich: -^03:36
jeremybthen what is lp.net/~swauth ?03:36
jeremybdead?03:36
creihtjeremyb: yeah03:36
creihtswauth is now middleware that sits on top of the cluster03:37
jeremybcool. should it be production ready you think? (at least as of release time)03:37
creihton top of the proxy03:37
*** glenc has joined #openstack03:37
*** jtimberman has joined #openstack03:37
*** JordanRinke has joined #openstack03:37
*** Dweezahr has joined #openstack03:37
*** brainproxy has joined #openstack03:37
*** _cerberus_ has joined #openstack03:37
*** klumpie has joined #openstack03:37
*** cory__ has joined #openstack03:37
*** smoser has joined #openstack03:37
*** letterj has joined #openstack03:37
*** sunech has joined #openstack03:37
*** zelazny.freenode.net sets mode: +vv _cerberus_ letterj03:37
creihtI hope so, as we are going to use it for our internal auth :)03:37
creihtIt is for sure better than the dev auth03:37
jeremyborly? for rax you mean?03:37
creihtyes03:37
*** lvaughn_ has quit IRC03:38
jeremybhuh. can it be used for nova too?03:38
creihtinternal fir cloudfiles03:38
*** adiantum has quit IRC03:38
creihtjeremyb: as is, I doubt it03:38
jeremybdoes nova come with an auth server or require one?03:38
*** rossij has joined #openstack03:38
creihtnot sure what would be needed to make it work for nova (or if that is even a good idea :))03:38
* jeremyb wonders who the main swauth ppl are03:38
creihtjeremyb: I think the default is ldap, but it is pluggable03:39
creihtjeremyb: gholt wrote most of it03:39
jeremybahh. can swauth use ldap too?03:39
creihtno03:39
jeremyb:/03:39
creihtit just does the auth that is needed for swift, and backends it with swift itself03:39
jeremybhuh, interesting03:40
gholtYeah, ldap backing would be ldauth or something. And does need to be written sometime.03:40
jeremybwell i have to get cracking, bbl03:40
jeremybnelson__: 11 03:36:17 <+creiht> swauth is now merged into swift trunk03:41
creihtjeremyb: I think xtoddx modified nova auth so that it can also be auth for swift03:41
jeremybalso interesting03:41
jeremybi wonder what Ryan_Lane's doing03:42
jeremybnova natively supports ldap or he hacked it?03:42
creihtnova's default auth is backed by ldap03:42
creihtHe added the needed keys to store swift like auth tokens (or something like that)03:43
jeremybright03:43
creihtI haven't really done anything with nova yet, so it would probably be best to try to ask him how it works :)03:44
jeremybi don't really care enough atm03:44
jeremybbut thanks03:44
*** adiantum has joined #openstack03:45
*** openstackhudson has joined #openstack03:45
creiht:)03:45
jarrodwhere can i get some more good systems for cloud servers for a good price03:53
*** zykes- has quit IRC03:54
jarrodive been using dell03:54
jarrodbut they get $$$03:54
* jeremyb </3 dell03:55
*** zykes- has joined #openstack03:55
jeremybi just heard today they sometimes sell a system with a given perc model and then a month later ship the same perc model but it's not really the same. (they don't make their own, sometimes adaptec, sometimes lsi)03:55
jeremyb(perc is raid controller)03:56
jarrodi was replacing with 3ware03:56
jarrodbut i believe that03:57
*** adiantum has quit IRC04:02
*** zykes- has quit IRC04:02
*** MarkAtwood has quit IRC04:07
*** adiantum has joined #openstack04:07
*** zykes- has joined #openstack04:09
*** rlucio has joined #openstack04:13
*** joearnold has joined #openstack04:19
*** kashyapc has quit IRC04:23
*** adiantum has quit IRC04:39
*** MarkAtwood has joined #openstack04:40
*** adiantum has joined #openstack04:44
*** kashyapc has joined #openstack04:45
*** troytoman has joined #openstack04:53
*** rlucio has quit IRC05:00
*** zykes- has quit IRC05:00
*** zykes- has joined #openstack05:01
*** adiantum has quit IRC05:09
*** adiantum has joined #openstack05:15
*** joearnold has quit IRC05:20
*** ccustine has joined #openstack05:22
*** pandemicsyn has joined #openstack05:26
*** ChanServ sets mode: +v pandemicsyn05:26
*** pandemicsyn has quit IRC05:28
*** ccustine has quit IRC05:28
*** adiantum has quit IRC05:37
*** syn_ has joined #openstack05:37
*** syn_ is now known as pandemicsyn05:38
*** ChanServ sets mode: +v pandemicsyn05:38
*** adiantum has joined #openstack05:39
openstackhudsonProject nova build #380: SUCCESS in 1 min 24 sec: http://hudson.openstack.org/job/nova/380/05:44
openstackhudsonTarmac: Adds support to nova-manage to modify projects05:44
XenithI usually pick up Supermicro systems. Very good hardware and support.05:47
openstackhudsonProject nova build #381: SUCCESS in 1 min 20 sec: http://hudson.openstack.org/job/nova/381/05:49
openstackhudsonTarmac: Fixes the metadata forwarding to work by default.05:49
