Friday, 2011-05-13

*** adjohn has joined #openstack00:02
*** joearnold has quit IRC00:04
*** j05h has quit IRC00:08
*** adjohn has quit IRC00:13
*** jeffjapan has joined #openstack00:14
*** AXiS_SharK has quit IRC00:14
*** AXiS_SharK has joined #openstack00:18
AXiS_SharKhow does multi-tenancy work in openstack?00:19
AXiS_SharKis it just users and projects?00:19
AXiS_SharKis there a high-level architecture diagram somewhere?00:19
*** jguerrero__ has left #openstack00:20
*** adjohn has joined #openstack00:20
*** Dumfries has quit IRC00:22
*** AXiS_SharK1 has joined #openstack00:24
*** AXiS_SharK has quit IRC00:24
*** j05h has joined #openstack00:25
*** AimanA has quit IRC00:28
*** AimanA has joined #openstack00:29
*** adjohn has quit IRC00:30
*** holoway has quit IRC00:31
*** holoway has joined #openstack00:33
*** AXiS_SharK1 has quit IRC00:39
*** AXiS_SharK has joined #openstack00:45
*** adjohn has joined #openstack00:51
nelsongholt and creiht: I'm working on checking my wikipedia code into our SVN, which will make it more public than it's been.00:58
nelsonI've got some needed changes to common/client.py which need to get into Swift.00:58
nelsonShould I just propose it as a patch, or do you want to give it a sanity/style look-over first?00:59
*** jdurgin has quit IRC00:59
nelsonIt's a brand-new function, for writing chunks, so it won't break anything. Safe enough to include in the next version.01:00
gholtWhichever you prefer, we'll take it. :)01:00
creihtnelson: easiest way is to propose as a merge proposal, and that will give us an easy way to look it over01:00
nelsonOkay, will do.01:00
nelsonHopefully you won't spew chunks all over it. :)01:00
gholtHeheh, if you have launchpad account, signed the CLA thingie and all that, doing a merge prop is the /best/ way.01:00
nelsonyep, I already have dox in 1.301:01
*** nerens has quit IRC01:01
gholtAh yeah, coolio then01:01
gholtAnd don't worry, every merge prop gets ripped up. Well, most every one. :)01:01
gholtI mean that in a good way of course, not just silly critiques dangit. ;)01:02
nelsonI'll be sure to leave a HT in there just to give you something to complain about.01:03
gholt\o/ Make sure you leave a PEP8 fail somewhere too; to see if we're doing our jobs.01:03
*** llang629 has joined #openstack01:04
*** llang629 has left #openstack01:04
nelsonIt's pretty low-tech -- just an unwravelling of put_object into a class with three functions: __init__, write, and close() And I suppose I ought to have a destructor which calls close() just in case they don't.01:05
*** jwilmes has quit IRC01:06
*** alex_openstack has joined #openstack01:09
*** AXiS_SharK has quit IRC01:10
*** AXiS_SharK has joined #openstack01:12
*** dittob has joined #openstack01:14
*** pguth66 has quit IRC01:24
*** AimanA has joined #openstack01:26
*** adjohn has quit IRC01:32
*** AimanA has quit IRC01:35
*** s1cz- is now known as s1cz01:38
*** nelson has quit IRC01:38
*** nelson has joined #openstack01:39
*** maple_bed has joined #openstack01:39
*** maplebed has quit IRC01:42
*** maple_bed has quit IRC01:42
*** MarkAtwood has quit IRC01:53
*** dittob has quit IRC02:04
*** kaz_ has quit IRC02:08
*** AXiS_SharK has quit IRC02:15
*** AXiS_SharK has joined #openstack02:16
*** AXiS_SharK has quit IRC02:21
*** paltman has quit IRC02:29
*** paltman has joined #openstack02:30
*** msivanes has quit IRC02:31
*** winston-d has quit IRC02:34
*** guynaor has joined #openstack02:34
*** hadrian has quit IRC02:35
*** shentonfreude has joined #openstack02:47
*** kennethkalmer has quit IRC03:02
*** kennethkalmer has joined #openstack03:10
*** adjohn has joined #openstack03:23
HugoKuoStill thinking about  Nova-network  HA :<03:23
*** photron_ has joined #openstack03:37
*** AlexNeef has quit IRC03:46
*** jbryce has quit IRC03:46
*** Binbinaway has quit IRC03:55
*** MotoMilind has quit IRC03:55
*** Binbinaway has joined #openstack03:55
*** guynaor has left #openstack03:55
*** MotoMilind has joined #openstack03:56
*** dysinger has quit IRC03:59
*** dysinger has joined #openstack04:00
*** kennethkalmer has quit IRC04:01
*** nelson has left #openstack04:05
*** kennethkalmer has joined #openstack04:09
*** krish|wired-in has joined #openstack04:17
alex_openstacki use euca-run-instances ami-47030828 -k mykey -t m1.tiny04:22
alex_openstackfor single node mode04:22
alex_openstackbut can not running04:22
alex_openstackINSTANCE        i-00000008      ami-47030828    192.168.2.9     192.168.2.9     shutdown        mykey (dubproject, cgtcloud)    0               m1.tiny 2011-05-13T04:20:50Z    nova04:22
*** cdbs has quit IRC04:23
*** cdbs has joined #openstack04:23
*** cdbs has joined #openstack04:23
*** Zangetsue has joined #openstack04:28
*** hagarth has joined #openstack04:32
*** Zangetsue has left #openstack04:34
*** Zangetsue has joined #openstack04:34
*** AXiS_SharK has joined #openstack04:34
*** John____ has joined #openstack04:40
John____Hi, I have download virtual box image on this link: http://wiki.openstack.org/NovaVirtually, but failed to boot, Error: alert /dev/mapper/ubuntu-root does not exist, dropping to a shell, anyone have idea?04:41
