*** sacharya1 has joined #openstack | 00:01 | |
*** sacharya1 has quit IRC | 00:01 | |
*** sacharya has quit IRC | 00:01 | |
*** sacharya has joined #openstack | 00:01 | |
*** dirakx has joined #openstack | 00:06 | |
*** dtroyer_zzz is now known as dtroyer | 00:06 | |
*** sacharya has quit IRC | 00:10 | |
*** eglynn has quit IRC | 00:11 | |
*** justinsb has joined #openstack | 00:13 | |
*** rkukura has quit IRC | 00:15 | |
*** koolhead17 has quit IRC | 00:17 | |
*** ShKoDrAnI has quit IRC | 00:22 | |
*** ShKoDrAnI has joined #openstack | 00:24 | |
*** sacharya has joined #openstack | 00:26 | |
*** tmichael has quit IRC | 00:28 | |
*** kaz__ has quit IRC | 00:37 | |
*** kaz_ has joined #openstack | 00:44 | |
*** Ryan_Lane has joined #openstack | 00:44 | |
nvez | Is there such thing as Superadmin in keystone (ex: superadmin Nova account) | 00:50 |
---|---|---|
*** dtroyer is now known as dtroyer_zzz | 00:54 | |
*** jamespage_ has joined #openstack | 00:57 | |
*** jamespage_ has quit IRC | 00:57 | |
*** rnorwood has joined #openstack | 01:09 | |
nvez | zykes- / CristianDM: ran into this while just looking for something: http://glance.openstack.org/configuring.html#configuring-the-rbd-storage-backend | 01:11 |
*** martine has quit IRC | 01:13 | |
*** dcramer_ has quit IRC | 01:19 | |
*** CaptTofu1 has joined #openstack | 01:19 | |
*** dwcramer has quit IRC | 01:20 | |
*** CaptTofu has quit IRC | 01:20 | |
*** rnorwood has quit IRC | 01:26 | |
*** CaptTofu1 has quit IRC | 01:28 | |
*** CaptTofu has joined #openstack | 01:30 | |
*** erkules has quit IRC | 01:31 | |
*** erkules has joined #openstack | 01:31 | |
*** agonellaa has quit IRC | 01:32 | |
*** dcramer_ has joined #openstack | 01:33 | |
*** CaptTofu has quit IRC | 01:33 | |
*** dwcramer has joined #openstack | 01:33 | |
*** livemoon has joined #openstack | 01:35 | |
*** martine has joined #openstack | 01:36 | |
*** martine has quit IRC | 01:38 | |
*** ejat has quit IRC | 01:40 | |
*** ejat has joined #openstack | 01:45 | |
*** ejat has joined #openstack | 01:45 | |
*** CaptTofu has joined #openstack | 01:48 | |
*** dirakx has left #openstack | 01:49 | |
*** asadpanda has quit IRC | 01:53 | |
*** e1mer has joined #openstack | 01:54 | |
*** darkhorse has quit IRC | 01:54 | |
uvirtbot | New bug: #989516 in openstack-manuals "Inconsistent service description for Object Storage service" [Undecided,Fix committed] https://launchpad.net/bugs/989516 | 01:55 |
creiht | davidha: Re: apache etc, the first question that comes to my mind is what problem are you trying to solve? | 01:58 |
*** wariola has joined #openstack | 01:58 | |
creiht | perhaps once we have that identified, then we can have more fruitful discussions on these topics? | 01:59 |
*** livemoon_ has joined #openstack | 02:09 | |
*** livemoon has quit IRC | 02:09 | |
*** livemoon_ is now known as livemoon | 02:09 | |
*** sniperd has joined #openstack | 02:12 | |
*** RicardoSSP has joined #openstack | 02:12 | |
*** dachary1 has joined #openstack | 02:13 | |
*** dachary has quit IRC | 02:13 | |
*** littleidea has joined #openstack | 02:15 | |
*** ryanpetr_ has joined #openstack | 02:17 | |
*** never2far has quit IRC | 02:19 | |
*** sdake has joined #openstack | 02:20 | |
*** ryanpetrello has quit IRC | 02:20 | |
*** RicardoSSP has quit IRC | 02:23 | |
*** osier has joined #openstack | 02:25 | |
nvez | creiht: I don't have an objection against eventlet but some people might feel more comfortable operating Apache as the HTTP frontend | 02:25 |
nvez | also, i would argue to some people feel that swift should take care of the object storage and let the user choose the http server to serve the content, in this case, apache or anything that can deal with wsgi | 02:27 |
*** ryanpetr_ has quit IRC | 02:27 | |
nvez | i do see a point that apache is really good at being an http server, so let it serve http and have swift be really good at operating an object storage system | 02:27 |
*** ShKoDrAnI has quit IRC | 02:28 | |
*** ShKoDrAnI has joined #openstack | 02:29 | |
creiht | nvez: There is nothing that prevents people from putting apache in front of swift in a reverse proxy setup, what I'm trying to get at is what that really buys them (other than familiarity) | 02:30 |
creiht | If there is a gap, then there is a problem for us to solve :) | 02:31 |
creiht | even if you put something in front of swift, you still have to maintain a whole swift stack, which doesn't make it any easier | 02:31 |
nvez | creiht: I made that exact suggestion which I agree with you, should help cover those odd requirements people have and it should be enough | 02:31 |
*** hvb has quit IRC | 02:31 | |
nvez | but I wonder if there is a way of having Apache serve files to lower the usage of swift itself | 02:32 |
creiht | lower which usage? | 02:32 |
nvez | aka, auth/get file and then once it's there, have apache serve it and it'll free up resources for swift, but again, i'm not that heavy of a user to require that | 02:32 |
*** hvb has joined #openstack | 02:33 | |
creiht | being able to get a file requires a certain amount of smarts to know where those objects are | 02:33 |
creiht | a majority of the smarts are in the proxy servers | 02:33 |
nvez | ah, see, i dont know that much about the swift infrastructure so in that case it can't be a straightforward implementation | 02:34 |
creiht | it turns out that actual server http is one of the smallest problems of swift right now :) | 02:34 |
creiht | serving http | 02:34 |
nvez | anyhow, i think that until someone hits a scalability issue with it (because any apache requirement can be fixed by reverse proxying), it probably won't be an issue | 02:34 |
creiht | in our production environment the number of nodes serving http a orders of magnitude fewer than the number of storage nodes | 02:35 |
creiht | if people are interested in helping solve real scalability issues with swift, there are plenty of hard problems to solve | 02:36 |
creiht | like better object/container replication | 02:36 |
*** ywu has quit IRC | 02:36 | |
nvez | agreed | 02:36 |
creiht | container sharding, or better handling of super large containers (other than throwing better hardware at it) | 02:36 |
creiht | dc to dc replication | 02:36 |
creiht | things like that :) | 02:36 |
creiht | http serving throughput is very low on my list :) | 02:37 |
creiht | that reminds me, that I need to write a couple of blog posts | 02:37 |
nvez | i see. you work mostly with swift? | 02:37 |
*** hugokuo has quit IRC | 02:37 | |
creiht | I helped write it :) | 02:37 |
creiht | I haven't been able to work on it as much lately as I have been working on another project | 02:38 |
creiht | but I still keep close tabs on it, and try to help where I can | 02:38 |
nvez | that's awesome, haha I hope you bring your distribution knowledge for distributed block storage for nova | 02:38 |
nvez | :p | 02:38 |
creiht | hah | 02:38 |
creiht | I'm actually working on block storage :) | 02:39 |
creiht | but not distributed | 02:39 |
*** sniperd_ has joined #openstack | 02:39 | |
nvez | aww | 02:39 |
creiht | that's a tougher nut to crack | 02:39 |
creiht | :) | 02:39 |
*** sniperd has quit IRC | 02:39 | |
*** dtroyer_zzz is now known as dtroyer | 02:39 | |
nvez | ceph+rbd sounds really "meh"… according to readings sheepdog doesnt offer that good performance | 02:39 |
nvez | if only someone just did that, would be so awesome, makes everything perfect! heh | 02:40 |
creiht | hah | 02:40 |
creiht | In our limited testing, we couldn't get ceph to perform very well | 02:40 |
creiht | but some people swear by it | 02:40 |
nvez | i havent done much reading but there doesnt seem to be a lot of documentation / users | 02:41 |
creiht | At the ceph meetup with piston, I remember someone saying that they got an order of magnitude lesser performance with ceph than raw disk | 02:41 |
nvez | that's not good :( | 02:42 |
creiht | but I don't want to disparage ceph | 02:42 |
creiht | you should try it and form your own opinions :) | 02:42 |
nvez | something designed for nova where it replicates it n times on servers and openstack should try to boot on servers that host it locally only | 02:43 |
nvez | the only problem I would think is just writes being as slow as the connection between the servers | 02:43 |
*** johnpur has joined #openstack | 02:44 | |
*** ChanServ sets mode: +v johnpur | 02:44 | |
creiht | One of the bigger problems with distributed block storage in my mind is consistency. It really requires strict consistency, which in turn adds a lot of constraints (and usually perf degradation) | 02:45 |
nvez | but i mean like, we have raid (and even software raid) over local drives, if only just have that over network + locking mechanisms | 02:47 |
creiht | if only :) | 02:47 |
nvez | heh | 02:47 |
nvez | im working a lot with nova and the only thing now is just finding the proper docs, there is a lot but some of stuff arent documented | 02:48 |
creiht | We may try to tackle the distributed block storage at some point, but for now I'm taking a more pragmatic, simple approach, and seeing how far that gets you | 02:49 |
nvez | ex: keystone + adding a new user to a tenant, i know it's possible because the dashboard does it, but no docs, so gotta reference the libs and my python stinks, heh | 02:50 |
creiht | yay open source :) | 02:50 |
nvez | and that's awesome, it would be great to see that, it would be an awesome feature | 02:50 |
nvez | what surprises me (and might be scary) that not even paid products offer distributed block storage | 02:50 |
creiht | at least the docs are always getting better | 02:50 |
creiht | hehe | 02:51 |
creiht | unfortunately I can't help with keystone, but ask the right question to the right people and you will find what you need | 02:53 |
creiht | you migh also look at how the devstack sets up keystone | 02:53 |
creiht | that might provide some insight | 02:53 |
creiht | good luck! :) | 02:53 |
nvez | creiht: i do use devstack and i've actually had good luck just reading the code :p | 02:56 |
nvez | i think i found the isue what my problem was and its all good now | 02:56 |
*** mdomsch has joined #openstack | 02:57 | |
nvez | though it would be cool if this chanel was a bit more active, i've been running to some really silly issues | 02:57 |
nvez | cant delete tenant cause user assigned to it, cant delete user cause tenant assigned to it, and user not being unassigned from tenant.. really silly lol | 02:57 |
creiht | nvez: you might also follow #openstack-dev | 02:57 |
creiht | there are usually more crunchy questions in there | 02:58 |
*** ryanpetrello has joined #openstack | 02:59 | |
*** bharata has joined #openstack | 02:59 | |
uvirtbot | New bug: #994957 in openstack-common "handle all mailmap with name and email address" [Low,In progress] https://launchpad.net/bugs/994957 | 03:01 |
*** paltman has quit IRC | 03:09 | |
*** paltman has joined #openstack | 03:10 | |
*** never2far has joined #openstack | 03:15 | |
*** hugokuo has joined #openstack | 03:21 | |
*** joearnold has joined #openstack | 03:21 | |
*** ljj-justin has joined #openstack | 03:21 | |
*** ljj-justin has left #openstack | 03:24 | |
uvirtbot | New bug: #995727 in openstack-manuals "openstack-api-site in openstack-manuals fails to build" [Undecided,New] https://launchpad.net/bugs/995727 | 03:26 |
*** esteimann_zz is now known as esteimann | 03:32 | |
*** guigui1 has joined #openstack | 03:32 | |
*** cryptk|offline is now known as cryptk | 03:33 | |
*** Karmaon has joined #openstack | 03:34 | |
*** CaptTofu has quit IRC | 03:35 | |
*** esteimann is now known as esteimann_zz | 03:37 | |
*** guigui1 has quit IRC | 03:40 | |
*** nigwil has joined #openstack | 03:40 | |
*** fifieldt has quit IRC | 03:42 | |
*** fifieldt has joined #openstack | 03:43 | |
*** tuf8 has quit IRC | 03:46 | |
*** jackh has joined #openstack | 03:58 | |
*** nigwil has quit IRC | 04:01 | |
*** egant has joined #openstack | 04:03 | |
*** nigwil has joined #openstack | 04:03 | |
*** nigwil has quit IRC | 04:09 | |
*** egant has quit IRC | 04:10 | |
*** wilmoore has quit IRC | 04:13 | |
*** sdake has quit IRC | 04:13 | |
*** ShKoDrAnI has quit IRC | 04:17 | |
*** ShKoDrAnI has joined #openstack | 04:18 | |
*** ShKoDrAnI has quit IRC | 04:21 | |
*** ShKoDrAnI has joined #openstack | 04:22 | |
*** sandywalsh_ has quit IRC | 04:37 | |
*** issackelly has quit IRC | 04:41 | |
uvirtbot | New bug: #994758 in quantum "xen installer crashes on quantum agent installation" [Undecided,In progress] https://launchpad.net/bugs/994758 | 04:42 |