openstackhudsonThis branch also moves get_my_ip into flags.py from utils.py and sets up a flag called my_ip.  Any flag that was using get_my_ip can just use '$my_ip' as the default value.  This allows a user to change all ips by changing one flag.05:49
*** f4m8_ is now known as f4m805:49
*** adiantum has quit IRC05:51
*** mray has joined #openstack05:55
*** adiantum has joined #openstack05:58
*** medmaster0 has joined #openstack05:59
medmaster0hey guys05:59
medmaster0i got a question can someone help?05:59
*** trin_cz has quit IRC06:00
*** adiantum has quit IRC06:05
*** adiantum has joined #openstack06:11
*** medmaster0 has quit IRC06:14
openstackhudsonProject nova build #382: SUCCESS in 1 min 22 sec: http://hudson.openstack.org/job/nova/382/06:19
openstackhudsonTarmac: A few more changes to the smoeketests.  Allows smoketests to find the nova package from the checkout.  Adds smoketests for security groups.  Also fixes a couple of typos.06:19
*** ramkrsna has joined #openstack06:26
*** MarkAtwood has quit IRC06:29
*** ramkrsna has quit IRC07:07
*** _0x44_ is now known as _0x4407:14
*** ChanServ sets mode: +v _0x4407:14
_0x44jaypipes-afk: Is there any chance we can get swift installed as a package on the glance hudson?07:19
*** ramkrsna has joined #openstack07:20
*** befreax has joined #openstack07:23
*** aimon_ has joined #openstack07:33
*** aimon has quit IRC07:35
*** aimon_ is now known as aimon07:35
*** adiantum has quit IRC07:36
*** adiantum has joined #openstack07:41
*** littleidea has joined #openstack07:52
*** brd_from_italy has joined #openstack07:57
*** rcc has joined #openstack07:59
*** adiantum has quit IRC08:07
*** calavera has joined #openstack08:17
*** adiantum has joined #openstack08:19
*** rcc has quit IRC08:19
*** adiantum has quit IRC08:33
*** ibarrera has joined #openstack08:37
*** miclorb__ has quit IRC08:38
*** damon__ has quit IRC08:42
*** fabiand_ has joined #openstack08:46
*** DigitalFlux has joined #openstack08:52
*** adiantum has joined #openstack08:53
DigitalFluxHi08:54
DigitalFluxI see that someone removed the sections "Setup a Cloud Controller" and "setup a Compute controller" from the MultiServer" installation wiki page ..08:54
DigitalFluxhttp://wiki.openstack.org/NovaInstall/MultipleServer08:54
*** adiantum has quit IRC09:11
*** kashyapc has quit IRC09:14
*** adiantum has joined #openstack09:24
*** miclorb has joined #openstack09:25
*** allsystemsarego has joined #openstack09:41
*** allsystemsarego has joined #openstack09:41
*** dizz|away has quit IRC09:45
*** kashyapc has joined #openstack09:45
*** adiantum has quit IRC09:51
*** miclorb has quit IRC10:09
*** omidhdl has joined #openstack10:09
*** irahgel has joined #openstack10:12
*** pcoca has joined #openstack10:14
*** kashyapc has quit IRC10:17
*** kashyapc has joined #openstack10:18
sorenttx: How are we doing on reviews?10:18
ttxsoren:  let me run the script10:19
ttx$ python brstats.py10:19
ttxglance: 1 / 5 completed (should be 3 by eod)10:19
ttxswift: 4 / 8 completed (should be 4 by eod)10:19
ttxnova: 3 / 22 completed (should be 13 by eod)10:19
ttxso swift is looking good, glance is a bit behind, nova is worryingly behind10:20
ttxAt this point I'm mostly concerned by the branches that didn't get reviewed even once10:21
ttxEspecially https://code.launchpad.net/~ntt-pf-lab/nova/ipv6-support/+merge/4522810:21
ttxsoren: could you review that one today ?10:22
sorenThe reason I haven't is that ipv6 is a bit of a mystery to me.10:23
sorenSo I'll have difficulties reviewing that actually interesting bits.10:23
* soren carefully glances at it.10:24
sorenOh, ffs. It's about as long as the live migration patch. That was a killer.10:24
ttxthen we need to find someone in nova-core for which it isn't a mystery10:24
ttxsoren: yes, hence me worrying. it's not exactly easy to review.10:25
ttxthe sheepdog thing is very simple in comparison :)10:25
sorenThat one kind of surprised me.10:26
sorenI thought they'd use sheepdog for instance storage, not just EBS volumes10:26
sorenTo easy live migration.10:26
sorens/easy/ease/10:26
sorenmeh10:26
*** kashyapc has quit IRC10:36
*** sandywalsh has quit IRC10:49
*** sandywalsh has joined #openstack11:03
sorenphew.11:06
fraggelnsoren: you fixed it? :D11:09
*** arthurc has joined #openstack11:11
*** omidhdl1 has joined #openstack11:12
*** omidhdl has quit IRC11:12
sorenfraggeln: I finished a review.11:13
sorenfraggeln: I fixed something right before then, though :)11:13
*** adiantum has joined #openstack11:13
sorenttx: Hey, how about a SupyBot plugin that notifies us when there's a new bmp?11:14