*** f4m8_ is now known as f4m804:42
*** katkee has joined #openstack04:44
*** AXiS_SharK has quit IRC04:47
*** BK_man has joined #openstack04:54
*** omidhdl has joined #openstack05:04
*** shentonfreude has quit IRC05:13
*** kashyap has joined #openstack05:15
*** santhosh has joined #openstack05:16
*** zenmatt_ has joined #openstack05:29
*** zenmatt has quit IRC05:29
*** boncos has joined #openstack05:29
*** hggdh has joined #openstack05:29
*** katkee has quit IRC05:33
*** katkee has joined #openstack05:35
*** MarkAtwood has joined #openstack05:36
*** guigui1 has joined #openstack05:46
*** miclorb has quit IRC05:46
*** MarkAtwood has left #openstack05:46
*** AXiS_SharK has joined #openstack05:47
*** zenmatt_ has quit IRC05:47
*** katkee has quit IRC05:57
*** anticw has quit IRC06:04
*** anticw has joined #openstack06:05
*** romero has joined #openstack06:12
*** matiu has quit IRC06:19
*** nerens has joined #openstack06:22
*** matiu has joined #openstack06:23
*** nerens has quit IRC06:24
*** obino has quit IRC06:24
*** matiu has quit IRC06:28
*** matiu has joined #openstack06:30
*** matiu has joined #openstack06:30
*** matiu has quit IRC06:39
*** fabiand__ has joined #openstack06:44
*** MarcMorata has joined #openstack06:51
*** MarcMorata has quit IRC06:52
*** AXiS_SharK has quit IRC06:57
*** med_out is now known as med07:02
*** med is now known as medberry07:02
*** ryker has quit IRC07:03
*** hggdh has quit IRC07:04
*** Kronick has joined #openstack07:08
*** obino has joined #openstack07:10
*** zul has joined #openstack07:10
*** lionel has quit IRC07:12
*** dobber has joined #openstack07:13
*** lionel has joined #openstack07:13
boncoshi ... how to setup staticweb in swift ?07:17
dobberHi. I just configured swift. Can I upload a big file that is stored on different servers at the same time. like chunk1 on storage6, chunk2 on storage3 etc07:21
*** CloudChris has joined #openstack07:22
*** CloudChris has left #openstack07:22
dobberis XenServer 5.6 supported ?07:28
*** zul has quit IRC07:30
*** zul has joined #openstack07:30
*** cuzoka has joined #openstack07:33
dobberFull openvz support for nova07:36
dobberThis was reported as started work in Feb, is it ready ?07:36
*** nerens has joined #openstack07:37
*** rcc has joined #openstack07:40
*** alex_openstack has quit IRC07:41
dobberwhat is approximate ETA for diablo ?07:43
*** ryker has joined #openstack07:44
photron_dobber: http://wiki.openstack.org/DiabloReleaseSchedule07:49
*** John____ has quit IRC07:51
*** nacx has joined #openstack07:51
dobbertnx07:52
*** obino has quit IRC07:52
*** adjohn has quit IRC07:53
*** zul has quit IRC07:53
*** krish|wired-in has quit IRC07:55
uvirtbotNew bug: #782050 in swift "Add Child Node to <subdir> in container listings with delimiters" [Undecided,New] https://launchpad.net/bugs/78205007:56
*** dirkx_ has joined #openstack07:57
*** mgoldmann has joined #openstack07:58
dobberok, i got permission to use 5 servers and setup proof of concept :)08:01
dobbergood buy VMs ;)08:01
*** daveiw has joined #openstack08:05
*** ChameleonSys has quit IRC08:07
*** rchavik has joined #openstack08:07
*** ChameleonSys has joined #openstack08:07
*** obino has joined #openstack08:08
*** Eyk has joined #openstack08:11
*** cloudgroups has joined #openstack08:11
*** dirkx_ has quit IRC08:12
*** obino has quit IRC08:13
*** obino has joined #openstack08:13
dobberhow can i monitor the daily patches in the source tree ?08:14
*** cloudgroups has left #openstack08:20
*** mgoldmann has quit IRC08:22
*** mgoldmann has joined #openstack08:23
*** Jordandev has joined #openstack08:24
*** allsystemsarego has joined #openstack08:29
*** dirkx_ has joined #openstack08:35
dobberwhat about hardware recommendations ? witch node, what cpu/ram needs ?08:36
*** Capashen has joined #openstack08:37
HugoKuobetter and better :>08:37
*** nerens has quit IRC08:39
*** katkee has joined #openstack08:40
*** anticw has quit IRC08:43
*** anticw has joined #openstack08:43
*** obino has quit IRC08:44
*** obino has joined #openstack09:00
*** boncos has quit IRC09:05
*** obino has quit IRC09:08
HugoKuolooking for network HA solution09:10
*** skiold has joined #openstack09:19
sorendobber: https://code.launchpad.net/~hudson-openstack/nova/trunk/+edit-subscription09:20
*** Binbinaway has quit IRC09:21
*** photron_ has quit IRC09:21
*** robbiew has joined #openstack09:22
*** rchavik has quit IRC09:24
*** GasbaKid has joined #openstack09:32
*** krish|wired-in has joined #openstack09:32
*** infinite-scale has joined #openstack09:40
*** kennethkalmer has quit IRC09:41
*** miclorb_ has joined #openstack09:41
radekI'm trying to attach cdrom image to existing instance by editing libvirt.xml in the instance folder but after reboot I can't see any change09:44
*** robbiew has quit IRC09:51
*** miclorb_ has quit IRC09:58
*** CloudChris has joined #openstack09:58
*** salv-orlando has joined #openstack10:06
*** zul has joined #openstack10:06
*** Ryan_Lane has joined #openstack10:11
*** obino has joined #openstack10:15
sorenradek: That file is only read once: At the first instantiation time.10:16
sorenradek: If you want to edit it later on, use "virsh edit <instance name>"10:17
radekok thanks so only way is to change libvirt.xml.template ?10:17
sorenNo, that'll do even less :)10:17
soren10:17 <+soren> radek: If you want to edit it later on, use "virsh edit <instance name>"10:18
radekvirsh edit and then reboot instance ?10:18
*** obino has quit IRC10:19
sorenradek: Yeah.10:19
*** miclorb_ has joined #openstack10:19
*** miclorb_ has quit IRC10:21
*** dirkx_ has quit IRC10:22
*** zul has quit IRC10:25
*** kennethkalmer has joined #openstack10:25
*** zul has joined #openstack10:27
*** obino has joined #openstack10:27
*** obino has quit IRC10:29
*** Ryan_Lane has quit IRC10:33
*** mgoldmann has quit IRC10:35
*** zul has quit IRC10:42
HugoKuodoes nova support instance or volume snapshot ?10:44