uvirtbot | New bug: #994774 in quantum "ovs_quantum_agent.py crashes on Citrix XCP because of old python version (2.4)" [Undecided,In progress] https://launchpad.net/bugs/994774 | 04:42 |
uvirtbot | New bug: #994831 in quantum "ovs_quantum_plugin incompatibility with python 2.4" [Undecided,In progress] https://launchpad.net/bugs/994831 | 04:42 |
*** masterhumper is now known as sejo | 04:46 | |
*** garyk has quit IRC | 04:48 | |
*** troytoman is now known as troytoman-away | 04:50 | |
*** littleidea has quit IRC | 04:51 | |
*** wilmoore has joined #openstack | 04:52 | |
*** sniperd_ has quit IRC | 04:54 | |
*** mindpixel has joined #openstack | 04:54 | |
*** clopez has quit IRC | 04:55 | |
*** egant has joined #openstack | 04:57 | |
*** koolhead17 has joined #openstack | 05:01 | |
*** dwcramer has quit IRC | 05:02 | |
*** dendro-afk is now known as dendrobates | 05:02 | |
*** dcramer_ has quit IRC | 05:03 | |
*** Blackavar has quit IRC | 05:03 | |
*** nvez has quit IRC | 05:06 | |
*** ejat has quit IRC | 05:07 | |
*** Blackavar has joined #openstack | 05:08 | |
*** rmt has joined #openstack | 05:12 | |
*** ryanpetrello has quit IRC | 05:12 | |
*** exekias has quit IRC | 05:20 | |
*** mindpixel has quit IRC | 05:26 | |
*** dtroyer is now known as dtroyer_zzz | 05:28 | |
*** ShKoDrAnI has quit IRC | 05:28 | |
*** saju has joined #openstack | 05:31 | |
saju | could you please share a doc whic describimh networking in openstack essex. | 05:32 |
saju | could you please share a doc which describe network configuration (All mode) in openstack essex. ? | 05:33 |
*** ShKoDrAnI has joined #openstack | 05:34 | |
*** PiotrSikora has quit IRC | 05:36 | |
*** PiotrSikora has joined #openstack | 05:38 | |
*** dendrobates is now known as dendro-afk | 05:41 | |
*** ShKoDrAnI has quit IRC | 05:44 | |
*** kodapa has quit IRC | 05:44 | |
*** ShKoDrAnI has joined #openstack | 05:45 | |
*** hvb has quit IRC | 05:51 | |
*** CaptTofu has joined #openstack | 05:54 | |
*** garyk has joined #openstack | 05:59 | |
*** MarkAtwood has quit IRC | 06:00 | |
*** hattwick has quit IRC | 06:03 | |
*** aa__ has joined #openstack | 06:06 | |
*** andresambrois has quit IRC | 06:07 | |
*** bjoernh has joined #openstack | 06:11 | |
*** dachary1 has quit IRC | 06:13 | |
*** shang_ has quit IRC | 06:15 | |
*** davidha has quit IRC | 06:15 | |
*** davidha has joined #openstack | 06:18 | |
*** rendar has joined #openstack | 06:22 | |
*** GheAway is now known as GheRivero | 06:25 | |
GheRivero | morning people | 06:25 |
*** davepigott has joined #openstack | 06:25 | |
*** Pebbles` has quit IRC | 06:36 | |
*** Pebbles` has joined #openstack | 06:36 | |
uvirtbot | New bug: #994652 in quantum "nova-scheduler crashes with the cisco Quantum scheduler" [Undecided,In progress] https://launchpad.net/bugs/994652 | 06:46 |
*** kpp has joined #openstack | 06:48 | |
*** davidha has quit IRC | 06:49 | |
*** GheRivero has quit IRC | 06:51 | |
*** krow has quit IRC | 06:53 | |
*** dnaori has joined #openstack | 06:54 | |
*** nigwil has joined #openstack | 06:56 | |
*** nigwil has quit IRC | 06:57 | |
*** paulmillar has joined #openstack | 06:58 | |
*** nigwil has joined #openstack | 06:58 | |
*** hvb has joined #openstack | 06:59 | |
*** davepigott has quit IRC | 06:59 | |
*** prakasha-log has quit IRC | 07:00 | |
*** prakasha-log has joined #openstack | 07:00 | |
*** nigwil has quit IRC | 07:01 | |
*** nigwil has joined #openstack | 07:02 | |
*** fifieldt has quit IRC | 07:02 | |
*** Stackops-Jorge has joined #openstack | 07:04 | |
*** GheRivero has joined #openstack | 07:06 | |
*** kpp has quit IRC | 07:07 | |
*** pnavarro has joined #openstack | 07:07 | |
*** journeeman has joined #openstack | 07:10 | |
*** azbarcea has joined #openstack | 07:14 | |
*** anderstj has quit IRC | 07:16 | |
*** Triade has joined #openstack | 07:19 | |
*** kokyu has quit IRC | 07:21 | |
*** dachary has joined #openstack | 07:22 | |
*** asavu has joined #openstack | 07:24 | |
*** kodapa has joined #openstack | 07:31 | |
*** CaptTofu1 has joined #openstack | 07:32 | |
*** willaerk has joined #openstack | 07:33 | |
*** CaptTofu has quit IRC | 07:34 | |
*** pmezard has joined #openstack | 07:34 | |
*** Neptu has joined #openstack | 07:35 | |
*** stewart has quit IRC | 07:36 | |
*** Free_maN has joined #openstack | 07:38 | |
*** Free_maN has joined #openstack | 07:38 | |
*** joearnold has quit IRC | 07:43 | |
*** stewart has joined #openstack | 07:48 | |
*** asavu has quit IRC | 07:51 | |
*** dieterdemeyer has joined #openstack | 07:53 | |
*** aa__ has quit IRC | 07:57 | |
Darksecond | if i use vlan networking in nova, i only need 1 nic in the machine that runs nova-network, right? | 07:58 |
*** hattwick has joined #openstack | 07:59 | |
*** rendar has quit IRC | 08:01 | |
uvirtbot | New bug: #995811 in keystone "updated prettytable api" [Undecided,New] https://launchpad.net/bugs/995811 | 08:01 |
*** _et has quit IRC | 08:02 | |
*** wilmoore has quit IRC | 08:02 | |
*** zeriouz has joined #openstack | 08:02 | |
*** rendar has joined #openstack | 08:03 | |
*** davidha has joined #openstack | 08:06 | |
*** krow has joined #openstack | 08:10 | |
*** lianhao has joined #openstack | 08:10 | |
uvirtbot | New bug: #995818 in nova "Prettytable updated with new api" [Undecided,New] https://launchpad.net/bugs/995818 | 08:11 |
*** miclorb has quit IRC | 08:13 | |
*** ShKoDrAnI has quit IRC | 08:13 | |
*** ShKoDrAnI has joined #openstack | 08:14 | |
*** nacx has joined #openstack | 08:16 | |
*** lianhao has left #openstack | 08:17 | |
*** kokyu has joined #openstack | 08:19 | |
kokyu | morning (: | 08:19 |
*** Razique has joined #openstack | 08:23 | |
Razique | morning :) | 08:25 |
uvirtbot | New bug: #995826 in glance "Prettytable updated api" [Undecided,New] https://launchpad.net/bugs/995826 | 08:25 |
*** clopez has joined #openstack | 08:35 | |
*** VivekVC has joined #openstack | 08:52 | |
kokyu | the nova-scheduler (/network/cert) only need to run on the controller node, in a basic multi-node setup, isn't it? | 08:55 |
*** davidha has quit IRC | 08:55 | |
*** asavu has joined #openstack | 08:56 | |
*** Fu4ny has joined #openstack | 08:58 | |
*** krow has quit IRC | 08:58 | |
*** davidha has joined #openstack | 09:04 | |
*** livemoon has quit IRC | 09:07 | |
*** mohits has joined #openstack | 09:18 | |
*** Ryan_Lane has quit IRC | 09:18 | |
*** Trixboxer has joined #openstack | 09:19 | |
*** davepigott has joined #openstack | 09:27 | |
*** _et has joined #openstack | 09:39 | |
_et | hi all... | 09:40 |
_et | wondering if anyone else is having issues with mysql.. | 09:40 |
_et | acting all crazy.. it won't let users access db even after setting up appropriate privileges. | 09:41 |
*** eglynn has joined #openstack | 09:45 | |
*** jackh has quit IRC | 09:47 | |
*** longhudou has joined #openstack | 09:51 | |
*** ik04ka has joined #openstack | 09:52 | |
*** ipl31_ has quit IRC | 09:53 | |
*** miclorb has joined #openstack | 09:55 | |
*** anniec72 has joined #openstack | 10:08 | |
*** ondergetekende has joined #openstack | 10:10 | |
*** wariola has quit IRC | 10:13 | |
*** never2far has quit IRC | 10:13 | |
*** rkukura has joined #openstack | 10:21 | |
*** lts has joined #openstack | 10:26 | |
*** wariola has joined #openstack | 10:27 | |
*** mhzarei has joined #openstack | 10:29 | |
uvirtbot | New bug: #989746 in nova "'FlatManager' object has no attribute 'allocate_floating_ip'" [Undecided,New] https://launchpad.net/bugs/989746 | 10:30 |
*** exekias has joined #openstack | 10:31 | |
Darksecond | if i use vlan networking in nova, i only need 1 nic in the machine that runs nova-network, right? | 10:32 |
*** mjfork has joined #openstack | 10:32 | |
rmt | Darksecond, correct. VLAN tagging just sticks a couple more bytes in the ethernet frame, and it all goes out via the same card. | 10:34 |
*** mhzarei has quit IRC | 10:34 | |
Darksecond | rmt: very nice, since my testing system only has one network card, and i want to tinker with openstack and the vlan networking mode :) | 10:35 |
*** davepigott has quit IRC | 10:35 | |
Darksecond | i'm also looking into openflow and openvswitch, but currently all it's doing for me is giving me a headache :D | 10:37 |
*** ttrifonov is now known as ttrifonov_zZzz | 10:37 | |
*** ttrifonov_zZzz is now known as ttrifonov | 10:37 | |
*** Rrocky has quit IRC | 10:39 | |
*** Rrocky has joined #openstack | 10:40 | |
Darksecond | or would it be better to use flatdhcp? | 10:44 |
*** CaptTofu has joined #openstack | 10:45 | |
*** CaptTofu1 has quit IRC | 10:46 | |
Darksecond | aaah, i think i understand, flatDHCP isn't less secure than vlan, it uses security groups and iptables to make it just as secure | 10:48 |
Darksecond | now i get it | 10:48 |
*** miclorb has quit IRC | 10:49 | |
*** mohits has quit IRC | 10:50 | |
*** ShKoDrAnI has quit IRC | 10:50 | |
*** ShKoDrAnI has joined #openstack | 10:51 | |
*** mhzarei has joined #openstack | 10:51 | |
*** bharata has quit IRC | 10:51 | |
*** gstaicu has joined #openstack | 10:51 | |
*** asavu has quit IRC | 10:55 | |
*** never2far has joined #openstack | 11:03 | |
*** ondergetekende has quit IRC | 11:05 | |
_val_ | ping koolhead17 | 11:05 |
*** wariola has quit IRC | 11:06 | |
*** never2far has quit IRC | 11:07 | |
*** asavu has joined #openstack | 11:09 | |
*** twayneprice has quit IRC | 11:10 | |
*** flaviamissi has quit IRC | 11:11 | |
uvirtbot | New bug: #995283 in quantum "Documentation build is broken" [Undecided,In progress] https://launchpad.net/bugs/995283 | 11:15 |
_val_ | Oh tell me something new. | 11:16 |
_val_ | A lot of bugs. | 11:16 |
*** rmartinelli has quit IRC | 11:16 | |
*** rkukura has quit IRC | 11:17 | |
Darksecond | if i want to use flatdhcp, can i use 2 vlan (virtual) adapters instead of real nics? | 11:27 |
*** rmt has quit IRC | 11:33 | |
*** renier has joined #openstack | 11:34 | |
*** never2far has joined #openstack | 11:34 | |
*** kiranmurari has joined #openstack | 11:40 | |
*** miclorb has joined #openstack | 11:44 | |
*** twayneprice has joined #openstack | 11:49 | |
*** bsza has joined #openstack | 11:49 | |
*** mohits has joined #openstack | 11:50 | |
*** hvb has quit IRC | 11:52 | |
*** journeeman has quit IRC | 11:52 | |
*** ayoung_ has joined #openstack | 11:52 | |
*** salgado has joined #openstack | 11:55 | |
*** adalbas has joined #openstack | 11:56 | |
*** hvb has joined #openstack | 11:58 | |
*** anniec72 has quit IRC | 11:58 | |
*** davidha has quit IRC | 12:07 | |
*** dmojoryder has left #openstack | 12:08 | |
*** davidha has joined #openstack | 12:08 | |
*** bjoernh has quit IRC | 12:08 | |
*** markvoelker has joined #openstack | 12:10 | |
Darksecond | if i want to use flatdhcp, can i use 2 vlan (virtual) adapters instead of real nics? | 12:12 |
*** rkukura has joined #openstack | 12:17 | |
*** Leseb has joined #openstack | 12:20 | |
*** Leseb has joined #openstack | 12:21 | |
*** aspiers has quit IRC | 12:23 | |
*** paltman has quit IRC | 12:25 | |
*** paltman has joined #openstack | 12:25 | |
*** sdake has joined #openstack | 12:28 | |
uvirtbot | New bug: #994481 in keystone "admin login in pam backend fails" [Undecided,In progress] https://launchpad.net/bugs/994481 | 12:31 |
*** dwcramer has joined #openstack | 12:31 | |
*** never2far has quit IRC | 12:31 | |
*** dcramer_ has joined #openstack | 12:33 | |
*** h0cin has joined #openstack | 12:39 | |
*** esm_ has joined #openstack | 12:39 | |
*** dtroyer_zzz is now known as dtroyer | 12:41 | |
*** paulmillar has quit IRC | 12:49 | |
*** paulmillar has joined #openstack | 12:50 | |
*** miclorb has quit IRC | 12:55 | |
*** aliguori has joined #openstack | 12:55 | |
*** Susanne-Balle has joined #openstack | 12:56 | |
*** davidha has quit IRC | 12:56 | |
*** tryggvil_ has joined #openstack | 12:58 | |
*** junaid has joined #openstack | 12:58 | |
*** gakott has joined #openstack | 12:59 | |
junaid | I have configured keystone on my machine and it is up and running. but when I run 'keystone --token ADMIN --endpoint http://localhost:35357/v2.0 tenant-list' I get an error | 12:59 |
junaid | OperationalError: (OperationalError) no such table: tenant u'SELECT tenant.id AS tenant_id, tenant.name AS tenant_name, tenant.extra AS tenant_extra \nFROM tenant' () | 12:59 |
*** DavidLevin has quit IRC | 13:00 | |
*** flaviamissi has joined #openstack | 13:00 | |
junaid | can any one guide me in fixing this problem? | 13:00 |