sorenttx: #twisted has a bot that posts the list of pending reviews every 5 hours.11:15
XenithThat is one big patch.11:15
ttxsoren: that would be a bit noisy. Also would fail to prioritize11:15
XenithGood functionality, though :)11:15
ttxsoren: I see some nagging value in it though :)11:15
sorenttx: it's all about the nagging.11:16
* soren nags himself to go to lunch11:16
* soren succeeds and wanders off11:16
*** brd_from_italy has quit IRC11:20
fraggelnsoren: lunch is the shit :)11:20
fraggelnsoren: btw, you are from the nordic region yes?11:21
*** rcc has joined #openstack11:30
*** brd_from_italy has joined #openstack11:36
*** littleidea has quit IRC11:36
*** littleidea has joined #openstack11:36
*** guigui1 has joined #openstack11:39
*** littleidea has quit IRC11:41
*** sandywalsh has quit IRC11:49
*** omidhdl1 has quit IRC11:49
*** sandywalsh has joined #openstack11:49
sandywalsho/11:52
ttxsandywalsh: o/12:00
*** arthurc has left #openstack12:01
*** befreax has quit IRC12:03
sorenfraggeln: I am indeed. Denmark, specifically.12:09
fraggelnMy first guess was sweden, but its the same shit anyway :D12:14
ttxahh, don't say that to him !12:16
fraggeln:/12:18
fraggelngreat work btw, we are planning on implementing swift as a backend storage for our backup-software.12:19
fraggelnits looking very good so far.12:19
ttxGot a spam selling Cocoa beans by the ton. That sounds rather specialized.12:21
ttxPAYMENT:                 NEGOTIABLE12:22
*** befreax has joined #openstack12:24
openstackhudsonProject nova build #383: SUCCESS in 1 min 24 sec: http://hudson.openstack.org/job/nova/383/12:24
openstackhudsonTarmac: Implementation of xs-console blueprint (adds support for console proxies like xvp)12:24
openstackhudsonIf you spin up the nova-console service, you should be able to see the xvp.conf being edited, and the xvp daemon started/stopped if you exercise the openstack console api (consoles sub-resource on servers)12:24
* soren likes how Tarmac now includes output there.12:24
ttxyes, I love that12:25
*** ctennis has quit IRC12:27
*** colinnich has quit IRC12:28
*** trin_cz has joined #openstack12:28
*** ctennis has joined #openstack12:50
*** irahgel has quit IRC12:52
*** brd_from_italy1 has joined #openstack12:56
*** brd_from_italy has quit IRC12:56
*** irahgel has joined #openstack13:04
*** rcc has quit IRC13:14
*** MarkAtwood has joined #openstack13:30
*** ramkrsna has quit IRC13:32
*** trin_cz has quit IRC13:46
*** uvirtbot has quit IRC14:01
*** uvirtbot has joined #openstack14:02
*** kashyapc has joined #openstack14:06
*** rcc has joined #openstack14:09
*** trin_cz has joined #openstack14:15
*** lvaughn has joined #openstack14:15
*** ppetraki has joined #openstack14:18
*** openstackhudson has quit IRC14:20
sorenopenstackhudson: See you in a bit.14:20
*** openstackhudson has joined #openstack14:20
sorenDarn it.14:21
*** BK_man has joined #openstack14:22
sorenmtaylor: Monty!14:22
*** skrusty has quit IRC14:26
*** irahgel1 has joined #openstack14:33
*** irahgel has quit IRC14:35
*** skrusty has joined #openstack14:40
jt_zgMorning @ everyone awake at this ungodly hour!14:42
*** troytoman has quit IRC14:42
notmynamejt_zg: heh, I've got kids that have been up for 3 hours :-)14:44
jt_zgnotmyname, I think they call that cruel and unusual punishment here.14:44
*** littleidea has joined #openstack14:46
*** f4m8 is now known as f4m8_14:48
*** gaveen has quit IRC14:51
*** MarkAtwood has quit IRC14:53
*** abecc has joined #openstack14:58
*** littleidea has quit IRC14:58
*** littleidea has joined #openstack14:58
*** gondoi has joined #openstack15:02
*** littleidea has quit IRC15:03
*** westmaas has joined #openstack15:04
*** rnirmal has joined #openstack15:08
*** reldan has joined #openstack15:11
*** hadrian has joined #openstack15:15
*** allsystemsarego has quit IRC15:16
*** jdarcy has joined #openstack15:16
*** ilya_ has joined #openstack15:20
*** lorinmh has joined #openstack15:21
*** ilya_ has quit IRC15:22
*** adiantum_ has joined #openstack15:23
*** joearnold has joined #openstack15:29
*** jdarcy has quit IRC15:40
*** joearnold has quit IRC15:41
*** littleidea has joined #openstack15:42
*** johnpur has joined #openstack15:46
*** ChanServ sets mode: +v johnpur15:46
*** rcc has quit IRC15:47
*** jdarcy has joined #openstack15:48
*** DigitalFlux has quit IRC15:53
*** guigui1 has quit IRC15:57
*** rcc has joined #openstack15:58
*** lvaughn has quit IRC16:04
*** dragondm has joined #openstack16:05
tr3buchetfeedback welcome: https://blueprints.launchpad.net/nova/+spec/instance-state-arbiter16:12
*** zul has joined #openstack16:15
sandywalshI see you looking at the bug list16:24
sandywalshI see you reviewing blueprints16:25