*** medberry is now known as med_out10:46
*** CloudChris has left #openstack10:48
*** GasbaKid has quit IRC10:55
*** kennethkalmer has quit IRC11:07
*** kennethkalmer has joined #openstack11:11
uvirtbotNew bug: #782138 in openstack-manuals "README text in rc-maven-cloud-docs dir just says, "Don't forget to put this under a license"" [Undecided,New] https://launchpad.net/bugs/78213811:11
*** kennethkalmer has quit IRC11:13
*** ctennis has quit IRC11:39
RichiH#782138 is somewhat ironic, isn't it? ;)11:40
*** skiold has quit IRC11:42
*** Ryan_Lane has joined #openstack11:45
*** mgoldmann has joined #openstack11:46
*** citral has joined #openstack11:54
*** ctennis has joined #openstack11:54
*** ctennis has joined #openstack11:54
*** xavicampa has quit IRC11:55
uvirtbotNew bug: #782152 in nova "Documentation typos" [Undecided,New] https://launchpad.net/bugs/78215211:56
*** paltman has quit IRC12:02
*** Eyk has quit IRC12:03
*** obino has joined #openstack12:03
*** paltman has joined #openstack12:03
*** obino has quit IRC12:04
*** arb10 has joined #openstack12:15
*** markvoelker has joined #openstack12:16
arb10can anyone point me in the direction of a reference architecture diagram for a production swift cluster.  The admin guide shows some good information but doesnt go into much detail regarding the network12:17
*** dendro-afk is now known as dendrobates12:18
*** m4dbi7 has joined #openstack12:25
m4dbi7Hi12:26
*** santhosh has quit IRC12:31
*** namaqua has joined #openstack12:33
*** dendrobates is now known as dendro-afk12:37
*** mgoldmann has quit IRC12:40
*** mgoldmann_ has joined #openstack12:40
*** radek has quit IRC12:48
*** kashyap has quit IRC12:51
*** dendro-afk is now known as dendrobates12:55
*** salv-orlando has quit IRC12:55
*** bkkrw has joined #openstack12:56
*** hadrian has joined #openstack12:57
*** kakella has joined #openstack12:58
*** kakella has left #openstack12:58
*** lorin1 has joined #openstack12:58
*** xavicampa has joined #openstack12:59
*** namaqua has quit IRC13:00
*** hggdh has joined #openstack13:04
*** hagarth has quit IRC13:07
*** krish|wired-in has quit IRC13:08
*** med_out is now known as medberry13:08
*** amccabe has joined #openstack13:09
*** msivanes has joined #openstack13:10
*** Kronick1 has joined #openstack13:10
*** Kronick has quit IRC13:14
*** dprince has joined #openstack13:17
*** thickski_ has joined #openstack13:20
*** thickskin has quit IRC13:21
alekibangoarb10: you need at least one proxy node... and how many other nodes as you wish, at least 5 for every service13:21
*** drogoh has quit IRC13:21
alekibangominimum is 3, but 5 makes sense13:22
*** drogoh has joined #openstack13:22
*** mgoldmann_ has quit IRC13:23
*** citral has quit IRC13:26
*** Zangetsue_ has joined #openstack13:27
*** amccabe has quit IRC13:28
*** Zangetsue has quit IRC13:28
*** Zangetsue_ is now known as Zangetsue13:28
*** alekibango has quit IRC13:30
*** cloudgroups has joined #openstack13:31
arb10Was thinking for a 5 zone production 1 PetaByte cluster I'd need 10 storage nodes per zone (so thats 50 nodes) with each node roughly having 100 TByte storage.  Was also thinking having two pools of 3 proxy nodes, load balanced.  is this overkill13:32
*** shentonfreude has joined #openstack13:32
*** infinite-scale has quit IRC13:33
*** Kronick1 has quit IRC13:33
*** dendrobates is now known as dendro-afk13:33
*** amccabe has joined #openstack13:34
*** amccabe has quit IRC13:35
creihtarb10: that would give you around 1.3-1.5 usable PB for customer data13:35
creiht100*50/3.0*0.813:36
creiht(assuming that you limit the storage to 80-85% utilization13:36
*** Ryan_Lane has quit IRC13:36
*** Ryan_Lane1 has joined #openstack13:36
*** Ryan_Lane1 has quit IRC13:36
*** Ryan_Lane1 has joined #openstack13:36
creiht3 proxy nodes should get you to at least 3000 requests/sec sustained13:36
creiht+/-13:37
*** Ryan_Lane1 is now known as Ryan_Lane13:37
creihtand if you need more requests, you can just add more proxies13:37
*** gondoi has joined #openstack13:37
*** jaypipes has joined #openstack13:39
*** medberry is now known as med_out13:44
*** ctennis has quit IRC13:48
*** f4m8 is now known as f4m8_13:49
arb10creiht: what about memory requirements for proxies, is 24 GB enough?13:49
*** Kronick has joined #openstack13:49
*** AlexNeef has joined #openstack13:50
*** amccabe has joined #openstack13:50
*** hggdh has quit IRC13:54
*** AlexNeef has quit IRC13:56
*** cp16net has joined #openstack13:57
*** katkee has quit IRC13:59
*** katkee has joined #openstack13:59
*** nelson has joined #openstack14:01
*** Zangetsue has quit IRC14:03
*** heden has quit IRC14:06
*** cp16net has quit IRC14:08
*** ctennis has joined #openstack14:12
gholtarb10: I think creiht had to run, but yeah 24G is plenty.14:20
*** krish|wired-in has joined #openstack14:22
*** xavicampa has quit IRC14:23
*** AXiS_SharK has joined #openstack14:27
*** daveiw has left #openstack14:27
*** joshfng has joined #openstack14:28
*** joshfng has quit IRC14:29
*** obino has joined #openstack14:31
*** AXiS_SharK has quit IRC14:31
*** cloudgroups has left #openstack14:32
*** zenmatt has joined #openstack14:33
*** jkoelker has joined #openstack14:36
*** hallyn has quit IRC14:36
*** hallyn has joined #openstack14:36
*** piken_afk is now known as piken14:39
*** nerens has joined #openstack14:39
*** cuzoka has quit IRC14:40
*** hallyn_ has joined #openstack14:41
*** Kronick has quit IRC14:41
*** Kronick has joined #openstack14:41
uvirtbotNew bug: #782237 in nova "logging loop when drivers don't implement set_admin_password" [Undecided,In progress] https://launchpad.net/bugs/78223714:41