*** garyk has quit IRC | 13:00 | |
*** jobicoppola has quit IRC | 13:01 | |
*** aliguori has quit IRC | 13:02 | |
*** koolhead17 has quit IRC | 13:02 | |
*** DavidLevin has joined #openstack | 13:03 | |
*** jsavak has joined #openstack | 13:04 | |
*** rmartinelli has joined #openstack | 13:05 | |
*** Leseb has quit IRC | 13:05 | |
Darksecond | if i use quantum and make an integration bridge, how can i attach eth0, but then vlan'ed? will this even work? | 13:07 |
*** e1mer has quit IRC | 13:07 | |
*** dprince has joined #openstack | 13:09 | |
*** tomjones has joined #openstack | 13:09 | |
*** tomjones has quit IRC | 13:11 | |
*** hvb has quit IRC | 13:11 | |
*** trevman2001 has joined #openstack | 13:12 | |
*** e1mer has joined #openstack | 13:13 | |
*** e1mer has joined #openstack | 13:13 | |
*** _rfz has quit IRC | 13:16 | |
*** nvez has joined #openstack | 13:18 | |
*** never2far has joined #openstack | 13:20 | |
*** cctrieloff has joined #openstack | 13:21 | |
*** blamar has joined #openstack | 13:21 | |
*** ayoung_ has quit IRC | 13:22 | |
*** ayoung has quit IRC | 13:25 | |
*** markmcclain has joined #openstack | 13:26 | |
*** cryptk is now known as cryptk|offline | 13:26 | |
*** johnpur has quit IRC | 13:27 | |
*** Glace has joined #openstack | 13:27 | |
*** Leseb has joined #openstack | 13:27 | |
*** tomjones has joined #openstack | 13:28 | |
*** jcannava has joined #openstack | 13:31 | |
sstent | Anyone know why i would see "The action you have requested has not been implemented. (HTTP 501)" when i do anything relating to endpoints in keystone? e.g. endpoint-list or endpoint-create | 13:32 |
*** avoine has joined #openstack | 13:33 | |
*** avoine has left #openstack | 13:34 | |
kokyu | I added 2 networks (via nova-manage) and associated them with 3 different projects. | 13:34 |
nvez | Are you using a static backend for the catalog, sstent ? | 13:35 |
kokyu | then I started a new instance in project (A), but now I see that it got assigned all three IPs (in the dashbaord) but I cannot log in via either IP | 13:35 |
sstent | nvez: yeah mysql | 13:36 |
nvez | Could you double check the keystone.conf file to make sure it is using sql? | 13:37 |
*** lazyshot has joined #openstack | 13:37 | |
*** avoine has joined #openstack | 13:37 | |
*** sacharya1 has joined #openstack | 13:38 | |
*** sacharya has quit IRC | 13:38 | |
*** mdomsch has quit IRC | 13:38 | |
gstaicu | juanid after you have installed keystone did you make keystone-manage db_sync? | 13:39 |
junaid | gstaicu: yes | 13:40 |
sstent | nvez: yup using mysql -- connection = mysql://user:pass@127.0.0.1/keystone | 13:40 |
sstent | nvez: i can create services etc, jsut not endpoints | 13:40 |
nvez | ah, I don't know in that case.. I just recently started tinkering around with Keystone and it's errors are pretty cryptic sometimes :p | 13:41 |
gstaicu | but using mysql | 13:41 |
*** Lumiere has joined #openstack | 13:42 | |
gstaicu | can you check your database keystone? | 13:42 |
junaid | gstaicu: I see in the code that it calls create_engine(CONF.sql.connenction, ...), which in our case is sqlite://keystone.db | 13:42 |
junaid | gstaicu: how do I connect to this database? | 13:42 |
nvez | sqlite3 keystone.db in shell | 13:43 |
*** GheRivero_ has joined #openstack | 13:44 | |
*** sandywalsh has joined #openstack | 13:44 | |
*** KavanS has joined #openstack | 13:44 | |
*** jgriffith has joined #openstack | 13:45 | |
junaid | nvez: sqlite keystone.db and then sqlite> .table returns zero output | 13:46 |
*** krow has joined #openstack | 13:46 | |
*** kbringard has joined #openstack | 13:46 | |
nvez | junaid: looks like your db is empty, follow what gstaicu mentioned and run the db_sync, it should build your db (i think) | 13:46 |
*** lazyshot has quit IRC | 13:47 | |
*** saju has quit IRC | 13:50 | |
*** sandywalsh_ has joined #openstack | 13:51 | |
*** sandywalsh has quit IRC | 13:51 | |
*** tryggvil_ has quit IRC | 13:55 | |
*** lazyshot has joined #openstack | 13:55 | |
*** littleidea has joined #openstack | 13:58 | |
*** davidha has joined #openstack | 14:00 | |
*** FlorianOtel has quit IRC | 14:00 | |
Razique | hi guys i'm validating Essex install | 14:01 |
Razique | when I want to list images via glance index, i've the message "Response from Keystone does not contain a Glance endpoint." | 14:01 |
Razique | I've enabled templated catalogs in keystone | 14:01 |
Razique | and the file contains the glance endpoints | 14:02 |
*** livemoon has joined #openstack | 14:02 | |
Razique | (ubuntu 12 via repo btw) | 14:02 |
*** ryanpetrello has joined #openstack | 14:02 | |
livemoon | can anyone help it? | 14:03 |
*** hvb has joined #openstack | 14:03 | |
nvez | Razique: it does sound like according to keystone, it cant find the glance endpoint… try running an strace and looking at recvfrom calls and see the JSON that keystone sends (specifically the one with "access" response that includes the token) | 14:04 |
Razique | thanks nvez I'll try | 14:05 |
livemoon | Razique: what is btw? | 14:06 |
Razique | strace is a tool that monitor system calls | 14:06 |
kbringard | btw == by the way | 14:06 |
nvez | Razique: you can run this -- strace -s 4096 <glance call> | 14:07 |
livemoon | kbringard: thanks | 14:07 |
nvez | look for recvfrom with some json in it | 14:07 |
Razique | nvez : is it that output http://paste.openstack.org/show/15949/ | 14:10 |
*** edygarcia has joined #openstack | 14:11 | |
junaid | nvez: when I run keystone-manage db_sync, this command creates a database in the current directory.. But in keystone.conf the connection option has sqlite:///keystone.db... So my question is what path does this sqlite:///keystone.db resolve to? | 14:12 |
nvez | Razique: looks like you're getting an empty service catalog | 14:12 |
nvez | are you providing a tenant id? | 14:12 |
*** jsavak has left #openstack | 14:12 | |
kbringard | junaid: usually the pwd you're in when you run the command | 14:12 |
*** mhzarei has quit IRC | 14:12 | |
kbringard | I'd recommend changing that to a static path so it always gets created in the same place | 14:12 |
kbringard | or, ideally, switch it to use a mysql DB | 14:13 |
Razique | nvez : I provide an tenant name | 14:13 |
Razique | via OS_TENANT_NAME | 14:13 |
nvez | Razique: can you scroll up a bit in the strace and see if that is being sent in the token request/ | 14:13 |
junaid | abringard: how to switch to mysql DB? | 14:13 |
Razique | nvez oh wait | 14:14 |
Razique | it worked | 14:14 |
nvez | what was it? | 14:14 |
Razique | I had to update my sourceable file | 14:14 |
nvez | aha, awesome | 14:14 |
Razique | the tenant name was incorrect | 14:14 |
Razique | sorry :s | 14:14 |
Razique | catalog empty but works now | 14:14 |
nvez | np! now you know how to debug those keystone issues | 14:14 |
livemoon | that's the problem? | 14:14 |
livemoon | nvez: using strace? | 14:15 |
kbringard | junaid: in the keystone.conf, change the sql_connection = line to read mysql://user:pass@host/db_name | 14:15 |
nvez | yep that's usually the way i do it.. | 14:15 |
kbringard | then, of course, you'll need a mysql DB setup | 14:15 |
*** tong has joined #openstack | 14:15 | |
kbringard | and you'll need to have the DB created (and empty) | 14:15 |
Razique | perfect thanks a ton nvez | 14:15 |
kbringard | and users created to have access to it | 14:15 |
Razique | hey kbringard | 14:15 |
Razique | mi friend | 14:15 |
nvez | np | 14:15 |
livemoon | Razique: since of incorrect tenant_name? | 14:15 |
kbringard | 'sup Razique | 14:16 |
Razique | livemoon yup | 14:16 |
Razique | into my sourceable file | 14:16 |
Razique | the OS_TENAME_NAME was the user | 14:16 |
Razique | not the tenant | 14:16 |
*** vmlinuz has joined #openstack | 14:16 | |
Razique | I hate those demo datas | 14:16 |
livemoon | :) | 14:16 |
Razique | the generic names are too...generic :D | 14:16 |
Razique | kbringard : fine, back on openstack after a kit of kvm/ libvirt :D | 14:17 |
tong | @notmyname, john, u there? | 14:17 |
*** jackh has joined #openstack | 14:17 | |
Razique | i'm working on diablo -> essex migration | 14:17 |
kbringard | ah, I will be interested in what you come up with there | 14:17 |
livemoon | Razique: go on your work ,good luck | 14:17 |
kbringard | it's on my list of things to do, but I'm stuck in firefighting hell at the moment | 14:17 |
notmyname | tong: yay! | 14:18 |
Razique | kbringard : first impression : that will be funny, much changes meanwhile | 14:18 |
notmyname | tong: we must chat :-) | 14:18 |
livemoon | nvez: what is "glance call" in"strace -s 4096 <glance call>" | 14:18 |
zykes- | notmyname: will swift stuff be packaged into different packages now or ? | 14:18 |
zykes- | i mean like rate limit etc | 14:18 |
notmyname | zykes-: some of it. we're going through that process now. similar to what happened with slogging and swauth | 14:19 |
notmyname | tong: pm | 14:19 |
*** dtroyer is now known as dtroyer_zzz | 14:20 | |
tong | @notmyname, john, just want to see if there is any new development on the 3rd party API? | 14:20 |
notmyname | tong: yes, I sent you a pm | 14:21 |
*** rnorwood has joined #openstack | 14:21 | |
*** pnavarro has quit IRC | 14:22 | |
tong | @notmyname, sorry John, what is pm? | 14:22 |
notmyname | tong: sorry. private message (ie /msg) | 14:22 |
kokyu | what does the status-column mean at the output of `nova-manage service list`? it shows enabled/disabled. but how is that actually being evaluated ? | 14:22 |
notmyname | tong: should have opened a new channel or window for you | 14:22 |
tong | oh. email? | 14:22 |
notmyname | tong: no, in IRC | 14:22 |
notmyname | tong: for example, type "/msg notmyname hi" | 14:23 |
* Razique thought flags were no longer used in Essex, why nova.conf contains flag notations :D | 14:26 | |
*** erwanzinux has joined #openstack | 14:27 | |
*** pengyong has joined #openstack | 14:27 | |
nvez | livemoon: sorry late reply but what i meant simply the glance command you were executing after | 14:28 |
livemoon | nvez: I see, thanks | 14:29 |
junaid | abringard: keystone --token ADMIN --endpoint http://localhost:35357/v2.0 tenant-list gives output as printt, what does this mean? | 14:29 |
nvez | junaid: I ran into this bug yesterday. | 14:29 |
nvez | Using Ubuntu? | 14:29 |
nvez | pip uninstall prettytable; apt-get install python-prettytable | 14:30 |
*** kodapa_ has joined #openstack | 14:30 | |
*** cryptk|offline is now known as cryptk | 14:30 | |
erwanzinux | quit | 14:30 |
*** erwanzinux has quit IRC | 14:30 | |
junaid | rhel6.2 | 14:30 |
*** saju has joined #openstack | 14:31 | |
nvez | junaid: yum search python-prettytable -- do you find anything there? | 14:31 |
*** kodapa_ has joined #openstack | 14:31 | |
*** kodapa has quit IRC | 14:31 | |
*** kodapa_ is now known as kodapa | 14:32 | |
junaid | nvez: yeah, python-prettytable.noarch | 14:32 |
*** joearnold has joined #openstack | 14:32 | |
nvez | is it version 0.5? | 14:32 |
*** dtroyer_zzz is now known as dtroyer | 14:33 | |
*** joearnold has quit IRC | 14:33 | |
junaid | 0.5-2 | 14:33 |
nvez | Ok, so do pip uninstall prettytable and install that one | 14:33 |
*** joearnold has joined #openstack | 14:34 | |
junaid | okay.. and then yum install python-prettytable.noarch? | 14:34 |
*** dendro-afk is now known as dendrobates | 14:34 | |
nvez | yep | 14:34 |
*** cryptk is now known as cryptk|offline | 14:35 | |
junaid | nvez: yeah that solved the problem :) | 14:35 |
sstent | can someone confirm that if i have SERVICE_ENDPOINT and | 14:35 |
*** dendrobates is now known as dendro-afk | 14:35 | |
*** dendro-afk is now known as dendrobates | 14:35 | |
*** dcramer_ has quit IRC | 14:35 | |
sstent | OS_TOKEN defined, that the follwing endpoint creation command is correct " keystone endpoint-create --region RegionOne --service_id=ca301908f664445792255fd5ef07327c --publicurl=http://10.5.96.35:5000/v2.0 --internalurl=http://10.5.115.33:5000/v2.0 --adminurl=http://10.5.115.33:35357/v2.0" | 14:35 |
*** dwcramer has quit IRC | 14:35 | |
*** livemoon has quit IRC | 14:38 | |
junaid | nvez: logging off now, I will try adding new tenants tomo... thanks for the help | 14:38 |
nvez | np, gluck | 14:38 |
nvez | sstent: shouldn't it generate a servcie id for you..? | 14:39 |
*** junaid has quit IRC | 14:39 | |
nvez | er no, nevermind, have you tried looking at the source? | 14:39 |
nvez | sstent: try runnign a strace on that keystone command and see what the json that is being sent and received.. | 14:40 |
*** joearnold has quit IRC | 14:40 | |
uvirtbot | New bug: #995219 in keystone "swift proxy ValueError: invalid literal for int() with base 10: 'true' (dup-of: 988523)" [Undecided,New] https://launchpad.net/bugs/995219 | 14:41 |