sandywalshyou should be reviewing this merge-prop16:25
sandywalshhttps://code.launchpad.net/~sandy-walsh/nova/api-parity/+merge/4537116:25
sandywalsh:)16:25
*** calavera has quit IRC16:27
*** hggdh has joined #openstack16:27
*** Cybo has joined #openstack16:32
*** kashyapc has quit IRC16:34
*** dubsquared1 has joined #openstack16:36
*** dfg_ has joined #openstack16:39
*** jaypipes-afk is now known as jaypipes16:42
*** lvaughn has joined #openstack16:44
tr3bucheti already reviewed that sandy!16:44
*** reldan has quit IRC16:47
*** grizzletooth has joined #openstack16:47
*** grizzletooth has left #openstack16:47
sandywalshhaha, wasn't talking to you specifically tr3buchet ... your place in heaven is already secured.16:48
Ryan_Lanejeremyb: I haven't done any LDAP work in swift ;)16:49
Ryan_Lanejeremyb: it would be really unlikely for us to use LDAP for that, so I haven't bothered16:50
jeremybright16:50
jeremybi thought maybe ldap in nova was you16:50
Ryan_LaneI've rewritten a bunch of it16:50
jeremybbut it sounds like that's the main way so it would have predated you16:50
jeremybahh16:50
Ryan_Lanemodified the schema, etc16:50
* jeremyb has to work quick so he can go meet ragesoss and friends16:51
jeremybhave you met him?16:51
Ryan_Laneheh16:51
Ryan_Laneyeah. at foundation officies16:51
jeremybk16:51
Ryan_Laneoffices*16:51
*** reldan has joined #openstack16:54
tr3buchetsweet!16:54
*** Whoop has joined #openstack16:55
* JordanRinke looks around16:57
*** kashyapc has joined #openstack16:58
*** brd_from_italy1 has quit IRC17:00
*** gondoi has quit IRC17:01
*** reldan has quit IRC17:05
dabogeneral question on async call problems: if we're not waiting around for the async call to return, should the call raise exceptions? IOW, there's nothing around to handle them. Would it be better to simply catch the exception and record it in the log?17:05
jeremybanyone know George Vanecek? <http://www.meetup.com/OpenStack-New-York-Meetup/calendar/15634525/>17:08
sandywalshdabo, I think so long as it's caught high enough up, it will permit us to do things like pubsubhubbub/http callbacks or just log it for the short term.17:09
dabosandywalsh: the API calls return immediately, so there is nothing waiting for the response from the async call. In the case of password reset, the timeout is 30 seconds, so the TimeoutError doesn't happen until then.17:10
daboI see tracebacks in the compute log for these; I think simply catching the timeout and logging it would be cleaner.17:11
BK_manhi all!17:12
sandywalshdabo, yeah, good point. There's no cohesive plan for notifications on completion (be it a callback/email/etc). Perhaps others have plans for that? New bp?17:12
BK_mangents, who is managing nova.openstack.org/tarballs/ ?17:12
sandywalshdabo, actually, come to think of it. It should just write to a notification queue17:12
dabosandywalsh: does sprinkle's proposed blueprint address this? https://blueprints.launchpad.net/nova/+spec/instance-state-arbiter/17:12
BK_manin trunk I can see bin/nova-combined but there is no bin/nova-combined in latest available tarball (bzr543).17:13
sandywalshdabo, no, different animal17:14
dabosandywalsh: really? Your idea of a notification queue would seem to tie in nicely17:15
sandywalshdabo, his proposal could be used when polling is required since it's kind of a cron-job service. But when we're actively waiting for results and have no where to report them ...17:17
*** opengeard has quit IRC17:18
dabosandywalsh: exactly: if we had somewhere to report them (e.g., a notification queue), his proposed system could listen for these, and update things accordingly.17:19
jarrodwhat are some good cost servers to use for cloud servers?17:22
sandywalshdabo, yeah, so long as the event gets written to the queue, tr3buchet's service or anything else can listen for it.17:24
*** gondoi has joined #openstack17:25
*** lvaughn has quit IRC17:31
*** lvaughn has joined #openstack17:32
*** littleidea has quit IRC17:34
JordanRinkejarrod: depends what you want to do, your buying capacity etc.17:34
*** mray has joined #openstack17:37
jarrodi know that17:37
jarrodim asking for any other cost sensitive solutiosn than major brands like dell17:37
*** ibarrera has quit IRC17:38
*** ccustine has joined #openstack17:42
critchjarrod: http://www.pricewatch.com/servers/17:42
fabiand_jarrod: In europe?17:43
*** pcoca has quit IRC17:44
*** troytoman has joined #openstack17:50
*** jdurgin has joined #openstack17:50
*** hadrian has quit IRC17:50
*** joearnold has joined #openstack17:50
*** reldan has joined #openstack17:50
*** hadrian has joined #openstack18:00
*** dendrobates is now known as dendro-afk18:07
*** maplebed has joined #openstack18:08
*** dendro-afk is now known as dendrobates18:09
*** rlucio has joined #openstack18:09
uvirtbotNew bug: #701575 in nova "install_venv.py does not install virtualenv if missing" [Undecided,New] https://launchpad.net/bugs/70157518:11