nhmHugo: for snapshoting there is this bug for KVM: https://bugs.launchpad.net/nova/+bug/75054414:41
uvirtbotLaunchpad bug 750544 in nova "Snapshotting VM instance failing for KVM" [High,Fix committed]14:42
*** AXiS_SharK has joined #openstack14:43
*** dubsquared has joined #openstack14:45
*** joloughlin has joined #openstack14:46
*** MarkAtwood has joined #openstack14:46
*** dragondm has joined #openstack14:49
*** koolhead17 has joined #openstack14:54
arb10is there any documentation on how replication on the Auth Server.  For instance if i have 3 nodes each running a proxy and auth server and I create a new user (or update a password), how do the user details get replicated across all nodes?14:54
gholtarb10: Are you talking about swauth? If so, swauth stores its information in Swift itself as objects. So the usual object replication covers that.14:56
*** msivanes has quit IRC14:57
arb10gholt: yes swauth, that makes sense :)15:00
*** shentonfreude has quit IRC15:02
dobber(swift) so, i got 3 storage nodes, 2 of them have only one disk, 1 has 11 disks. do you have an idea for an ok-ish proof of concept setup15:03
dobberlike replica 215:03
gholtAre all the disks the same size?15:04
dobberno15:04
*** boncos has joined #openstack15:04
dobbermaybe i'll make replica 3 with 3 disks (one on each storage node)15:05
dobberfor starters15:05
gholtYeah, that might make more sense to prototype.15:06
gholtAnd just ignore that you have extra drives.15:06
gholtThere are ways to make use of oddly balanced nodes and drives, but I doubt you really want to fiddle with that sort of thing right off the bat.15:07
*** kashyap has joined #openstack15:08
dobberyou are right15:08
dobberi'll need strange configurations - later ;)15:08
*** maple_bed has joined #openstack15:10
*** zenmatt has quit IRC15:10
*** joloughlin has quit IRC15:11
*** maple_bed has quit IRC15:12
*** cuzoka has joined #openstack15:13
*** maple_bed has joined #openstack15:13
*** CloudChris has joined #openstack15:15
*** ccustine has joined #openstack15:15
CloudChrisHello together - is there any "Openstack in a Box" solution to quickly have a look at the whole stack on 1 machine?15:16
notmynameCloudChris: are you looking for object storage or compute or both?15:18
*** cweidenkeller has joined #openstack15:19
CloudChrishi, I am looking for the whole package :)15:19
CloudChrisof course I can install it in a few virtual machines, but a live CD or something like that would be finde15:19
notmynamefor swift there is http://swift.openstack.org/development_saio.html to run an entire swift cluster in one VM for testing/dev purposes15:21
notmynamenot sure what there is like that on the nova side15:21
openstackjenkinsProject nova build #894: SUCCESS in 2 min 44 sec: http://jenkins.openstack.org/job/nova/894/15:23
openstackjenkinsTarmac: Update the compute manager so that it breaks out of a loop if set_admin_password is not implemented by the driver. This avoids excessively logging NotImplementedError exceptions.15:23
*** cweidenkeller has quit IRC15:25
*** camm has quit IRC15:25
*** camm has joined #openstack15:26
openstackjenkinsProject nova build #895: SUCCESS in 2 min 41 sec: http://jenkins.openstack.org/job/nova/895/15:31
openstackjenkinsTarmac: Lost a flag pulling from another branch. Whoops.15:31
*** skiold has joined #openstack15:31
*** Kronick has quit IRC15:31
*** zenmatt has joined #openstack15:32
vishyCloudChris: http://docs.openstack.org/cactus/openstack-compute/admin/content/installing-openstack-compute-on-ubuntu.html15:32
*** cweidenkeller has joined #openstack15:32
CloudChris^^ thankx, I#ll check that15:33
uvirtbotCloudChris: Error: "^" is not a valid command.15:33
CloudChristhanks, ill check that vishy15:33
vishyCloudChris: or if you are more into bleeding edge trunk code, you can check out: http://cloudbuilders.github.com/deploy.sh/hacking-nova.html15:34
*** dprince has quit IRC15:34
*** rnirmal has joined #openstack15:37
jaypipesfriggin HVAC guy better get here soon... I think I'm melting.15:39
*** dobber has quit IRC15:40
*** guigui1 has quit IRC15:42
*** AXiS_SharK has quit IRC15:45
*** rnirmal has quit IRC15:46
boncosHi, how to setup "staticweb" (public container) in swift ?15:46
*** arun_ has quit IRC15:49
*** rnirmal has joined #openstack15:52
*** CloudChris has quit IRC15:52
annegentleboncos: here's the doc I could find on it - http://swift.openstack.org/misc.html#module-swift.common.middleware.staticweb sorry I don't know much about the operations side.15:52
boncosaahh, thx for pointing to that docs annegentle :)15:53
creihtarb10: and yeah the proxies are mostly cpu bound15:54
*** enigma1 has joined #openstack15:54
*** raggi_ has left #openstack15:54
*** shentonfreude has joined #openstack15:59
katkeehello, i have created and terminated 32 instances on a 32 ip range, now i cannot create instances anymore, any ideas? the ip seems all "frozen"...15:59
*** zenmatt has quit IRC16:02
katkeeit works again... just had to wait a certain amount of time for ip to be "freed"16:04
*** krish|wired-in has quit IRC16:06
*** obino has quit IRC16:07
openstackjenkinsProject nova build #896: SUCCESS in 2 min 38 sec: http://jenkins.openstack.org/job/nova/896/16:08
openstackjenkinsTarmac: Make set_admin_password non-blocking to API.16:08
*** omidhdl has quit IRC16:08
*** lborda has joined #openstack16:09
*** fabiand__ has left #openstack16:13
*** maple_bed is now known as maplebed16:15
*** llang629 has joined #openstack16:16
*** amccabe has joined #openstack16:17
boncos!ping gholt16:17
openstackpong16:17
notmynameboncos: he just went to lunch16:19