sstent | looks like "{\"endpoint\": {\"adminurl\": \"http://10.5.115.33:35357/v2.0\", \"service_id\": \"ca301908f664445792255fd5ef07327c\", \"region\": \"RegionOne\", \"internalurl\": \"http://10.5.115.33:5000/v2.0\", \"publicurl\": \"http://10.5.96.35:5000/v2.0\"}" | 14:43 |
*** joearnold has joined #openstack | 14:43 | |
nvez | looks alright, and the response? | 14:43 |
*** trevman2001 has quit IRC | 14:43 | |
sstent | "{\"error\": {\"message\": \"The action you have requested has not been implemented.\", \"code\": 501, \"title\": null}}" | 14:44 |
*** dmlb2000 has quit IRC | 14:45 | |
nvez | do you see what URL it's POSTing too? | 14:45 |
uvirtbot | New bug: #991913 in horizon "Deleting a tenant where a user is a member leaves the user with a tenant link that's not existant." [Low,Confirmed] https://launchpad.net/bugs/991913 | 14:46 |
uvirtbot | New bug: #994154 in quantum "Quantum / Nova not killing DHCP properly on instance termination" [Undecided,Incomplete] https://launchpad.net/bugs/994154 | 14:46 |
uvirtbot | New bug: #995976 in keystone "PrettyTable 0.6 breaks Keystone CLI" [Undecided,New] https://launchpad.net/bugs/995976 | 14:46 |
*** cmagina_ is now known as cmagina | 14:46 | |
sstent | POST /v2.0//endpoints HTTP/1.1\r\nHost: 10.5.115.33:35357\r\nAccept-Encoding: identity\r\nContent-Length: 223\r\nx-auth-token: <token>\r\ncontent-type: application/json\r\nuser-agent: python-keystoneclient\r\n\r\n" | 14:47 |
*** dwcramer has joined #openstack | 14:47 | |
nvez | that's really odd, i really don't see why | 14:47 |
*** davidha has quit IRC | 14:47 | |
sstent | me either | 14:47 |
*** esteimann_zz is now known as esteimann | 14:48 | |
sstent | even tried going back to an sqlite db - but no luck | 14:48 |
nvez | can you show your keystone.conf ? | 14:48 |
*** hub_cap has joined #openstack | 14:48 | |
nvez | https://github.com/openstack/keystone/blob/master/keystone/catalog/backends/sql.py#L114 -- if you're using the right driver .. it should work | 14:48 |
*** dcramer_ has joined #openstack | 14:48 | |
nvez | really, the only thing I suspect is that you're using the TemplatedCatalog. | 14:49 |
Ruetobas | hi, has anyone tried libvirt + xen deployment ? | 14:50 |
sstent | http://pastebin.com/Wv2aEYpw | 14:51 |
critch | odd bit of a problem, Looking at the raw mysql db to check out some stuff. I see in the show tables output I have a projects table, but when you select on it, mysl says projects doesn't exist. | 14:51 |
*** Neptu has quit IRC | 14:51 | |
uvirtbot | New bug: #993244 in quantum "QuantumManager not creating gw-%s interfaces upon instance creation" [Undecided,Triaged] https://launchpad.net/bugs/993244 | 14:51 |
critch | Part of the interesting bit is that the dns_domains table has a foreign key pointing at the projects table. | 14:52 |
*** gstaicu has quit IRC | 14:52 | |
*** ayoung_ has joined #openstack | 14:52 | |
*** ayoung has joined #openstack | 14:52 | |
*** anderstj has joined #openstack | 14:52 | |
critch | for anyone wanting to know, my install is based on the ubuntu 12.04 install. | 14:53 |
*** MarkAtwood has joined #openstack | 14:53 | |
*** flaviamissi has quit IRC | 14:55 | |
*** anderstj has quit IRC | 14:55 | |
*** saju has quit IRC | 14:55 | |
*** rnirmal has joined #openstack | 14:55 | |
uvirtbot | New bug: #995078 in devstack "./stack.sh: line 715: cd: /opt/stack/nova: No such file or directory " [Undecided,Incomplete] https://launchpad.net/bugs/995078 | 14:56 |
uvirtbot | New bug: #995205 in devstack "devstack stack.sh error" [Medium,In progress] https://launchpad.net/bugs/995205 | 14:56 |
*** aliguori has joined #openstack | 14:56 | |
*** heyho_ has joined #openstack | 14:57 | |
*** hvb has quit IRC | 14:58 | |
*** heyho_ has quit IRC | 14:58 | |
*** heckj has joined #openstack | 15:00 | |
*** KavanS has quit IRC | 15:01 | |
*** KavanS has joined #openstack | 15:02 | |
*** jamespage_ has joined #openstack | 15:04 | |
*** jackh has quit IRC | 15:05 | |
*** livemoon has joined #openstack | 15:06 | |
*** hggdh has joined #openstack | 15:07 | |
*** zeriouz has quit IRC | 15:08 | |
*** ErwanZinux has joined #openstack | 15:09 | |
*** vds has joined #openstack | 15:10 | |
*** vds has left #openstack | 15:11 | |
*** judu has joined #openstack | 15:12 | |
judu | Hello | 15:12 |
*** jamespage_ has quit IRC | 15:12 | |
judu | I'm using openstack with libvirt, and I need the guest to know its own uuid. How do I do that? | 15:13 |
*** livemoon has left #openstack | 15:13 | |
*** livemoon has joined #openstack | 15:13 | |
*** DavidLevin has quit IRC | 15:14 | |
*** joearnold has quit IRC | 15:14 | |
*** ondergetekende has joined #openstack | 15:14 | |
kokyu | hey, I have a `nova-manage service list` output like this: http://pastie.org/private/5dybafky5qv9g4sf1hfqqa and I'd like to get rid completeley of the "colossus11" entries, but how? | 15:15 |
*** rmt has joined #openstack | 15:15 | |
kokyu | I also wonder how even they got in there :) | 15:15 |
*** jj0hns0n has joined #openstack | 15:15 | |
*** aspiers has joined #openstack | 15:15 | |
judu | (Also, does someone know When in the nova boot we get the instance uuid?) | 15:18 |
*** DavidLevin has joined #openstack | 15:18 | |
*** Blackavar has quit IRC | 15:18 | |
*** ondergetekende has quit IRC | 15:19 | |
kokyu | judu: maybe you can hack something together via the cloud-init package, that places an init script into /etc/init (/etc/init.d) to inject pubkeys at least. | 15:19 |
kokyu | judu: and if you've found the solution, tell me, because I want the same, but for the hypervisor hostname :) | 15:20 |
*** rnorwood has quit IRC | 15:20 | |
judu | kokyu: the init script is placed in the instance /etc/init.d or the hypervisor's one? | 15:21 |
*** _et has quit IRC | 15:22 | |
judu | kokyu: There might be a solution for your problem. Look into the _inject_data function. It adds things like server ssh key when creating the instance. | 15:22 |
kokyu | of the instance's one of course (: | 15:23 |
*** aliguori has quit IRC | 15:23 | |
judu | So the hypervisor hostname is easy to get and put there. | 15:23 |
kokyu | function? judu, I'm (currently) by far not a python guy. maybe this'll change once I got my cloud production ready and can think more forward then :) | 15:23 |
judu | My problem is to get the (libvirt generated?) uuid. | 15:23 |
judu | AFAIK, I will need to patch the python that creates the instance >_< | 15:24 |
ErwanZinux | quit | 15:24 |
ErwanZinux | exit | 15:24 |
*** ErwanZinux has quit IRC | 15:24 | |
*** esteimann is now known as esteimann_zz | 15:24 | |
*** imsplitbit has joined #openstack | 15:24 | |
judu | And I'm not a python guy either… | 15:25 |
kokyu | well, once you've the IP address, you could do a little hack like this: nova list | grep 10.10.40.2 | awk '{print $2}' | xargs nova show | grep ':instance_name' | awk '{print $4}' | 15:25 |
*** mael5trom has quit IRC | 15:25 | |
kokyu | works for me at least :) | 15:25 |
*** rnorwood has joined #openstack | 15:25 | |
sstent | nvez: any ideas on my config? | 15:26 |
kokyu | replace the IP with the one you get from within your instance. this trick requires the guest to have nova client installed and env vars being set | 15:26 |
nvez | ahhhh Stackops-Jorge | 15:27 |
nvez | oops! | 15:27 |
nvez | sstent | 15:27 |
nvez | driver = keystone.catalog.backends.templated.TemplatedCatalog | 15:27 |
nvez | you need to switch to the sql onw | 15:27 |
nvez | https://github.com/openstack/keystone/blob/master/etc/keystone.conf.sample#L36 | 15:28 |
*** esteimann_zz is now known as esteimann | 15:30 | |
sstent | nvez: WOOT! works now - thanks | 15:30 |
*** edygarcia_ has joined #openstack | 15:31 | |
*** dnaori has quit IRC | 15:31 | |
sstent | one final question for keystone - publicurl, is it necessary to expose this to the public network or can i jsut expose it to a secured network for our cluster admins? | 15:31 |
nvez | dont know about that technicality! | 15:33 |
*** Glace has quit IRC | 15:33 | |
*** livemoon has quit IRC | 15:33 | |
*** edygarcia has quit IRC | 15:33 | |
*** edygarcia_ is now known as edygarcia | 15:33 | |
*** hggdh has quit IRC | 15:33 | |
*** Gordonz has joined #openstack | 15:34 | |
*** Glace has joined #openstack | 15:34 | |
sstent | hmm ok. yeah mainly wiondering if endusers need access to keystone | 15:34 |
sstent | or if it's just admins/other nodes | 15:34 |
*** hggdh has joined #openstack | 15:35 | |
*** vitiho_ has quit IRC | 15:35 | |
*** tomjones has quit IRC | 15:36 | |
*** jgriffith has quit IRC | 15:38 | |
*** rmartinelli has quit IRC | 15:39 | |
*** anniec72 has joined #openstack | 15:39 | |
*** krow has quit IRC | 15:40 | |
*** hggdh has quit IRC | 15:41 | |
*** philips has quit IRC | 15:41 | |
*** ayoung has quit IRC | 15:42 | |
*** ayoung_ is now known as ayoung\ | 15:42 | |
*** ayoung\ is now known as ayoung | 15:42 | |
*** aa__ has joined #openstack | 15:42 | |
*** CaptTofu has quit IRC | 15:42 | |
*** philips has joined #openstack | 15:43 | |
*** ewindisch has joined #openstack | 15:44 | |
*** dendrobates is now known as dendro-afk | 15:44 | |
*** CaptTofu has joined #openstack | 15:45 | |
*** broussardc has joined #openstack | 15:46 | |
*** Glace has quit IRC | 15:47 | |
*** mikal has quit IRC | 15:47 | |
*** jmcboots has joined #openstack | 15:47 | |
*** tmichael has joined #openstack | 15:48 | |
*** rmartinelli has joined #openstack | 15:49 | |
*** mikal has joined #openstack | 15:49 | |
*** tryggvil_ has joined #openstack | 15:50 | |
*** dieterdemeyer has quit IRC | 15:50 | |
*** markvoelker has quit IRC | 15:53 | |
*** anniec72 has quit IRC | 15:55 | |
*** Razique has quit IRC | 15:55 | |
*** vitiho has joined #openstack | 15:55 | |
*** salgado has quit IRC | 16:01 | |
*** salgado has joined #openstack | 16:01 | |
sstent | another keystone question. with endpoints do you create an entry for each node of that type? or should we create loadbalancer for each class? | 16:02 |
*** jcannava has left #openstack | 16:02 | |
heyho | hello.. quick question: is nova-api service mandatory in every compute-host if I use multi-host networking mode? Or can I just use nova-api-metadata? | 16:03 |
*** Gordonz has quit IRC | 16:03 | |
sstent | heyho: you should be able to jsut use metadata | 16:03 |
*** Gordonz has joined #openstack | 16:03 | |
zykes- | heyho: nova-api > your front end node towards users, nova-api-metata should run on each compute node i think even if not in multihost mode no ? | 16:04 |
*** jemartin has joined #openstack | 16:06 | |
heyho | sstent zykes.. I thought i read somwhere that nova-api also manage the assignement of the local private gateways ip on the physical compute nodes.. So these operations are done only by the nova-api on the controller? | 16:06 |
*** rmt has quit IRC | 16:07 | |
*** trevman2001 has joined #openstack | 16:09 | |
*** dhellmann has joined #openstack | 16:09 | |
heyho | I expressed myself bad, sorry.. nova-api should forward every public (internet) vm request outside the private network via the local gateways.. or these operations are performed by local nova-network services? | 16:11 |
*** asadpanda has joined #openstack | 16:11 | |
*** jj0hns0n has quit IRC | 16:11 | |
*** jj0hns0n has joined #openstack | 16:12 | |
sstent | heyho: this should be nova-networks job | 16:12 |
*** asavu has quit IRC | 16:13 | |
sstent | heyho: api should recieved service requests from users, pass them onto the scheduler, which farms them out to the approriate nova-network to create the networks | 16:13 |
*** Blackavar has joined #openstack | 16:14 | |
heyho | so "local" nova-apis make no sense if the requests are always made via the controller node, right? | 16:14 |
sstent | correct | 16:14 |
*** ondergetekende has joined #openstack | 16:14 | |
heyho | ok, thank you very much | 16:14 |
sstent | meta is jsut there to service the 169.254.169.254 requests | 16:15 |
*** melmoth has joined #openstack | 16:16 | |
*** sniperd has joined #openstack | 16:17 | |
*** shang has joined #openstack | 16:19 | |
*** asavu has joined #openstack | 16:20 | |
*** Leseb has quit IRC | 16:22 | |
*** gakott has quit IRC | 16:22 | |
*** osier has quit IRC | 16:24 | |
*** lborda has joined #openstack | 16:24 | |
*** rods has joined #openstack | 16:24 | |
*** shwetaap has joined #openstack | 16:25 | |
*** salgado is now known as salgado-lunch | 16:27 | |
*** maplebed has joined #openstack | 16:27 | |
*** CaptTofu has quit IRC | 16:28 | |
*** judu has quit IRC | 16:28 | |
*** aa__ has quit IRC | 16:28 | |
*** joearnold has joined #openstack | 16:28 | |
*** warik has joined #openstack | 16:29 | |
*** kokyu has quit IRC | 16:30 | |
*** timello has quit IRC | 16:32 | |
*** MarkAtwood has quit IRC | 16:32 | |
kiranmurari | FYI, OpenStack Beginner's Guide (Essex on Ubuntu 12.04) -> http://wp.me/pT1iY-q9 | 16:32 |