*** ccustine has quit IRC18:13
*** hggdh has quit IRC18:13
*** ccustine has joined #openstack18:15
*** alekibango has quit IRC18:16
*** rcc has quit IRC18:16
jarrodis there a maximum amount of memory that openstack supports?18:21
jarrodim currently only using 16gb18:21
*** Ryan_Lane is now known as Ryan_Lane|food18:24
*** Ryan_Lane|food has quit IRC18:26
*** westmaas has quit IRC18:27
*** befreax has quit IRC18:28
sirp-dubs, jk0: these are the errors im seeing in nova/trunk: http://paste.openstack.org/show/464/18:30
jk0sirp-: I just got the same thing18:31
jk0I'll try on my mac18:31
dubssame here18:31
* jk0 waits for venv18:34
sirp-jk0: btw, im running with out venv for the moment, had good success in the past with tests passing without the hassle of setting up the venv18:35
jk0yeah, I don't use venv either18:36
jk0only when running tests on OS X18:36
jk0just want to see if it works outside of our env18:36
*** ccustine has quit IRC18:38
jk0sirp-: tests fail in fresh OS X trunk pull18:43
jk0oh, nm, different issue18:43
jk0openssl18:43
jk0I'll try in a fresh ubuntu 10.10 slice18:44
sirp-jk0: cool, thx18:44
lorinmhI just posted my first bug report to OpenStack (#701575), as well as a potential bugfix branch. Is there anything else I need to do in order to get the proposed fix reviewed? Wasn't sure if I need to propose the branch for a merge or if it will get checked when somebody on the team takes a look at the bug.18:47
notmynamelorinmh: you should propose it for merging. if nothing else, it gives a central place to comment on the patch18:49
lorinmhnotmyname: Will do.18:49
*** ccustine has joined #openstack18:50
dabojk0: sirp- I'm seeing the same errors on OS X, no venv18:54
jk0sirp-: all tests pass on a fresh ubuntu 10.10 slice with trunk18:57
*** lorinmh1 has joined #openstack18:57
*** lorinmh1 has left #openstack18:57
sirp-jk0: without venv?18:57
jk0yep, w/o18:57
*** lorinmh1 has joined #openstack18:57
jk0I'll try it with too18:57
*** lorinmh1 has left #openstack18:58
dabojk0: sirp- Same fail on my 10.10 system in the castle lab18:58
*** mray has quit IRC18:59
*** colinnich has joined #openstack19:00
*** lorinmh has quit IRC19:00
ttxTeam meeting in two hours in #openstack-meeting19:02
*** mray has joined #openstack19:02
*** Ryan_Lane has joined #openstack19:04
*** lorinh1 has joined #openstack19:05
jk0sirp-: tests *fail* using venv on fresh ubuntu 10.10 slice, pass not using venv19:09
* nelson__ causes trouble.19:09
*** sophiap has joined #openstack19:09
*** phymata has joined #openstack19:11
dabojk0: I'm not using venv on 10.10, and they're failing19:11
jk0dabo: you are trying it within our dev env, this is on a fresh slice19:11
jk0something weird is going on19:12
dabojk0: yes, but until the last pull it was working fine19:12
jk0right, that's what we're trying to figure out19:12
jk0what changed that it works on a fresh slice but not out dev env19:12
jk0*our19:13
*** kpepple has joined #openstack19:13
*** dirakx has joined #openstack19:14
*** trin_cz has quit IRC19:16
*** littleidea has joined #openstack19:17
*** BK_man has quit IRC19:25
*** irahgel1 has quit IRC19:29
*** adiantum has quit IRC19:30
sorenmtaylor: Ok, so the nova-tarball job was busted on Hudson.19:33
sorenmtaylor: http://hudson.openstack.org/job/nova-tarball/136/console19:34
mtaylorsoren: yup. I saw19:34
sorenmtaylor: Did you see my workaround?19:34
mtaylorsoren: no. what did you do?19:34
sorenmtaylor: nova-tarball was set to use the SCM settings from nova.19:34
mtaylorah. ok. cool19:34
sorenmtaylor: Instead, I copied the actual settings from nova.19:34
sorenThat was it.19:34
mtaylorcool. I think I can start poking at where to fix that for real then19:35
sorenNo clue why that worked. No clue why I tried.19:35
sirp-jk0: i re-ran the tests (with no changes!), and they just passed19:35
jk0hm, strange19:35
jk0I'll try again too19:35
jk0sirp-: they just passed for me too, wtf19:37
sorenmtaylor: Oh, and once again, I found a bunch of files in /usr/lib/python that were owned by hudson.19:37
mtaylorsoren: WEIRD19:38
sorenmtaylor: Yeah. Make it stop, please :)19:38
sorenmtaylor: Does anything still use virtualenv19:38
soren?19:38
mtaylornot in my world19:38
* soren goes looking in /var/lib/dhuson19:38
sorenhudson, even.19:38
dabosirp-: jk0: no love here. Just ran again, and same fail19:39
sirp-jk0: is it possible that the tests broke at the same *time*19:39