boncosnotmyname, ow...thanks16:19
*** joearnold has joined #openstack16:21
*** nacx has quit IRC16:21
*** katkee has quit IRC16:22
*** llang629 has left #openstack16:22
*** bkkrw has quit IRC16:26
*** hadrian has quit IRC16:29
jaypipeskpepple: hey, just read http://ken.pepple.info/openstack/2011/04/22/openstack-nova-architecture/. Excellent article, man.16:30
dubsquared+1, and OmniGraffle = win16:31
*** lborda has quit IRC16:32
jaypipesdubsquared: :)16:32
dubsquaredjaypipes:  hi, friend!  how is life?16:33
*** r5 has joined #openstack16:33
jaypipesdubsquared: good! trying to wrap up the new Glance versioned API stuff. In Greece all next week, so trying to get a bit ahead of schedule for D1 release.16:34
*** kashyap has quit IRC16:34
*** krish|wired-in has joined #openstack16:34
dubsquaredjaypipes: i see, living the tough life, traveling to Greece ...16:34
dubsquaredjaypipes:  you deserve it though, hope you have a great time!16:35
jaypipesdubsquared: hehe, you know it!16:35
jaypipesdubsquared: me too :) never been.16:35
*** krish|wired-in has quit IRC16:36
*** skiold has quit IRC16:39
*** cuzoka has quit IRC16:45
*** kashyap has joined #openstack16:46
*** BK_man has quit IRC16:48
blamarsirp_, If you get a second can you check this again? https://code.launchpad.net/~throughnothing/nova/natty_python_2_7/+merge/6014216:48
sirp_blamar: sure thing16:48
blamarmuch appreciated16:49
*** enigma1 has quit IRC16:50
openstackjenkinsProject nova build #897: SUCCESS in 2 min 39 sec: http://jenkins.openstack.org/job/nova/897/16:58
openstackjenkinsTarmac: XenAPI was not implemented to allow for multiple simultaneous XenAPI requests.  A single XenAPIConnection (and thus XenAPISession) is used for all queries.  XenAPISession's wait_for_task method would set a self.loop = for looping calls to _poll_task until task completion.  Subsequent (parallel) calls to wait_for_task for another query would overwrite this.  XenAPISession._poll_task was pulled into the XenAPISession16:58
*** jdurgin has joined #openstack16:58
*** BK_man has joined #openstack16:58
*** mszilagyi has joined #openstack16:58
BK_manhi all!16:59
BK_mando I need to configure a network switch to be able to run several hosts (multi-node Nova) with VlanManager?17:00
BK_manI can ping and ssh to instances which are hosted by cloud controller, but can't go to instances which are residing on both compute nodes.17:01
*** llang629__ has joined #openstack17:01
vishyBK_man: you have to configure the vlans on your switch17:04
*** mszilagyi has quit IRC17:04
*** mgoldmann has joined #openstack17:04
*** mgoldmann has joined #openstack17:04
BK_manvishy: nice. Can't find it on wiki.17:04
*** rnirmal has quit IRC17:04
vishyBK_man: essentially it means allowing vlan trunking on the ports that your hosts are using for vlan100 + X17:05
vishywhere X is the number of projects you want to support17:05
vishyBK_man: perhaps a general comment should be added to the wiki, but the actual config is switch-specific17:05
BK_manvishy: thanks, now going to Cisco guys...17:05
openstackjenkinsProject nova build #898: SUCCESS in 2 min 43 sec: http://jenkins.openstack.org/job/nova/898/17:06
openstackjenkinsTarmac: I'm assuming that openstack doesnt work with python < 2.6 here (which I read somewhere on the wiki).  This patch will check to make sure python >= 2.6 is installed, and also allow it to work with python 2.7 (and greater in the future).17:06
openstackjenkinsIn order for nova to work with python 2.7, it needs Cheetah 2.4.4, which also works fine in python 2.6 (all tests passed with Cheetah 2.4.4 and python 2.6 for me).17:06
*** Capashen has quit IRC17:06
*** mgoldmann_ has joined #openstack17:07
*** dragondm has quit IRC17:07
*** hadrian has joined #openstack17:09
*** mgoldmann has quit IRC17:10
*** JSharp has joined #openstack17:11
JSharpHey folks :) I've got a FlatDHCP setup, and the DHCP offers don't seem to be leaving the dhcp server, I'm looking at packet traces and I see the requests/discovers, but no offers, but in the syslog file I see the DHCP server offering the IP to the right MAC17:13
JSharpoddly enough, vms on the machine acting as the dhcp server can receive leases just fine17:14
*** Ryan_Lane has quit IRC17:15
*** Ryan_Lane has joined #openstack17:15
*** mattray has joined #openstack17:16
*** kennethkalmer has joined #openstack17:18
*** msivanes has joined #openstack17:26
BK_manquick question: InstanceLimitExceeded: Instance quota exceeded. You can only run 0 more instances of this type. How to allow more than 10 instances?17:29
*** krish|wired-in has joined #openstack17:32
*** rcc has quit IRC17:33
boncoshi, when i configure staticweb in swift, what URL is i can access to public container ?17:43
boncoshttp://swift.openstack.org/misc.html#module-swift.common.middleware.staticweb17:43
vishyBK_man: nova-manage quota instances 10017:44
vishyBK_man: er sorry, nova-manage project quota <proj_id> instances 10017:44
vishyBK_man: and nova-manage project quota <proj_id> cores 40017:45
*** Ryan_Lane has quit IRC17:46
*** dprince has joined #openstack17:47
*** rblackwe has quit IRC17:50
boncosah .. forget it .. i found it17:51
vishyJSharp: do you have --flat_interface and --my_ip set in your flagfile?17:52
koolhead17hi all17:52
JSharpvishy, yes, I do17:52
JSharplet me pastie my nova.conf17:52
vishyis it possible another dhcp server is handing out ips?17:53
JSharpvishy, not on that network, no. http://pastebin.com/RdH7cxZq17:54
JSharpthere's an eth0 control plane, with a public/flat_interface=eth117:54
vishyJSharp: why not use flat interface on eth0?17:55