*** sniperd has quit IRC | 16:33 | |
*** joearnold has quit IRC | 16:33 | |
*** sniperd has joined #openstack | 16:33 | |
*** littleidea has quit IRC | 16:35 | |
*** vincentricci has joined #openstack | 16:35 | |
*** littleidea has joined #openstack | 16:36 | |
*** Free_maN has quit IRC | 16:36 | |
*** andrewsben_zz is now known as andrewsben | 16:37 | |
*** llang629 has joined #openstack | 16:39 | |
*** krow has joined #openstack | 16:39 | |
*** kiranmurari has quit IRC | 16:40 | |
*** sniperd has quit IRC | 16:43 | |
*** sniperd has joined #openstack | 16:44 | |
*** markvoelker has joined #openstack | 16:45 | |
*** CaptTofu has joined #openstack | 16:45 | |
*** llang629 has left #openstack | 16:47 | |
*** rafaduran has joined #openstack | 16:47 | |
*** hvb has joined #openstack | 16:47 | |
*** krow has quit IRC | 16:48 | |
*** joearnold has joined #openstack | 16:48 | |
*** RichardRaseley has joined #openstack | 16:48 | |
*** askstack has joined #openstack | 16:48 | |
RichardRaseley | If I were building some dedicated swift proxy servers, would it be worthwhile to look at enterprise class SSDs? | 16:49 |
*** sante has quit IRC | 16:49 | |
*** jgriffith has joined #openstack | 16:54 | |
*** kokyu has joined #openstack | 16:54 | |
LarsErikP | in flat networking, when injecting network config into an instance. How do that actually happen? Mounting disk image, and edit /etc/network/interfaces ? | 16:55 |
kokyu | check the cloud-init package that does its kun-fu on the instance-side :) | 16:55 |
LarsErikP | thanks | 16:56 |
*** spiffxp has joined #openstack | 16:57 | |
kokyu | nova-volumes service needs to *only* run on the controller node, right ? | 16:57 |
*** CatKiller is now known as Guest25068 | 16:58 | |
kokyu | that means, the "nova-volumes" LVM volume group just needs to exist on that host. | 16:58 |
kokyu | and remote compute nodes connect to them via iSCSI. or am I mistaken ? | 16:58 |
*** melmoth has quit IRC | 16:59 | |
*** shang has quit IRC | 16:59 | |
*** lborda has quit IRC | 17:01 | |
*** shang has joined #openstack | 17:02 | |
*** Glace has joined #openstack | 17:03 | |
*** davidha_who_took has joined #openstack | 17:05 | |
LarsErikP | so, i guess my assumption was right.. So, flatManager and windows instances can't work very well right? :s | 17:05 |
*** dnaori has joined #openstack | 17:06 | |
kokyu | LarsErikP: did you check what actually cloud-init is doing? in the end it'll be just one kind of dhcp request (and injecting ssh pub keys) | 17:07 |
kokyu | just look at what happens on a linux instance | 17:07 |
LarsErikP | maybe i looked in the wrong place.. where did you find that? | 17:08 |
*** milner has joined #openstack | 17:08 | |
*** ondergetekende has quit IRC | 17:08 | |
*** ahasenack has joined #openstack | 17:09 | |
kokyu | LarsErikP: I'm in the traing right now, I cannot look :-) | 17:09 |
*** davidha_who_took is now known as davidha | 17:09 | |
LarsErikP | oki :p | 17:09 |
*** nacx has quit IRC | 17:10 | |
LarsErikP | dhcp request? I was talking about flatmanager without dhcp | 17:10 |
*** jdurgin has joined #openstack | 17:10 | |
*** Glace has quit IRC | 17:11 | |
*** natea has joined #openstack | 17:12 | |
*** Guest25068 has quit IRC | 17:13 | |
*** marrusl has joined #openstack | 17:15 | |
anticw | OT; how do i stop emails coming from people who click "Contact this team" ? | 17:17 |
kokyu | LarsErikP: FlatManager AFAIK is using dnsmasq, wich (in fact) has a dhcp server included. but whether FlatManager make use of it. tell me :) | 17:17 |
*** bbcmicrocomputer has joined #openstack | 17:17 | |
*** bbcmicrocomputer has joined #openstack | 17:17 | |
*** bbcmicrocomputer has joined #openstack | 17:17 | |
kokyu | #spam | 17:17 |
*** salgado-lunch is now known as salgado | 17:17 | |
*** anderstj has joined #openstack | 17:19 | |
*** Triade has quit IRC | 17:19 | |
anticw | i guess i could always delete my LP account; it's one less cancer to worry about | 17:20 |
*** Glace has joined #openstack | 17:22 | |
LarsErikP | is there support for Windows7 guests? Or only 2008 Server? | 17:23 |
anticw | it's a licensing issue | 17:23 |
anticw | win7 should work | 17:23 |
*** jj0hns0n has quit IRC | 17:24 | |
*** jj0hns0n has joined #openstack | 17:24 | |
LarsErikP | ok | 17:25 |
Darksecond | what if 2 nodes don't have the same nic card names? fedora names them by pci slot these days, should i rename them to all be the same? (ethX) | 17:26 |
*** clopez has quit IRC | 17:27 | |
*** epim has joined #openstack | 17:27 | |
*** melmoth has joined #openstack | 17:27 | |
*** nphase has joined #openstack | 17:27 | |
*** dachary has quit IRC | 17:29 | |
*** hggdh has joined #openstack | 17:29 | |
*** dieterdemeyer has joined #openstack | 17:30 | |
*** e1mer has quit IRC | 17:30 | |
*** alrs has joined #openstack | 17:31 | |
*** gakott has joined #openstack | 17:32 | |
*** timello has joined #openstack | 17:32 | |
anticw | oh jesus, it's like Hotel California | 17:33 |
*** reed has quit IRC | 17:33 | |
*** GheRivero_ has quit IRC | 17:33 | |
*** dieterdemeyer has quit IRC | 17:33 | |
*** Leseb has joined #openstack | 17:34 | |
*** RichardRaseley has quit IRC | 17:35 | |
*** dendro-afk is now known as dendrobates | 17:35 | |
*** dieterdemeyer has joined #openstack | 17:35 | |
*** whenry has quit IRC | 17:36 | |
*** aa has joined #openstack | 17:37 | |
*** vitiho has quit IRC | 17:39 | |
*** dieterdemeyer has quit IRC | 17:39 | |
*** dieterdemeyer has joined #openstack | 17:40 | |
apergos | are you trying to leave? | 17:40 |
*** whenry has joined #openstack | 17:41 | |
anticw | apergos: i can deactivate it, that's probably good enough | 17:42 |
anticw | though it's not clear how to get mailing list feed then | 17:43 |
apergos | heh | 17:43 |
anticw | launchpad is without a doubt one of the most frustrating, obtuse and invasive development tools i've encountered (me and the 89000 others who comaplain about this as well) | 17:44 |
apergos | oh. yeah I'm not a fan | 17:45 |
uvirtbot | New bug: #993859 in quantum "quamtum client --debug option needed" [Undecided,New] https://launchpad.net/bugs/993859 | 17:46 |
*** alrs has quit IRC | 17:46 | |
*** ShKoDrAnI has quit IRC | 17:49 | |
*** marrusl has quit IRC | 17:50 | |
*** jog0 has joined #openstack | 17:50 | |
*** ShKoDrAnI has joined #openstack | 17:51 | |
*** lborda has joined #openstack | 17:51 | |
*** dieterdemeyer has quit IRC | 17:51 | |
*** e1mer has joined #openstack | 17:52 | |
*** alrs has joined #openstack | 17:53 | |
*** johnpostlethwait has joined #openstack | 17:54 | |
*** ahasenack has quit IRC | 17:54 | |
*** dieterdemeyer has joined #openstack | 17:55 | |
*** mdrnstm has joined #openstack | 17:55 | |
*** pvo is now known as pvo-away | 17:56 | |
*** bbcmicrocomputer has quit IRC | 17:56 | |
*** avoine has quit IRC | 17:57 | |
*** mdomsch has joined #openstack | 17:58 | |
*** asavu has quit IRC | 17:59 | |
*** bbcmicrocomputer has joined #openstack | 17:59 | |
*** hggdh has quit IRC | 17:59 | |
*** swift-india has joined #openstack | 18:00 | |
*** bpaluch has joined #openstack | 18:00 | |
*** albert23 has joined #openstack | 18:00 | |
*** avoine has joined #openstack | 18:01 | |
bpaluch | hey does fedora or centos make openstack images simlar to UEC ubuntu images? I havne't found any useful links everything seems to be people making their own, I don't feel like recreating the wheel. | 18:01 |
*** Ryan_Lane has joined #openstack | 18:02 | |
*** sniperd has quit IRC | 18:03 | |
*** sniperd has joined #openstack | 18:03 | |
*** hub-cap has joined #openstack | 18:03 | |
*** armax has joined #openstack | 18:03 | |
*** armax has left #openstack | 18:03 | |
_val_ | bpaluch: maybe http://uksysadmin.wordpress.com/2011/03/03/openstack-nova-centos-instance/ | 18:04 |
*** hub-cap has quit IRC | 18:04 | |
*** hub-cap has joined #openstack | 18:04 | |
*** Stackops-Jorge has quit IRC | 18:05 | |
*** llang629_ has joined #openstack | 18:06 | |
bpaluch | hmm this is what I've found but thanks i might as well just bite the bullet and start | 18:06 |
*** llang629_ has left #openstack | 18:06 | |
*** hub_cap has quit IRC | 18:06 | |
*** hub-cap is now known as hub_cap | 18:06 | |
*** lborda has quit IRC | 18:07 | |
*** dendrobates is now known as dendro-afk | 18:07 | |
*** jgriffith has quit IRC | 18:07 | |
_val_ | you probably have somewhere preinstalled images | 18:07 |
*** jgriffith has joined #openstack | 18:09 | |
*** ahasenack has joined #openstack | 18:10 | |
*** maplebed has quit IRC | 18:10 | |
*** littleidea has quit IRC | 18:12 | |
*** ik04ka has quit IRC | 18:13 | |
*** apetrescu has joined #openstack | 18:16 | |
*** maplebed has joined #openstack | 18:18 | |
*** sniperd has quit IRC | 18:19 | |
*** sniperd has joined #openstack | 18:19 | |
*** jgriffith has quit IRC | 18:19 | |
*** dpippenger has joined #openstack | 18:19 | |
*** marrusl has joined #openstack | 18:19 | |
*** okapi14 has joined #openstack | 18:21 | |
*** jgriffith has joined #openstack | 18:21 | |
*** jgriffith is now known as jgriffith_away | 18:21 | |
uvirtbot | New bug: #995280 in openstack-ci "replace launchpad translations with transifex" [High,Incomplete] https://launchpad.net/bugs/995280 | 18:21 |
apetrescu | I feel stupid for asking this, but how do I add user accounts for the OpenStack Dashboard in Ubuntu 12.04? It won't take my normal user credentials. What does it auth against? | 18:23 |
*** reed has joined #openstack | 18:23 | |
*** jog0 has left #openstack | 18:23 | |
apetrescu | I can't seem to log into it at all | 18:24 |
*** camm has quit IRC | 18:25 | |
*** Trixboxer has quit IRC | 18:27 | |
*** mrmartin has joined #openstack | 18:29 | |
mrmartin | re | 18:29 |
*** mrmartin has quit IRC | 18:29 | |
apetrescu | Hm? | 18:30 |
*** timello has quit IRC | 18:30 | |
*** davepigott has joined #openstack | 18:32 | |
*** mrmartin has joined #openstack | 18:33 | |
*** vincentricci has quit IRC | 18:34 | |
*** davepigott has quit IRC | 18:34 | |
*** vincentricci has joined #openstack | 18:34 | |
*** jj0hns0n has quit IRC | 18:39 | |
*** kokyu has quit IRC | 18:41 | |
*** ShKoDrAnI has quit IRC | 18:42 | |
*** kokyu has joined #openstack | 18:43 | |
*** timello has joined #openstack | 18:44 | |
*** ShKoDrAnI has joined #openstack | 18:44 | |
*** vitiho has joined #openstack | 18:44 | |
*** marrusl has quit IRC | 18:44 | |
*** esteimann has left #openstack | 18:45 | |
*** askstack has quit IRC | 18:45 | |
*** broussardc has quit IRC | 18:45 | |
*** longhudou has quit IRC | 18:45 | |
*** Fu4ny has quit IRC | 18:45 | |
*** praefect has quit IRC | 18:45 | |
*** heyho has quit IRC | 18:45 | |
*** kiranmurari has joined #openstack | 18:45 | |
*** mnewby has joined #openstack | 18:47 | |
*** mnewby has quit IRC | 18:47 | |
*** mnewby has joined #openstack | 18:48 | |
*** reed has quit IRC | 18:49 | |
*** melmoth has quit IRC | 18:49 | |
*** mdrnstm has quit IRC | 18:50 | |
*** melmoth has joined #openstack | 18:50 | |
*** bbcmicrocomputer has quit IRC | 18:50 | |
*** pretec has quit IRC | 18:50 | |
*** KarinLevenstein has joined #openstack | 18:51 | |
*** pretec has joined #openstack | 18:51 | |
*** alrs has quit IRC | 18:52 | |
*** avoine has quit IRC | 18:52 | |
*** never2far has quit IRC | 18:52 | |
*** mdrnstm has joined #openstack | 18:52 | |
*** bbcmicrocomputer has joined #openstack | 18:54 | |
kokyu | nova-volumes service needs to *only* run on the controller node, right ? | 18:55 |
*** lborda has joined #openstack | 18:56 | |
uvirtbot | New bug: #994471 in quantum "dnsmasq config is not updated upon instance creation" [Undecided,New] https://launchpad.net/bugs/994471 | 18:56 |
kokyu | and remote compute nodes connect to them via iSCSI. or am I mistaken ? | 18:57 |
*** alrs has joined #openstack | 18:59 | |
*** KavanS has quit IRC | 18:59 | |
*** KavanS has joined #openstack | 18:59 | |
kiranmurari | kokyu, true... your compute nodes need not have nova-volume installed on them as well.... | 19:00 |
*** dolphm has joined #openstack | 19:01 | |
*** pnavarro has joined #openstack | 19:01 | |
uvirtbot | New bug: #994296 in glance "Glance can't deploy an image from Swift if the Swift account name contains an '@'" [Medium,In progress] https://launchpad.net/bugs/994296 | 19:01 |