mtaylorsoren: well, bash run_tests.sh19:39
mtayloris in test_nova.sh19:39
sirp-hmm19:39
sorenbash run_tests.sh -N, surely?19:40
sorenYes.19:40
mtaylorsoren: btw- it would be EVERSOGREAT if we didn't leave so many thousands of tmpdirs in /tmp19:40
mtaylorsoren: what's -N ?19:40
mtaylorsoren: and yes, there is a -N there19:40
mtaylorsoren: I _really_ want to change that to be python setup.py test intead - can I do that yet?19:41
soren-N means "don't use virtualenv"19:41
mtaylorsigh19:41
sorenThat's good!19:41
* mtaylor punches virtualenv in the kidney19:42
sorenWhat are those tmpdirs all about?19:42
sorenOh, images.19:42
sorenOk, I'm going to run hudson's postinst again just to see if it mangles something in /usr/lib/python again.19:43
mtaylorok19:43
mtaylorjaypipes: whatever happened to your branch to get rid of run_tests.py ?19:44
sorenthe hell?19:44
mtaylorsoren: did it?19:44
edaymtaylor: I got rid of it, then termie added it back again, but simply as a nose wrapper for pretty output19:46
edayit's no longer a nova-specific test harness doing some setup beforehand19:46
sorenmtaylor: Yes. Yes, it did.19:46
mtayloreday: python setup.py test does not work19:46
mtaylorsoren: it's that find command19:47
sorenmtaylor: It is. I just dont' quite understand how or why.19:47
mtayloreday: without run_tests.sh being called.19:47
edaymtaylor: hmm, so, in theory, nosetests should work just fine still19:47
sorenI will soon, though.19:47
* soren applied stubbornness19:47
mtayloreday: it doesn't19:47
edaymtaylor: oh, with 1 issue, does the sqlite db needs to be removed between noe runs19:47
edaynova.sqlite19:48
sirp-jk0, dabo: tests pass on the second try :/19:48
mtayloreday: ok. that works19:48
jk0sirp-: they're still good for me, too. not sure what the deal is19:48
mtaylorthat's pretty fail. why doesn't it clean up after itself? grumble grumble19:48
dabosirp-: I'm on like my tenth try. None of them, even the second, passed19:48
openstackhudsonProject nova build #384: SUCCESS in 1 min 24 sec: http://hudson.openstack.org/job/nova/384/19:48
openstackhudsonTarmac: Fixed a number of issues with the iptables firewall backend:19:48
openstackhudson* Port specifications for firewalls come back from the data store as integers, but were compared as strings.19:48
openstackhudson* --icmp-type was misspelled as --icmp_type (underscore vs dash)19:48
openstackhudson* There weren't any unit tests for these issues.19:48
sorenBam!19:48
edaymtaylor: I tried, it's a mess sinse there is session caching between tests with a singleton19:48
edaysinse? man brain not working today19:49
mtayloreday: hehe19:50
jk0eday: have a bacon cheeseburger, that ought to get the cells moving ;)19:50
mtayloreday: oh well. one of these days19:50
*** hadrian has quit IRC19:50
edaymtaylor: I tried switching to a memory-only db for tests, but that broke too because *sometimes* there is a different session var19:50
edaymtaylor: I decided to save it for another day :)19:50
mtaylor++19:51
edayjk0: that would only slow things down further, and make it more toxic :)19:51
jk0haha19:51
edayalthough I did have an amazing vegan bacon cheeseburger when I was in NYC in oct, mmm19:51
edayalong with a vegan butterfinger shake19:52
jk0I tried the veggie burger from BK last night. wasn't impressed19:52
* creiht ponders how that can even be possible :)19:52
edayjk0: does BK have anything that impresses you?19:52
jk0not really :(19:52
jk0well, except the BK stackers19:52
sorenmtaylor: Ok, I've found the culprit.19:52
sorenmtaylor: /var/lib/hudson/src/nova/test-venv19:53
sorenmtaylor: More specifically /var/lib/hudson/src/nova/test-venv/.nova-venv/lib/python2.619:53
mtaylorha19:54
sorenmtaylor: Is that needed for anything?19:54
edaycreiht: http://www.blossomnyc.com/  try it if you are ever there :)19:54
mtaylorsoren: nope19:54
edaycreiht: I was surprised too19:54
*** westmaas has joined #openstack19:54
mtaylorsoren: although - shouldn't that find command be fixed to not follow symlinks?19:54
sorenmtaylor: Quite likely, yes.19:54
sorenmtaylor: Not sure where to report bugs about it though.19:55
creihteday: interesting19:55
sorenmtaylor: It most likely involves something insane like Jira or whatnot.19:55
* soren shudders at the thought19:55
mtaylorsoren: directly on hudson ... if you can fix the find command, I can follow up on the bug19:55
*** trin_cz has joined #openstack19:55
creihtThe components sound like they would taste good, but at what point does it really stop being a bacon cheesburger, and just let it stand on its own merits? :)19:55
*** sophiap has quit IRC19:56
edaycreiht: agreed, could stand on its own. I guess it's just easier to describe it that way19:57