vishyJSharp: in any case you may need --fixed_range=10.252.0.0/1617:55
JSharpvishy, what does the fixed_range do?17:56
vishyand clear out iptables rules and try again, I'm thinking you're getting conflicts with your host ips17:56
*** dprince has quit IRC17:56
JSharpvishy, okay :)17:56
JSharpthanks17:56
vishyfixed range is to set up snatting rules17:56
JSharpah17:56
JSharpthat might explain why everything is just going off into space17:56
JSharpthanks :)17:56
JSharpI'll give it a try17:57
*** katkee has joined #openstack17:57
vishyJSharp: I don't see anything else in there that looks troublesome, but you might want to do a iptables -L -n -v and an iptables -t nat -L -n -v and see if anything looks weird.17:57
vishyJSharp: I've had times where I create rules as i'm setting things up and forget to flush them in between conf changes17:58
vishygl i have to head out17:58
JSharpvishy, thanks :) I'm doing automated builds, so I start from scratch when I have to make huge changes17:59
vishyJSharp: one other thing.  In my experience if you use the same interface for flat network and public network, you have to turn promisc on on br100 in order to get public routing from vms on the other compute hosts.  Also make sure ip_forward is set on your network host.18:01
JSharpyep :)18:01
vishyJSharp: the promisc br100 only needs to be set on the network host18:01
JSharpthank you18:01
JSharp:)18:01
*** Tv has joined #openstack18:03
*** alekibango has joined #openstack18:03
openstackjenkinsProject nova build #899: SUCCESS in 2 min 44 sec: http://jenkins.openstack.org/job/nova/899/18:18
openstackjenkinsTarmac: Adds proper error handling for images that can't be found and a test for deregister image.18:18
*** gondoi has quit IRC18:21
JSharpvishy, thanks for the hint.. I'm no longer losing outside connectivity to my compute nodes (now they can reach the wider internet), but I'm still not getting any dhcp server response (I can see the message coming into the logs for dnsmasq but nothing coming out of the ip stack on to the wire)18:25
JSharpwireshark on the compute node shows the offer packets being generated.. so it's probably getting blocked in iptables somewhere18:27
JSharpbut I also see ARP reqs.18:27
*** dendro-afk is now known as dendrobates18:28
uvirtbotNew bug: #782328 in swift "Improve logging in object auditor" [Undecided,New] https://launchpad.net/bugs/78232818:31
*** vernhart has quit IRC18:31
*** vernhart has joined #openstack18:34
*** maplebed has quit IRC18:34
*** mgoldmann_ has quit IRC18:43
*** Jordandev has quit IRC18:44
*** krish|wired-in has left #openstack18:53
*** dragondm has joined #openstack18:56
*** xavicampa has joined #openstack18:59
*** MarkAtwood has quit IRC18:59
*** MarkAtwood has joined #openstack19:04
*** markvoelker has quit IRC19:04
arb10is there a mechanism to change a password in swauth?19:08
*** infinite-scale has joined #openstack19:15
*** msivanes has quit IRC19:17
*** vernhart has quit IRC19:17
*** omidhdl has joined #openstack19:19
*** pguth66 has joined #openstack19:25
*** guynaor has joined #openstack19:29
*** guynaor has left #openstack19:29
*** iRTermite has quit IRC19:32
*** iRTermite has joined #openstack19:33
*** jtran has joined #openstack19:33
*** mgius has joined #openstack19:36
vishyJSharp: how do you know the compute nodes can get out if they aren't getting dhcp?19:37
*** omidhdl has left #openstack19:37
*** aliguori has quit IRC19:37
*** koolhead17 has quit IRC19:38
*** fysa has quit IRC19:40
*** pguth66 has quit IRC19:40
*** dobber_ has joined #openstack19:41
uvirtbotNew bug: #782364 in nova "better to send gratuitous ARPs to support HA" [Undecided,New] https://launchpad.net/bugs/78236419:41
*** zenmatt has joined #openstack19:57
JSharpvishy, well, sniffing the eth1 interface on the network node yields dhcp requests/discovery packets, so they're getting out, and then the network node is responding to them.. my nova-network-floating-snat rule chain is empty on the network machine20:00
JSharpsorry, if that's not related.. maybe I should post an iptables-save20:00
*** joearnold has quit IRC20:00
*** joearnold has joined #openstack20:01
JSharpvishy, here's the iptables-save output: http://pastebin.com/b7fT9JPL20:02
*** maplebed has joined #openstack20:03
*** maplebed has quit IRC20:04
arb10anyone know if you can change a users password in swift once its been created?20:06
*** lele_ has joined #openstack20:07
lele_Hi all, someone can help me to figure out a network issue , i cant connect to my instances20:07
lele_Im able to launch them well, y see them running but i cant ping or ssh to i20:09
lele_them20:09
*** pguth66 has joined #openstack20:13
*** fabiand__ has joined #openstack20:16
*** fabiand__ has left #openstack20:20
vishyJSharp: do the offers show up on eth1/br100  on the compute host?20:21
JSharpvishy, no, they don't oddly20:21
JSharpbut they do show up on the eth1/br100 of the network host20:21
vishythat seems very strange20:21
vishyJSharp: is eth1 dhcping?20:22
JSharpeth1 has no IP assigned20:22
JSharpon either box20:22
JSharplet me do an iptables-save of the compute node that can't get dhcp stuff20:23
*** allsystemsarego has quit IRC20:23
vishyi have seen messages fail sometimes because ip addresses are in the wrong order on br10020:24
vishyso it sends the offer to the wrong broadcast address20:24
vishythat is the only thing i can think of20:24
*** ctennis has quit IRC20:24
JSharpvishy, iptables bits http://pastebin.com/fCuwbpnW20:25
JSharplet me check order of ips on br10020:25