uvirtbot | New bug: #994609 in glance "wsgi.Server() starts but is broken on osx (test_multiprocessing never ends)" [Critical,In progress] https://launchpad.net/bugs/994609 | 19:01 |
kokyu | kiranmurari: very nice then :) | 19:02 |
*** marrusl has joined #openstack | 19:02 | |
kokyu | kiranmurari: but still, compute nodes need plenty of GB for the main OS disk image to be stored locally, or can I make it a iSCSI-target aswell ? | 19:03 |
*** marrusl has quit IRC | 19:03 | |
*** marrusl has joined #openstack | 19:03 | |
*** avoine has joined #openstack | 19:03 | |
*** alrs has quit IRC | 19:04 | |
*** mrunge has joined #openstack | 19:05 | |
uvirtbot | New bug: #992319 in glance "Patch to openstack.common.cfg should be removed" [High,In progress] https://launchpad.net/bugs/992319 | 19:05 |
uvirtbot | New bug: #996120 in devstack "libmysqlclient-dev missing from nova packages" [Undecided,In progress] https://launchpad.net/bugs/996120 | 19:05 |
*** llang629__ has joined #openstack | 19:06 | |
*** MarkAtwood has joined #openstack | 19:06 | |
*** llang629__ has left #openstack | 19:06 | |
*** hggdh has joined #openstack | 19:07 | |
*** asavu has joined #openstack | 19:07 | |
*** milner has quit IRC | 19:08 | |
*** kiranmurari has quit IRC | 19:09 | |
*** askstack has joined #openstack | 19:09 | |
*** littleidea has joined #openstack | 19:13 | |
*** reed has joined #openstack | 19:13 | |
*** kvad32 has joined #openstack | 19:14 | |
*** kfox1111 has joined #openstack | 19:15 | |
kfox1111 | Question. Does openstack support something like EC2's Tag api? | 19:15 |
*** JesperA has joined #openstack | 19:16 | |
kbringard | with the nova api you can set meta-data for instances at boot time | 19:17 |
*** dieterdemeyer has quit IRC | 19:17 | |
kbringard | I don't think you can do it with the ec2 API, though | 19:17 |
kbringard | nova api == openstack API | 19:17 |
*** dieterdemeyer has joined #openstack | 19:17 | |
kbringard | you specify —meta key=value for as many arbitrary values as you'd like | 19:17 |
kbringard | when you run nova boot | 19:18 |
kfox1111 | how do you fetch it in the instance? | 19:19 |
*** rods has quit IRC | 19:19 | |
*** rods has joined #openstack | 19:19 | |
kfox1111 | hmm.. looks like dashboard doesn't support setting them either. :/ | 19:19 |
*** hggdh has quit IRC | 19:20 | |
*** dprince has quit IRC | 19:20 | |
*** hggdh has joined #openstack | 19:21 | |
*** CaptTofu has quit IRC | 19:21 | |
*** pengyong has quit IRC | 19:21 | |
*** DavidLevin has quit IRC | 19:22 | |
*** eglynn has quit IRC | 19:22 | |
askstack | I have nova + nova volume + swift . when I launch an instance with 1GB RAM and 10GB disk, where does this 10GB disk gets its disk space from? I don't see empty space decrease in swift nor noav-volume. | 19:22 |
*** eglynn has joined #openstack | 19:22 | |
kfox1111 | so, is openstack moving away from supporting the ec2 api? | 19:23 |
*** jgriffith_away has quit IRC | 19:24 | |
kvad32 | askstack: Im not sure but maybe because if i member nova volume uses LVM so that space is already allocated? | 19:25 |
kvad32 | askstack: Just a quess | 19:25 |
*** asavu has quit IRC | 19:25 | |
askstack | kvad32: i ran pvdisplay before and after i created this instance, i don't see "Free PE " decrease. | 19:28 |
*** joearnol_ has joined #openstack | 19:28 | |
askstack | also, my nova-volume is 20GB. When I launch an instance, dashboard shows "1000GB available" | 19:28 |
*** rgoodwin is now known as rgoodwin_away | 19:28 | |
askstack | Kvad32: thanks, I wll keep looking | 19:29 |
kvad32 | askstack: kk good luck srry couldnt help | 19:30 |
*** anderstj has quit IRC | 19:30 | |
*** joearnold has quit IRC | 19:30 | |
kvad32 | askstack: When you find out please let me know | 19:30 |
askstack | kvad32: sure will. | 19:30 |
*** camm has joined #openstack | 19:30 | |
*** shang has quit IRC | 19:31 | |
*** hggdh has quit IRC | 19:31 | |
*** rkukura has quit IRC | 19:31 | |
VivekVC | annegentle: Hi | 19:32 |
*** rgoodwin_away is now known as rgoodwin | 19:33 | |
*** marrusl has quit IRC | 19:34 | |
*** CaptTofu has joined #openstack | 19:35 | |
*** Razique has joined #openstack | 19:36 | |
*** dieterdemeyer has quit IRC | 19:37 | |
kfox1111 | How do you get the tags a guest has set from the guest? | 19:37 |
sstent | askstack: unless you're booting from volume, the instance data is in /var/lib/nova/instances on the compute node | 19:37 |
*** dieterdemeyer has joined #openstack | 19:37 | |
*** melmoth has quit IRC | 19:37 | |
*** anderstj has joined #openstack | 19:37 | |
*** vitiho has quit IRC | 19:37 | |
*** cctrieloff has quit IRC | 19:37 | |
*** egant has quit IRC | 19:38 | |
sstent | askstack: as for dashboard shows "1000GB available" | 19:40 |
*** bbcmicrocomputer has quit IRC | 19:40 | |
sstent | askstack: that sounds like the quota for your account | 19:40 |
askstack | sstent: kvad32: you are right, /var/lib/nova/instances/ has the disk file, I download an ISO in the vm and the "disk" file increased from 6MB to 3.8G | 19:43 |
*** lborda has quit IRC | 19:44 | |
askstack | kvad32: i found this to be helpful, http://forums.openstack.org/viewtopic.php?f=16&t=536 | 19:45 |
askstack | kvad32: The default implementation in OpenStack uses local-to-host storage for the / partition and any additional storage specified in the "flavor" at guest creation. | 19:45 |
*** clopez has joined #openstack | 19:45 | |
sstent | askstack: also if your on kvm, i believe you'll see 2 disks in your vm, 1 for the OS + 1 of instance size | 19:46 |
*** sniperd has quit IRC | 19:46 | |
*** jj0hns0n has joined #openstack | 19:46 | |
askstack | sstent: yes, one named "disk" and one "disk.local" | 19:46 |
*** egant has joined #openstack | 19:47 | |
askstack | sstent: disk.local didn't change size | 19:47 |
*** viralpatadiya has joined #openstack | 19:48 | |
askstack | sstent: according to the libvirt.xml, "disk" is for vda and "disk.local" is for vdb. | 19:49 |
sstent | askstack: that's right | 19:49 |
uvirtbot | New bug: #996139 in keystone "KVS tokens not shared between processes" [Undecided,In progress] https://launchpad.net/bugs/996139 | 19:50 |
*** ahasenack has quit IRC | 19:51 | |
*** swift-india has quit IRC | 19:51 | |
*** jgriff has joined #openstack | 19:51 | |
*** BLZbubba has joined #openstack | 19:53 | |
BLZbubba | how painful would it be to make nova-network limit outbound connections too? | 19:53 |
kokyu | if nova-volumes is comparable to Amazon's EBS, can I use them as root disks, too? | 19:53 |
*** joearnol_ has quit IRC | 19:54 | |
*** anderstj has quit IRC | 19:55 | |
*** dieterdemeyer has quit IRC | 19:55 | |
*** sniperd has joined #openstack | 19:55 | |
*** dieterdemeyer has joined #openstack | 19:56 | |
*** melmoth has joined #openstack | 19:57 | |
*** asavu has joined #openstack | 19:57 | |
askstack | kokyu: i have no expericne in it, but i just saw this, http://forums.openstack.org/viewtopic.php?f=16&t=536, read the second message . | 19:59 |
kokyu | askstack: hehe, funny name you have. and thanks for the reference :) | 19:59 |
*** praefect has joined #openstack | 20:01 | |
*** dolphm has quit IRC | 20:02 | |
*** viralpatadiya has quit IRC | 20:03 | |
*** swift-india has joined #openstack | 20:03 | |
*** mohits has quit IRC | 20:03 | |
*** jj0hns0n has quit IRC | 20:03 | |
*** jj0hns0n has joined #openstack | 20:04 | |
*** DavidLevin has joined #openstack | 20:04 | |
*** reed has quit IRC | 20:04 | |
*** jj0hns0n has quit IRC | 20:04 | |
*** jj0hns0n has joined #openstack | 20:05 | |
*** milner has joined #openstack | 20:05 | |
uvirtbot | New bug: #995251 in openstack-ci "Run tempest in the devstack gate" [High,In progress] https://launchpad.net/bugs/995251 | 20:06 |
*** rmartinelli has quit IRC | 20:07 | |
*** Ruetobas has quit IRC | 20:07 | |
*** dieterdemeyer has quit IRC | 20:07 | |
*** egant has quit IRC | 20:08 | |
*** hub-cap has joined #openstack | 20:08 | |
*** dieterdemeyer has joined #openstack | 20:09 | |
*** cctrieloff has joined #openstack | 20:10 | |
kokyu | askstack: for some reason, I cannot access this site at all (forums.openstack.org) :( | 20:11 |
*** hub_cap has quit IRC | 20:11 | |
*** hub-cap is now known as hub_cap | 20:11 | |
*** dieterdemeyer has quit IRC | 20:14 | |
*** Ruetobas has joined #openstack | 20:15 | |
uvirtbot | New bug: #996157 in glance "backend storage error not logged properly" [Undecided,New] https://launchpad.net/bugs/996157 | 20:15 |
*** gourneau has joined #openstack | 20:16 | |
gourneau | Hey Y'all. What is a good modern alternative to Sun Grid Engine? I need a task queue for a cluster of machines. Does open stack have this (I am a noob) | 20:17 |
kokyu | define "task queue" | 20:17 |
kokyu | gourneau: some software that enqueues jobs to be performed by worker nodes? (that's how it sounds like) | 20:18 |
*** natea is now known as natea|busy | 20:18 | |
*** joearnold has joined #openstack | 20:19 | |
gourneau | kokyu: yup that is it. | 20:19 |
kokyu | gourneau: I very hardly doubt that OpenStack provides one, because this is an entirely different subject and layer of problem to solve. | 20:20 |
*** egant has joined #openstack | 20:20 | |
kokyu | gourneau: I do not know what platform you are on, and what your needs are, but maybe (for now) you might want to have a look at Resque ( http://github.com/bvandenbos/resque-scheduler ) | 20:21 |
*** anderstj has joined #openstack | 20:22 | |
askstack | kokyu: https://lists.launchpad.net/openstack/msg05284.html | 20:22 |
gourneau | okay, that is what I thought but I was just curious. I will keep looking around. Thanks! | 20:22 |
kokyu | askstack: thanks | 20:23 |
*** gyee has joined #openstack | 20:23 | |
*** heckj has quit IRC | 20:25 | |
*** hvb has quit IRC | 20:25 | |
*** vitiho has joined #openstack | 20:26 | |
uvirtbot | New bug: #996163 in quantum "remove unneeded imports in quantum/common/config.py" [Undecided,New] https://launchpad.net/bugs/996163 | 20:26 |
*** wilmoore has joined #openstack | 20:28 | |
*** jaypipes is now known as jaypipes-afk | 20:30 | |
*** xmltok has joined #openstack | 20:31 | |
kokyu | anyone experienced with VLAN within OpenStack? | 20:34 |
zykes- | kokyu: experienced how ? | 20:34 |
zykes- | i've used Quantum & VLANManager | 20:35 |
*** heckj has joined #openstack | 20:35 | |
*** heckj has quit IRC | 20:35 | |
*** mrunge has quit IRC | 20:35 | |
uvirtbot | New bug: #996171 in openstack-ci "add an openvz enabled builder" [High,Triaged] https://launchpad.net/bugs/996171 | 20:35 |
uvirtbot | New bug: #996173 in openstack-ci "run devstack tests on kvm and xen in addition to qemu" [Medium,Triaged] https://launchpad.net/bugs/996173 | 20:35 |
kokyu | I would like to let all OpenStack Projects (/ its networks) make use of VLANs, so it won't interfere with the management network (where the openstack resides in, and the switches, IPMIs, etc) | 20:35 |
kokyu | (pretty buggy day ^^) | 20:36 |
*** vmlinuz has quit IRC | 20:39 | |
*** ahasenack has joined #openstack | 20:39 | |
*** hub_cap has quit IRC | 20:39 | |
*** Free_maN has joined #openstack | 20:40 | |
*** Free_maN has joined #openstack | 20:40 | |
*** hggdh has joined #openstack | 20:42 | |
kokyu | has anyone ever prepared/tested live migrations via Nova (using KVM/libvirt)? | 20:44 |
kokyu | reading the docs, i read, that you MUST NFS-mount the /var/lib/nova/instances dir, implying that the actual compute nodes (that mount this dir from remote master) do really not need much disk space then. or am I mistaken ? | 20:45 |
zynzel | if you need live migration yes | 20:45 |
zynzel | but you can use kvm block migration. | 20:46 |
zynzel | it works after 'some' changes in code. (maybe this was fault of my env) | 20:46 |
kokyu | oh | 20:46 |
kokyu | never heard of "kvm block migration"s yet :-) | 20:47 |
avoine | I would be curious to know how it works also | 20:47 |
kokyu | hehe | 20:48 |
zykes- | why use KVM ? | 20:48 |
avoine | I guess this is what your talking about: http://wiki.libvirt.org/page/FAQ#What_setup_is_required_for_QEMU.2FKVM_migration.3F | 20:48 |
zykes- | use Gluster fs instead shared across nodes ? | 20:49 |
zynzel | problem with block migration in multi_node env is: 1) vm xml (on every compute node, other vnc listen address) 2) no support for floating ip migration 3) fucked up code for getting image from glance (maybe in essex it works) | 20:49 |
zykes- | zynzel: i wonder if Kiall was looking at it :p | 20:49 |
zynzel | dunno, but i was ;) and now its works :) | 20:50 |
zynzel | 3day of work... | 20:50 |
zykes- | zynzel: what works ? | 20:51 |
zynzel | kvm block migration ;) | 20:51 |
zykes- | ah, you use multi_host or ? | 20:51 |