sorenmtaylor: Wicked. Give me a minute to test my changes.19:58
*** kpepple has quit IRC19:58
sorenmtaylor: ...but are you saying Hudson has a bugtracker?19:58
sorenmtaylor: I mean... The software, not the project.19:58
*** hadrian has joined #openstack19:58
*** gustavomzw has joined #openstack19:59
mtaylorsoren: oh no. they use jira19:59
mtaylorsoren: I'm just saying I'm in there already, so getting a patch submitted shouldn't be too terrible19:59
ttxTeam meeting in one hour in #openstack-meeting20:01
*** gustavomzw has quit IRC20:01
*** irahgel has joined #openstack20:01
sorenmtaylor: http://pastebin.com/GvswrHc320:02
* soren breaks20:04
dabosirp-: jk0: I've verified that tests pass in r542 but fail in r543. Looking further...20:05
*** mray has quit IRC20:07
*** fabiand_ has quit IRC20:09
nelson__Is there a standard cluster sysadmin solution recommended for OpenStack? I'm looking in the direction of Puppet, but are there other choices?20:11
*** kpepple has joined #openstack20:13
ttxnelson__: The other choice would be Chef.20:15
nelson__ja sure! B0rk, b0rk, b0rk!20:16
*** sophiap has joined #openstack20:21
*** ctennis has quit IRC20:22
creihtnelson__: Chef and puppet are both mentioned often on here20:22
creihtI doubt there will be an "annointed" deployment system for openstack20:23
*** PiotrSikora has quit IRC20:26
*** WonTu has joined #openstack20:26
*** WonTu has left #openstack20:26
nelson__Would make the installation instructions easier if there was.20:27
nelson__Cuz once there was, the "other" unchosen one would feel the pressure to contribute docs.20:27
nelson__Two sets for the price of one. :)20:28
*** PiotrSikora has joined #openstack20:30
dabosirp-: jk0: deleting the nova.sqlite file first fixed the errors20:32
*** littleidea has quit IRC20:32
jk0ah20:34
*** ccustine has quit IRC20:34
*** adiantum has joined #openstack20:36
*** irahgel has left #openstack20:38
*** trin_cz has quit IRC20:39
*** pandemicsyn has quit IRC20:39
*** zykes- has quit IRC20:39
*** pothos has quit IRC20:39
*** ewanmellor has quit IRC20:39
*** sparkycollier has quit IRC20:39
*** hky has quit IRC20:39
*** jaypipes has quit IRC20:39
*** ChrisAM has quit IRC20:39
*** gholt has quit IRC20:39
*** Daviey has quit IRC20:39
*** dabo has quit IRC20:39
*** soosfarm_ has quit IRC20:39
*** herki has quit IRC20:39
*** mjmac has quit IRC20:39
*** termie has quit IRC20:39
*** creiht has quit IRC20:39
*** Adri2000 has quit IRC20:40
*** patri0t has quit IRC20:40
*** xtoddx has quit IRC20:40
*** morfeas has quit IRC20:40
*** pothos has joined #openstack20:41
*** littleidea has joined #openstack20:42
*** kpepple has left #openstack20:42
*** trin_cz has joined #openstack20:42
*** pandemicsyn has joined #openstack20:42
*** zykes- has joined #openstack20:42
*** ewanmellor has joined #openstack20:42
*** sparkycollier has joined #openstack20:42
*** hky has joined #openstack20:42
*** jaypipes has joined #openstack20:42
*** ChrisAM has joined #openstack20:42
*** gholt has joined #openstack20:42
*** dabo has joined #openstack20:42
*** soosfarm_ has joined #openstack20:42
*** herki has joined #openstack20:42
*** mjmac has joined #openstack20:42
*** termie has joined #openstack20:42
*** Adri2000 has joined #openstack20:42
*** creiht has joined #openstack20:42
*** patri0t has joined #openstack20:42
*** xtoddx has joined #openstack20:42
*** morfeas has joined #openstack20:42
*** zelazny.freenode.net sets mode: +vv pandemicsyn creiht20:42
*** critch has left #openstack20:42
*** adiantum has quit IRC20:43
*** adiantum has joined #openstack20:50
*** Daviey has joined #openstack20:51
*** ctennis has joined #openstack20:55
ttxTeam Meeting in 4 minutes in #openstack-meeting20:56
sorenmtaylor: Can you take that and run with it, or do I need to do more?20:56
sorendubsquared1: https://code.launchpad.net/~soren/nova/iptables-security-groups/+merge/45826 should address that problem you mentioned yesterday, by the way.20:57
sorendubsquared1: (It already landed)20:57
*** adiantum has quit IRC20:59
mtaylorsoren: I'm working on submitting a pull request right now - but github seems to be a bit failwhale atm21:00
ttxMeeting starting NOW!21:00
*** littleidea has quit IRC21:02
*** lvaughn_ has joined #openstack21:12
*** hggdh has joined #openstack21:15
*** lvaughn has quit IRC21:15
sorenmtaylor: Thanks. You are hereby granted an alcoholic beverage of your own choice for saving me from having to deal with jira.21:20
dubsquared1soren:  awesome, ill give it a test here in a moment :D21:20
mtaylorsoren: woot21:20