vishy-A nova-network-POSTROUTING -s 10.252.0.0/16 -d 10.128.0.0/24 -j ACCEPT20:25
vishythat could cause some problems20:25
*** piken is now known as piken_afk20:26
*** j05h has quit IRC20:26
JSharpip addr output for machine having issues, http://pastebin.com/4HnZskBx20:27
JSharpand for network node: http://pastebin.com/iM1LtxdY20:27
vishyalso seems a little odd 10.252.0.0/2520:28
vishythat you are using a /2520:28
vishyinstead of a /2420:28
*** pguth66 has quit IRC20:28
JSharpvishy, indeed... I don't know why it's doing that... perhaps I should delete that network and recreate?20:28
*** aliguori has joined #openstack20:29
*** pguth66 has joined #openstack20:29
vishyprobably because you are doing the same error everyone else does20:29
vishynova-manage network create 10.252.0.0/24 1 25620:29
vishyeveryone puts in 255 for some reason20:29
vishyi think i see your problem20:30
vishyi don't see an ip on br10020:30
vishyah nm there it is20:31
JSharpvishy, thanks for your help, this has been driving me a bit spare :)20:32
vishyi would definitely switch to the full /24 instead of the /2520:32
JSharpyeah20:33
JSharpoaky20:33
JSharper20:33
JSharpokay20:33
vishybut i'm not sure what could be causing the packets to not get back to the other node20:33
vishyif you do brctl show on both machines20:33
vishyit shows br100 bridged into eth1 right?20:33
vishyoh by the way, since you have no ips on eth0, how are these getting out to the net?20:33
vishyer no ips on eth1 that is20:34
JSharpthe 10.254.0.0/24 space is natted20:34
vishyit looks like you should be using --public_interface=eth120:34
JSharpam I not? *checks*20:35
JSharpI am20:35
vishysorry eth020:35
vishyyour "public" needs to go out eth020:35
vishyunless you have some other route out on eth1?20:35
JSharpwell, I have a /24 floating IP block I'd like to assign to instances20:36
JSharpotherwise, I want them on rfc1918 only20:36
vishyok so the network host generally has a public ip20:36
vishyand the public_interface is the interface that that ip is on20:36
JSharpah, I see20:37
vishyif you have a private management network20:37
vishygenerally you put the intravm communication on the private interface20:37
vishyat least that is how we have always done it20:37
JSharpright right20:37
JSharpso, the floating public IPs are assigned to vms, but are routed through the network box in this configuration?20:38
vishyyes20:39
vishythey are natted to the private range20:39
JSharpah, I see...20:39
JSharpis the a way to avoid that, since that might become a bottleneck at some point?20:40
*** aliguori has quit IRC20:40
lele_vishy, im having a similar issue to JSharp, could you trough me a ladder ? , im lost20:42
*** enigma has joined #openstack20:43
*** KnuckleSangwich has quit IRC20:43
JSharpthe Vlan manager might do the trick20:44
vishyJSharp: no, can't avoid it yet.  People are working on various other approaches that would allow natting straight through20:44
*** jtran has left #openstack20:44
vishycompute hosts20:44
JSharpI see20:45
*** lorin1 has quit IRC20:46
Tvsmoser: yt?20:46
*** enigma has quit IRC20:47
*** ctennis has joined #openstack20:50
*** ctennis has joined #openstack20:50
*** enigma has joined #openstack20:52
*** MotoMilind has quit IRC20:53
JSharpvishy, I checked my recipes and yes, 255 was getting put in20:53
*** enigma has quit IRC20:55
*** MotoMilind has joined #openstack21:03
*** lele_ has quit IRC21:07
*** arb10 has quit IRC21:08
*** zenmatt has quit IRC21:09
*** AXiS_SharK has joined #openstack21:10
*** MotoMilind has quit IRC21:13
*** sam_itx has joined #openstack21:13
*** MotoMilind has joined #openstack21:14
sam_itxhi, learning / looking in to openstack architecture and details. have bunch of questions. first one, how authentication works between different components ? i.e once authenticated, how does "storage server" for example, know that the request is authenticated ?21:16
*** zenmatt_ has joined #openstack21:16
*** enigma has joined #openstack21:18
*** brd_from_italy has joined #openstack21:18
*** enigma has quit IRC21:19
notmynamesam_itx: (assuming swift because you mentioned "storage server") each authenticated request to a swift cluster includes an auth token. the swift cluster checks the token validity with the appropriate auth service21:24
*** aliguori has joined #openstack21:24
sam_itxyes. I meant swift server.. so flow is api request (with token from auth service) --> swift server--> validate token with auth service --> if valid.. proceed ?21:27
*** enigma has joined #openstack21:28
*** llang629__ is now known as llang62921:37
*** amccabe has quit IRC21:39
sam_itxis auth server <--> switft communication via messaging or http ?21:39
*** enigma has quit IRC21:42
*** anticw is now known as FUKAH21:46
*** FUKAH is now known as anticw21:48
*** enigma has joined #openstack21:52
*** MotoMilind has quit IRC21:52
*** vernhart has joined #openstack21:57
*** MotoMilind has joined #openstack21:57
*** MotoMilind has quit IRC22:02
*** MotoMilind has joined #openstack22:02
*** joearnold has quit IRC22:03
*** shentonfreude has quit IRC22:03
*** enigma has quit IRC22:05
*** maplebed has joined #openstack22:10
*** MotoMilind has quit IRC22:13
BK_mannova-manage network create 10.20.30.0/20 1 4000 -> ValueError: IP('10.20.30.0/20') has invalid prefix length (20)22:14
*** MotoMilind has joined #openstack22:14
BK_manwhat's wrong?22:14
*** enigma has joined #openstack22:14
*** brd_from_italy has quit IRC22:14
notmynamesam_itx: sorry for the delay. didn't see your response22:15
notmynamesam_itx: you have the flow correct22:15