zynzel | yeah, multihost and flatdhcp | 20:51 |
zykes- | hmm, you running trunk or ? | 20:51 |
zynzel | nope, e3 (before stable essex) | 20:52 |
kokyu | what else despite flatdhcp does exist? | 20:52 |
zykes- | hmm, I thought Kiall said it didn't work with migrations and multi_host (Running nova network on all compute nodes) | 20:52 |
zynzel | kokyu: ? i dont understand | 20:53 |
*** sacharya1 has quit IRC | 20:53 | |
zynzel | zykes-: we have multi_host, so i dont have much choice ;) | 20:54 |
kokyu | zynzel: nova supports different networking modes, FlatDHCP, and what else? | 20:54 |
zykes- | zynzel: weird then that he said it didn't work | 20:54 |
*** monst has joined #openstack | 20:54 | |
zynzel | kokyu: flatdhcp, flat, vlan i think | 20:54 |
kokyu | zynzel: as far as I read in this channel, everybody is using FlatDHCP. but if that's configurable, there must be other choices, too | 20:54 |
monst | anyone had the issue where vcpu wont go higher then cores? | 20:55 |
zynzel | zykes-: i have to fix 'a few lines' | 20:55 |
zykes- | vishy: would multi_host networking work with live migrations ? | 20:55 |
zykes- | you had to change code zynzel ? | 20:55 |
monst | once I go past 16vcpu instances stay in build state. | 20:55 |
monst | i have 16 cores | 20:55 |
zynzel | zykes-: yeah | 20:56 |
kokyu | zynzel: oha, and when using vlan, it'll make all IP packets from the instances tagged, right? that's the point I am interested in! - do I need to configure anything special on the switches to support that? | 20:56 |
zykes- | what did you need to fix then ? | 20:56 |
kokyu | zynzel: I know I can configure vlan (tagged and untagged) in our switches, but that is for segmenting the ethernet ports only | 20:56 |
zynzel | zykes-: i dont have access now to git, if you are intrested plz remind me about it tomorrow | 20:57 |
vishy | zykes-: yes | 20:57 |
vishy | zykes-: it just doesn't automatically move floating ips | 20:57 |
zynzel | kokyu: i dont use vlan mode, so i cant help you :) | 20:57 |
zykes- | vishy: ah, I wonder if it was that that Kiall was working on :p | 20:57 |
zykes- | zynzel: did you fix floating ? | 20:57 |
zynzel | zykes-: yes | 20:57 |
zykes- | and fw rules + gw ? | 20:58 |
kokyu | zynzel: I'm still on mission to find one though :) | 20:58 |
zynzel | zykes-: yes, gw i 'auto fixed' after vm request address from dhcp | 20:58 |
zynzel | (after about ~1mint) | 20:58 |
zynzel | min* | 20:58 |
vishy | zykes-: http://wiki.openstack.org/ReleaseNotes/Essex#Known_Issues | 20:58 |
zynzel | gw is* | 20:58 |
vishy | 5th paragraph | 20:58 |
apetrescu | I'm trying to set up a brand new OpenStack on some hardware, but for some reason `keystone-manager db_sync` doesn't seem to do anything (no errors, but nothing happens in the DB). As far as I can tell, I've configured the connection properly in keystone.conf. Anybody have an inkling of what might be happening? | 20:58 |
*** ahasenack has quit IRC | 20:59 | |
*** GheRivero_ has joined #openstack | 20:59 | |
*** CaptTofu has quit IRC | 20:59 | |
*** maplebed has quit IRC | 21:00 | |
zynzel | zykes-: with fw rules was problem only with dhcp server ip, that was easy ;) | 21:00 |
kokyu | apetrescu: into what db do you expect to find the schema/data ? mysql? did you set it to use mysql, too? check the sqlite's file then, maybe you're still using sqlite ;) | 21:00 |
zynzel | floating ip migration is also easy (release, migrate, assign) | 21:00 |
*** mrmartin has quit IRC | 21:00 | |
zykes- | zynzel: ebtables ? | 21:01 |
zynzel | bigest problem was with changing kvm configuration on the fly (other ip address for vnc, depends on compute node) | 21:01 |
apetrescu | kokyu: Sorry, I should've specified, yes, I changed the connection string in /usr/share/keystone/keystone.conf from the default sqlite keystone.db file to a mysql:/// one | 21:01 |
*** milner has quit IRC | 21:01 | |
*** RichardRaseley has joined #openstack | 21:02 | |
*** ywu has joined #openstack | 21:02 | |
apetrescu | Even deleting the old keystone.db file doesn't help | 21:03 |
apetrescu | Hmmm | 21:03 |
apetrescu | Actually it recreated it | 21:03 |
apetrescu | Even though I changed keystone.conf | 21:03 |
apetrescu | So that's the problem I suppose | 21:03 |
apetrescu | Is there anything else I have to do besides change the connection string in keystone.conf to make it use MySQL instead of sqlite? | 21:04 |
*** ahasenack has joined #openstack | 21:04 | |
RichardRaseley | Can anyone provide more information on how I would support a multi-site Swift installation using Keystone for authentication? | 21:04 |
*** asadpanda has quit IRC | 21:04 | |
kokyu | apetrescu: I do not know your environment, but the keystone.conf you are supposed to edit should be in /etc/keystone/ isntead :) | 21:05 |
*** asadpanda has joined #openstack | 21:05 | |
* apetrescu feels really stupid right now | 21:05 | |
apetrescu | I must've been reading some old doc that specified /usr/share instead of /etc. Thanks :) That's probably it | 21:06 |
apetrescu | :q | 21:06 |
monst | anyone have an idea why my compute node would stop building instance once vcpu reached my core count? | 21:07 |
monst | I assumed you could over allocate and have thread ques like ESX? | 21:07 |
*** ahasenack has quit IRC | 21:07 | |
*** ahasenack has joined #openstack | 21:08 | |
monst | a compute node cant be limited to 16 machines if there is 16 cores, right? | 21:08 |
kbringard | I think the simple scheduler will do a resource check and won't assign a VM to node unless it has N number of cores available | 21:09 |
kokyu | apetrescu: if you're on Ubuntu (ideally 12.04), then try http://www.hastexo.com/resources/docs/installing-openstack-essex-20121-ubuntu-1204-precise-pangolin as a first start, and don't make the same mistake I did: read carefully! after it, I highly highly recommend to read the official docu at docs.openstack.org, no matter how much it feels like, but it helped me greatly, too (but still, I'm having issues :D) | 21:09 |
kbringard | but chance scheduler just does a rand() % and dolls them out indiscriminately | 21:09 |
monst | I am using a single node setup | 21:10 |
apetrescu | Thanks, kokyu, I'd only been using the official docs so far. | 21:10 |
monst | is there a limitation in openstack that vcpus cant exceed cores? | 21:10 |
kokyu | apetrescu: then check the URL path, it should start with /trunk and not /diabo or alike ;) | 21:11 |
monst | cause that would straight kill the project I am working on | 21:11 |
*** vitiho has quit IRC | 21:11 | |
kokyu | monst: it can, see cpu_allocation_ratio (=1.5) | 21:11 |
kokyu | monst: same goes for RAM | 21:11 |
monst | ah thank god! | 21:11 |
*** judu has joined #openstack | 21:12 | |
*** zz_jce is now known as jce | 21:12 | |
kokyu | I wasn't able to verify that yet, because there's also a mac_cores setting in nova.conf, that bothers me, but you can tell me as soon as you did ;) | 21:12 |
*** Free_maN has quit IRC | 21:12 | |
monst | gonna try it now | 21:13 |
kbringard | I believe max_cores is the number of VCPUs that a compute can have assigned to it | 21:13 |
kbringard | before the scheduler will stop giving it VMs | 21:13 |
kokyu | kbringard: yeah, but how does that play together with the cpu_allocation_ratio, that explictely means, that you can overcommit the number of physical CPUs on that compute node | 21:14 |
kbringard | no idea, I ws not aware of the allocation ratio | 21:15 |
kokyu | :) | 21:16 |
*** jgriff has quit IRC | 21:17 | |
kokyu | can anyone tell me the relation between the networks created via `nova-manage network ...` and the nova.conf's flat_network_dhcp_start=$IP value? the latter seems to work for just one network, but if you've created multiple networks via nova-manage (each for one project, being associated), then this does not fit. | 21:17 |
kokyu | same for the fixed_range flag | 21:18 |
*** GheRivero_ has quit IRC | 21:19 | |
monst | anyone run stackops distro? | 21:20 |
*** hggdh has quit IRC | 21:20 | |
*** johnpur has joined #openstack | 21:21 | |
*** ChanServ sets mode: +v johnpur | 21:21 | |
kokyu | monst: we tried at our company (a college of mine), but we ended up with installing directly on a Ubuntu 12.04 ;) | 21:21 |
uvirtbot | New bug: #996190 in openstack-ci "Selenium support on CI infrastructure" [Undecided,New] https://launchpad.net/bugs/996190 | 21:21 |
monst | kokyu yeah we will probably go that route too | 21:22 |
monst | the nova.conf is located @ /var/lib/nova/bin/nova.conf | 21:22 |
monst | and says not to change the settings manually | 21:22 |
monst | any idea what that is about? | 21:22 |
kokyu | because stackops is meant to configure everything for ya | 21:23 |
*** kbringard has quit IRC | 21:23 | |
kokyu | that's what stackops is all about. having a one-click install OpenStack deployment. (didn't work out for us either ;) | 21:23 |
monst | yeah I wonder if i can just copy the nova.conf to /etc/nova/ | 21:23 |
monst | and add my parameters | 21:23 |
kokyu | doesn't work unless the processes know about it (check process list to see) | 21:24 |
*** CaptTofu has joined #openstack | 21:24 | |
monst | hmm | 21:25 |
kokyu | Has anyone ever tried setting up multiple nova networks? my new VMs now always get an IP from *both* networks, created, and neither works for SSH'ing in. | 21:25 |
uvirtbot | New bug: #992228 in tempest "Testcase: Some new tests to be added to test_server_actions.py" [Undecided,New] https://launchpad.net/bugs/992228 | 21:25 |
*** markvoelker has quit IRC | 21:29 | |
*** tong has quit IRC | 21:29 | |
*** kbringard has joined #openstack | 21:30 | |
*** epim_ has joined #openstack | 21:33 | |
*** drno_ has joined #openstack | 21:33 | |
drno_ | Hiya folks. | 21:33 |
*** hggdh has joined #openstack | 21:33 | |
*** epim__ has joined #openstack | 21:33 | |
*** azbarcea has quit IRC | 21:33 | |
drno_ | Working off of "uksysadmin's" video, got openstack essex installed... but running into a couple errors with images. | 21:34 |
drno_ | Has anyone run into problems with glance following those instructions recently? | 21:34 |
*** epim has quit IRC | 21:36 | |
*** epim__ is now known as epim | 21:36 | |
uvirtbot | New bug: #996199 in swift "Strange problem with account autocreate and swift dispersion" [Undecided,New] https://launchpad.net/bugs/996199 | 21:36 |
*** epim_ has quit IRC | 21:37 | |
*** dolphm has joined #openstack | 21:38 | |
kokyu | drno_: do you mind giving that link away? :) | 21:39 |
*** lts has quit IRC | 21:39 | |
*** milner has joined #openstack | 21:39 | |
*** melmoth has quit IRC | 21:40 | |
*** natea|busy has quit IRC | 21:41 | |
*** Gordonz has quit IRC | 21:43 | |
*** kokyu has quit IRC | 21:44 | |
*** ohnoimde1d has quit IRC | 21:45 | |
*** reed has joined #openstack | 21:46 | |
*** kokyu has joined #openstack | 21:47 | |
*** hggdh has quit IRC | 21:47 | |
*** vitiho has joined #openstack | 21:47 | |
*** hggdh has joined #openstack | 21:48 | |
*** RamJett has joined #openstack | 21:51 | |
*** h0cin has quit IRC | 21:51 | |
*** bsza has quit IRC | 21:51 | |
*** sniperd has quit IRC | 21:52 | |
*** ahasenack has quit IRC | 21:53 | |
RamJett | I'm setting up all nova stuff on a box and having problems. Any help me with a error. I manually start nova-api and get this "CRITICAL nova [-] [Errno 13] Permission denied: '/usr/lib/python2.7/dist-packages/keys'" . Where should I start looking? | 21:53 |
*** reed has quit IRC | 21:56 | |
kokyu | RamJett: at the permissions and ownership of the path described. | 21:57 |
*** melmoth has joined #openstack | 21:58 | |
RamJett | /usr/lib/python2.7/dist-packages/keys ... there in no file there. Did I not install something that is needed ?? | 21:58 |
kokyu | sorry, can't help ya there then (: - maybe try starting as root | 21:59 |
*** milner has quit IRC | 22:00 | |
*** koolhead17 has joined #openstack | 22:00 | |
RichardRaseley | How do I make the network which the swift proxy uses to communicate with the storage nodes different from that which the swift storage nodes use to replicate the objects? All I can see is the stuff located in the ring builder which seems to be the storage network... | 22:00 |
*** kbringard has quit IRC | 22:01 | |
*** alrs has joined #openstack | 22:01 | |
Glace | RichardRaseley: Not sure I understand your question properly.. but you can configure two variables for the PROXY ips and STORAGE ips in your conf file | 22:03 |
RichardRaseley | That would be in the proxy conf? | 22:03 |