*** johnpur has quit IRC21:24
*** sophiap has quit IRC21:28
*** mray has joined #openstack21:28
ttxjaypipes: for image-service-use-glance-clients, does sirp-'s branch cover it, or do you have more in store ?21:29
ttxsoren: that's a cheap price to pay to avoid dealing with JIRA.21:32
sorenttx: Yes!21:33
*** sophiap has joined #openstack21:37
*** gustavomzw has joined #openstack21:37
*** gustavomzw has quit IRC21:39
*** jt_zg has quit IRC21:45
*** alekibango has joined #openstack21:51
*** miclorb has joined #openstack21:51
openstackhudsonProject nova build #385: SUCCESS in 1 min 32 sec: http://hudson.openstack.org/job/nova/385/21:53
openstackhudsonTarmac: Read Full Spec for implementation details and notes on how to boot an instance using OS API.21:53
openstackhudsonhttp://etherpad.openstack.org/B2RK0q1CYj21:53
openstackhudsonLook at these notes for known issues:21:53
openstackhudsonhttp://etherpad.openstack.org/BXOU0TTj9M21:53
*** adiantum has joined #openstack21:55
*** opengeard has joined #openstack21:56
*** lorinh1 has quit IRC22:00
*** adiantum has quit IRC22:01
*** brd_from_italy has joined #openstack22:03
*** westmaas has quit IRC22:05
*** schisamo has joined #openstack22:06
*** nelson__ has quit IRC22:13
*** nelson__ has joined #openstack22:13
*** kainam- is now known as arcane22:13
*** adiantum has joined #openstack22:14
*** jdarcy has quit IRC22:18
*** ccustine has joined #openstack22:18
*** kashyapc has quit IRC22:30
*** littleidea has joined #openstack22:30
*** adiantum has quit IRC22:33
*** brd_from_italy has quit IRC22:35
*** rnirmal has quit IRC22:35
*** DubLo7 has joined #openstack22:35
dubsquared1soren:  has the iptables fix already been rolled into the packages?22:36
*** kashyapc has joined #openstack22:38
*** DubLo7 has quit IRC22:41
*** ppetraki has quit IRC22:44
*** adiantum has joined #openstack22:45
gholtperm: Can you check storage21-z2.dfw1 out? Just noticed that if you cat /etc/bash.bashrc you get garbage.22:47
gholtwc :(22:48
*** sophiap has quit IRC22:53
rluciodubsquared1: i think you are referring to commit 544... the trunk ppa is now on 54523:00
*** adiantum has quit IRC23:01
uvirtbotNew bug: #701706 in nova "Security Group w/o enough parameters should fail more gracefully" [Wishlist,New] https://launchpad.net/bugs/70170623:06
dubsquared1rlucio:  just got the new packages, and still getting this when launching an image —> http://paste.openstack.org/show/465/23:06
*** adiantum has joined #openstack23:06
rluciowhoa23:07
rlucioim getting an error as well, but not the same as yours23:07
dubsquared1ha23:07
xtoddxQuick bugfix for a condition nobody will probably ever hit needs approves: https://code.launchpad.net/~anso/nova/lp701706/+merge/4592923:07
dubsquared1what are you seeing?23:09
*** blamar has quit IRC23:10
rluciohttp://paste.openstack.org/show/466/23:10
rluciolooks like another typo on mine23:10
rlucioif you look at the line 2723:11
dubsquared1i see23:11
rlucio /shrug23:11
rlucioi was getting a different issue earlier too, but maybe they are all related23:12
*** adiantum has quit IRC23:14
dubsquared1have i already asked you what images you are using?23:14
rluciono23:14
rlucioyou mean nova build or vm images23:14
dubsquared1rlucio:  what images are you using?23:15
dubsquared1vm images23:15
dubsquared1:D23:15
rluciooh, i made my own23:15
rlucio8.04 and 10.04 atm23:16
dubsquared1coo, im using my own as well23:16
dubsquared1and a UEC 10.10 image23:16
rlucioah, yea i dont exactly get the whole uec thing, so i haven't messed with it23:16
dubsquared1uec-publish-tarball (imagetarball) *bucketname* $arch23:17
dubsquared1donezo23:17
dubsquared1well, supposed to be...23:17
*** gondoi has quit IRC23:20
*** kashyapc has quit IRC23:25
*** adiantum has joined #openstack23:26
*** kashyapc has joined #openstack23:27
openstackhudsonProject nova build #386: SUCCESS in 1 min 24 sec: http://hudson.openstack.org/job/nova/386/23:33
openstackhudsonTarmac: Quick bugfix.  Also make the error message more specific and unique in the equivalent code in the revoke method.23:33
*** kashyapc has quit IRC23:37
*** gustavomzw has joined #openstack23:38
*** gustavomzw has quit IRC23:40
*** adiantum has quit IRC23:43
rluciodubsquared1: oh thats interesting, that script works with nova?23:44
rluciouec-publish-tarball i mean23:44
*** kashyapc has joined #openstack23:44
dubsquared1ja, with images that are packaged for it23:45
dubsquared1it's great when it works23:45
*** reldan has quit IRC23:46
*** zul has quit IRC23:51
*** dubsquared1 has quit IRC23:51
*** DubLo7 has joined #openstack23:53
*** abecc has quit IRC23:54
*** dirakx has quit IRC23:55
*** adiantum has joined #openstack23:55

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