*** xavicampa has quit IRC22:15
notmynamesam_itx: the auth server <-> swift protocol is dependent on your auth server. if you are using the auth system that ships with swift (swauth), then it's all REST-based22:16
*** dubsquared has quit IRC22:16
notmynamesam_itx: see http://swift.openstack.org/development_auth.html for more info on writing your own auth integration with swift22:17
* BK_man .oO dumb subnet-calculator.com22:17
gholtBK_man: :)22:18
sam_itx notmyname: Thanks let me take a look at the url.22:18
BK_manI have a question on network service in nova22:18
BK_manIs that true that compute-node can offload ext traffic from instances only to one nova-network?22:19
*** kennethkalmer has quit IRC22:19
BK_manDo I need HA on nova-network?22:19
*** AXiS_SharK has quit IRC22:20
*** AXiS_SharK has joined #openstack22:23
*** joearnold has joined #openstack22:29
*** enigma has quit IRC22:29
*** MotoMilind has quit IRC22:31
*** m4dbi7 has quit IRC22:35
*** enigma has joined #openstack22:38
*** Lou_ has joined #openstack22:38
*** lvaughn has quit IRC22:39
*** katkee has quit IRC22:40
*** MotoMilind has joined #openstack22:42
*** enigma has quit IRC22:42
*** Eyk has joined #openstack22:50
*** enigma has joined #openstack22:51
*** mattray has quit IRC22:53
*** dragondm has quit IRC22:53
*** aliguori has quit IRC22:54
*** enigma has quit IRC22:55
Eykamazon is talking about a reliability of 99,999999999%/year for the stored data, is there similar value known for a normal swift setup ?22:57
EykI could not find any information for cloud files22:57
*** jkoelker has quit IRC22:58
*** infinite-scale has quit IRC23:01
*** llang629 has left #openstack23:02
*** MotoMilind has quit IRC23:03
*** MotoMilind has joined #openstack23:09
*** rdw has left #openstack23:10
uvirtbotNew bug: #782437 in glance "API should respect the mimetype in Content-Type and Accept headers" [Undecided,New] https://launchpad.net/bugs/78243723:11
*** dobber_ has quit IRC23:16
*** nerens has quit IRC23:18
JSharpvishy, sorry to bug you again, but you said earlier that (<vishy> -A nova-network-POSTROUTING -s 10.252.0.0/16 -d 10.128.0.0/24 -j ACCEPT) could cause problems. Can you explain why?23:19
*** Lou_ has quit IRC23:19
vishysorry i misread it23:19
vishyi thought it was 10.128.0.0/1623:19
vishyer /923:19
vishywhich would mean that 10.252 and 10.254 would hit that rule23:20
*** r5 has quit IRC23:25
BK_manvishy: is there a way to run nova-network in HA?23:30
vishyBK_man: probably not the way you are thinking23:30
vishyMK_man: you can run with failover23:31
vishyBK_man: ^^23:31
BK_manvishy: ok. can nova-compute talk to several or just one nova-network?23:31
vishynot several at the same time23:31
BK_manvishy: I don't want to loose ext connectivity in case if nova-network fails.23:31
vishyyeah that would be failover, it takes 3-5 seconds23:32
BK_manvishy: ok. BTW, how can I specify for particular compute to use different nova-network?23:32
* BK_man tried to run 200 instances using 1 API call and got 4% percent failure rate :)23:32
vishyBK_man: whith ec2 api?23:33
BK_manvishy: yep23:33
vishyhow many computes?23:33
vishyI've done 100 with no failures on 4 computes23:33
BK_manvishy: 3 UCS blades23:33
vishyany error for the failures?23:33
BK_manvishy: yep. Socket failure while talking to Rabbit23:33
vishyah same thing that comstud was running into23:34
vishythere is a patched version of eventlet that may help with that23:34
vishyshould be in the ppa soon23:34
*** shentonfreude has joined #openstack23:34
BK_manvishy: is that a new fix? I can remember that I patched this also23:34
vishyalso comstud is looking at some refactoring to help there23:34
vishybrand new23:35
vishy:)23:35
BK_manok. do you have a patch on github/another site?23:35
vishyit is on bitbucket and also attached to our bug23:35
BK_manvishy: you know, I'm RHEL man with my own repo23:35
BK_manvishy: BTW, what is XXX meaning in nova-manage service list output?23:36
vishycan you see if this helps? https://bugs.launchpad.net/nova/+bug/77151223:36
uvirtbotLaunchpad bug 771512 in nova "Timeout from API with 50 Simultaneous Builds" [High,Triaged]23:36
vishyBK_man: that means that the service hasn't reported in recently23:36
vishymost of the time it is because of time skew23:37
vishyinstall ntp on all of your hosts23:37
BK_manvishy: good point, I'll check the time23:37
*** infinite-scale has joined #openstack23:39
uvirtbotNew bug: #782446 in swift "st large object upload should retry segments" [Undecided,New] https://launchpad.net/bugs/78244623:41
*** lvaughn has joined #openstack23:42
*** infinite-scale has quit IRC23:44
*** cloudgroups has joined #openstack23:44
*** shentonfreude has quit IRC23:47
BK_manvishy: NTP sync solved XXX problem23:47
devcamcari actually wrote code today!23:48
devcamcarwe're going to be adding swift support to the dashboard23:48
*** AXiS_SharK has left #openstack23:48
BK_manvishy: do I need to configure something to allow failover for nova-network?23:53
*** ohnoimdead has joined #openstack23:55
* BK_man got nova-network not starting up at all...23:55
BK_manAttempting to grab file lock "iptables" for method "apply"...23:55
BK_manlast line in log...23:55
sam_itxwhere can I find some information about scalability (study/usage) of openstack ? # of compute nodes, # of VMs, etc. ? # of accounts, etc.23:57
* BK_man found stale locks in /var/lib/nova/tmp23:57
BK_mansam_itx: I think the only way to go for today is to do it by youself23:58
sam_itxThanks.. from the conversations, it seems you are doing that with UCS. :)23:59

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