*** gyee has quit IRC | 22:03 | |
*** imsplitbit has quit IRC | 22:04 | |
Glace | Depending on your setup.. in the proxy Conf.. I put the IP in my access network that the load balancer hit swift with.. no storage network | 22:04 |
RichardRaseley | Glace: In my test environment I set it up so that all interfaces of both the proxy and the storage nodes are on the same subnet. | 22:04 |
Glace | in tje object.conf for example.. I enter the IP on the storage network | 22:04 |
*** milner has joined #openstack | 22:04 | |
*** milner has quit IRC | 22:04 | |
*** milner has joined #openstack | 22:04 | |
RichardRaseley | object.conf? | 22:04 |
Glace | object-server.conf | 22:05 |
*** llang629_ has joined #openstack | 22:05 | |
Glace | and other storage conf files | 22:05 |
Glace | you can BIND your IP there | 22:05 |
*** melmoth has quit IRC | 22:05 | |
*** sacharya has joined #openstack | 22:05 | |
RichardRaseley | I don't see an object-server.conf file on my proxy anywhere. | 22:05 |
Glace | on your object server | 22:05 |
*** KarinLevenstein has quit IRC | 22:05 | |
*** rnirmal has quit IRC | 22:06 | |
Glace | Ohh I see your question! | 22:06 |
RichardRaseley | But what tells the proxy server which IPs to use to get to the storage nodes? | 22:06 |
*** llang629_ has left #openstack | 22:06 | |
Glace | Well. im not sure you can do what you ask | 22:06 |
RichardRaseley | OK, Imagine this: | 22:06 |
Glace | you want a different network only for replication | 22:06 |
Glace | ? | 22:06 |
*** jemartin_ has joined #openstack | 22:06 | |
RichardRaseley | Load Balancer -> [Network 1] -> Swift Proxy -> [Network 2] -> Swift Store | 22:07 |
*** rnorwood has quit IRC | 22:07 | |
Glace | Ya | 22:07 |
RichardRaseley | So the two networks are two different subnets. | 22:07 |
RichardRaseley | Then the swift object servers do storage traffic over [Network 3] which is yet another subnet. | 22:07 |
*** maplebed has joined #openstack | 22:08 | |
*** maplebed has quit IRC | 22:08 | |
*** maplebed has joined #openstack | 22:08 | |
Glace | Good question.. not sure you can do that | 22:08 |
*** natea has joined #openstack | 22:08 | |
RichardRaseley | So basically I am saying on my proxy (because I did it all in one network in my test setup) the only IPs that the swift proxy knows about are those in the ring files that are really the IPs that are used for storage replication (I think). | 22:08 |
Glace | The proxy needs access to the network that is in the ring file | 22:08 |
RichardRaseley | Ah, OK | 22:08 |
Glace | again I might be wrong.. but I dont see how | 22:08 |
RichardRaseley | Isn't the network defined in the ring file the "storage" network? Or is that not correct? | 22:09 |
Glace | yes exactly | 22:09 |
Glace | so on my proxy servers | 22:09 |
Glace | I have 2 interfaces.. one on the access network and one on the storage network | 22:09 |
*** cmagina has quit IRC | 22:09 | |
Glace | so my proxy can talk to the storage/ring network | 22:09 |
RichardRaseley | Hmmm... | 22:10 |
Glace | So it talks to the storage node.. on the same network that replicates.. but if you set it up well that shouldnt be too much of a problem | 22:10 |
*** jemartin has quit IRC | 22:11 | |
*** jemartin_ is now known as jemartin | 22:11 | |
RichardRaseley | OK. It is just odd because in the setup walk through @ openstack.org it says you should have a dedicated storage network seperate from that which the proxy uses (looking for the diagram). | 22:11 |
*** lborda has joined #openstack | 22:11 | |
Glace | yes.. well thats what I do | 22:11 |
*** cmagina has joined #openstack | 22:12 | |
Glace | I have a didcated storage network.. but the proxy server needs to talk to it anyhow | 22:12 |
Glace | thats why I have 2 different interfaces | 22:12 |
*** Turicas has joined #openstack | 22:13 | |
RichardRaseley | Glace: I think I was confused by looking at this: http://docs.openstack.org/trunk/openstack-compute/install/content/example-object-storage-installation-architecture.html | 22:13 |
RichardRaseley | Doesn't it look like there is a seperate network on the back-side of the storage nodes there? | 22:13 |
*** hggdh has quit IRC | 22:14 | |
RichardRaseley | I get what you are saying though - I guess I can't think of the downside to that unless replication traffic was so high that it impacted the ability of the nodes to service proxy requests? | 22:14 |
Glace | Yes.. well if you check the diagram.. proxy still need access to storage.. | 22:15 |
Glace | Exactly Richard.. that will happen if you do not operate it properly :) | 22:15 |
Glace | In the small.. swift is hard to operate.. on a large cluster you wont be affected too much | 22:15 |
Glace | You can find creative ways to.. add nodes in a small cluster.. like adding the drives with a low weight.. and increase later.. etc | 22:16 |
*** rendar has quit IRC | 22:16 | |
Glace | dont let a node crapped down too long .. etc etc | 22:16 |
*** bbcmicrocomputer has joined #openstack | 22:16 | |
RichardRaseley | I just though it might be worthwhile to explore separating the networks that the proxy uses to talk to the nodes and that which the nodes use to talk to each other *but it looks like that may not be the case). | 22:16 |
RichardRaseley | Glace: Well, we are in a situation where we will have up to 1M file creations per day (several hundred GB). | 22:17 |
RichardRaseley | And that will (of course) grow even more. | 22:17 |
*** cmagina_ has joined #openstack | 22:17 | |
RichardRaseley | So I am just trying to future-proof it without overengineering it. =X | 22:17 |
Glace | Can I ask where you work? | 22:17 |
RichardRaseley | I appreciate your input. | 22:17 |
Glace | hehe ok.. but if you have a fast enought cluster.. your replications shouldnt be queued | 22:17 |
Glace | That depends.. on many things though :) | 22:17 |
*** cmagina has quit IRC | 22:18 | |
*** dolphm has quit IRC | 22:18 | |
RichardRaseley | Glace: I work for a company called Concur Technologies. This is just one of many pilot programs we are testing to deal with this volume / capacity issues we are having. | 22:19 |
Glace | lol ok :) | 22:19 |
kokyu | wtf, sometimes dns_domains sql table doesn't exist, and in between the nova.projects table did not exist. (I recreated the whole database for the tests) - what could cause this miss ? | 22:19 |
*** s0mik has joined #openstack | 22:19 | |
RichardRaseley | Not so much an "issue" now as, will become an issue in the future. | 22:19 |
Glace | hehe | 22:20 |
Glace | well.. A swift cluster is all about operation creativity :) | 22:20 |
Glace | If you have good operations and a good architecture.. you should do well | 22:20 |
RichardRaseley | So you are running Swift in a production environment right now? | 22:21 |
Glace | In Beta | 22:21 |
RichardRaseley | Can you give me an idea of your capacity? | 22:21 |
RichardRaseley | Ah | 22:21 |
uvirtbot | New bug: #996217 in nova "nova.log.JSONFormatter uses record.processName" [Undecided,New] https://launchpad.net/bugs/996217 | 22:21 |
RichardRaseley | Single or multi datacenter? | 22:21 |
RichardRaseley | (or site) | 22:21 |
Glace | Multi | 22:21 |
*** shang has joined #openstack | 22:21 | |
*** anderstj has quit IRC | 22:21 | |
RichardRaseley | Hey maybe you can give me an idea about one other thing. =] | 22:21 |
koolhead17 | Ryan_Lane, hey | 22:22 |
RichardRaseley | I am looking at Keystone for identity management, and trying to figure out how that will work in a multi-site environemnt. | 22:22 |
Ryan_Lane | howdy | 22:22 |
RichardRaseley | Are you using it? | 22:22 |
koolhead17 | where are you? :) | 22:22 |
Glace | Nope.. we have our own auth system | 22:22 |
RichardRaseley | Hmm... something custom? Or you use LDAP, AD, etc? | 22:22 |
Glace | Contacting a 3rd party auth system yes | 22:23 |
Glace | not LDAP or AD though | 22:23 |
Glace | but it could be used for that with some modifications | 22:23 |
Ryan_Lane | koolhead17: not at UDS | 22:23 |
Ryan_Lane | I may only show up one day | 22:23 |
koolhead17 | cool. :) | 22:23 |
kokyu | "ERROR: n/a (HTTP 400)" is not human-understandable. It should say "Could not connect to nova-api: Connection refused." (or similar) instead. | 22:24 |
RichardRaseley | Glace: Do you call that 3rd party auth system directly from the proxies or do you use a custom piece as an intermediary? | 22:25 |
Glace | https://github.com/cloudops/cs_auth | 22:25 |
Glace | It calls the 3rd party system directly.. (API CALLS) | 22:25 |
*** melmoth has joined #openstack | 22:26 | |
RichardRaseley | Ah, I see. Thanks. | 22:26 |
*** xmltok has quit IRC | 22:26 | |
*** xmltok has joined #openstack | 22:26 | |
*** xmltok has joined #openstack | 22:27 | |
*** lazyshot has quit IRC | 22:27 | |
*** xmltok has joined #openstack | 22:27 | |
*** dachary has joined #openstack | 22:28 | |
*** dachary has joined #openstack | 22:28 | |
*** alrs has joined #openstack | 22:30 | |
*** Leseb has quit IRC | 22:31 | |
koolhead17 | annegentle, hey | 22:32 |
*** Leseb has joined #openstack | 22:32 | |
*** vitiho has quit IRC | 22:33 | |
*** vitiho has joined #openstack | 22:33 | |
*** sniperd has joined #openstack | 22:34 | |
*** judu has quit IRC | 22:36 | |
zykes- | koolhead17: :D | 22:36 |
koolhead17 | hey zykes- :) | 22:36 |
*** sniperd_ has joined #openstack | 22:37 | |
*** oubiwann has quit IRC | 22:37 | |
*** sniperd has quit IRC | 22:38 | |
kokyu | what happens if I actively shutdown an instance? is it is still marked as ACTIVE in the UI (and command-line). will it auto-restart after some time or can I enforce it? | 22:42 |
kokyu | I hoped to see the status reflected in the dashboard/commandline, but it isn't (yet) | 22:42 |
*** dtroyer is now known as dtroyer_zzz | 22:42 | |
monst | I think I figured out part of my cpu_allocation_ratio issue, the Dashboard is reporting "Hypervisor: QEMU( rdtscp, pdpe1gb, x2apic, dca, xtpr, tm2, est, vmx, ds_cpl, monitor, pbe, tm, ht, ss, acpi, ds, vme)" | 22:43 |
monst | instead of kvm hypervisor | 22:43 |
monst | even though --libvirt_type=kvm is set | 22:43 |
monst | =\ | 22:43 |
monst | is there a way to check hypervisor via command line | 22:43 |
uvirtbot | New bug: #995323 in glance "glance-api-caching pipeline should use unauthenticated-context" [Medium,Fix committed] https://launchpad.net/bugs/995323 | 22:46 |
*** dnaori has quit IRC | 22:47 | |
*** alrs has quit IRC | 22:50 | |
*** clopez has quit IRC | 22:51 | |
*** esm_ has quit IRC | 22:51 | |
*** sniperd_ has quit IRC | 22:51 | |
*** RichardRaseley has quit IRC | 22:52 | |
*** bbcmicrocomputer has quit IRC | 22:52 | |
*** Glace has quit IRC | 22:53 | |
*** llang629_ has joined #openstack | 22:56 | |
*** llang629_ has left #openstack | 22:56 | |
*** bbcmicrocomputer has joined #openstack | 22:57 | |
*** rafaduran has quit IRC | 23:00 | |
*** miclorb has joined #openstack | 23:00 | |
kokyu | monst: virsh list --all | 23:00 |
kokyu | (on the compute node in question) | 23:00 |
*** xmltok has quit IRC | 23:01 | |
uvirtbot | New bug: #996233 in python-novaclient "nova and python-novaclient disagree on volumes API URLs" [Undecided,New] https://launchpad.net/bugs/996233 | 23:01 |
*** llang629__ has joined #openstack | 23:02 | |
*** llang629__ has left #openstack | 23:02 | |
*** duerF has joined #openstack | 23:03 | |
*** jemartin has quit IRC | 23:05 | |
*** jemartin has joined #openstack | 23:05 | |
*** trevman2001 has quit IRC | 23:05 | |
uvirtbot | New bug: #995224 in glance "DELETE /v2/images/<image_id> should return 204 No Content" [Low,Fix committed] https://launchpad.net/bugs/995224 | 23:06 |
uvirtbot | New bug: #996238 in glance "Refactor location class" [Undecided,New] https://launchpad.net/bugs/996238 | 23:06 |
*** Turicas has quit IRC | 23:06 | |
monst | kokyu: that lists instances, but does not show what hypervisor is running | 23:07 |
kokyu | monst: oh lol, I misunderstood you ;) | 23:07 |
monst | np =) | 23:08 |
*** melmoth has quit IRC | 23:08 | |
kokyu | monst: try `nova-manage vm list` ;) | 23:08 |
kokyu | and make sure your terminal has enough columns to display nicely ;) | 23:08 |
kokyu | otherwise try: nova-manage vm list | less -S | 23:08 |
*** jgriff has joined #openstack | 23:09 | |
kokyu | monst: | 23:09 |
monst | yeah | 23:09 |
monst | still doesnt show hypervisor it is using =( | 23:10 |
*** albert23 has left #openstack | 23:10 | |
monst | according to dashboard it is using qemu, instead of kvm =( | 23:10 |
*** ahasenack has joined #openstack | 23:11 | |
monst | wanted to verify via command line | 23:11 |
kokyu | you mean the hypervisor *TYPE* | 23:11 |
kokyu | please be exact next time :( | 23:11 |
*** shwetaap has quit IRC | 23:11 | |
*** willaerk has quit IRC | 23:11 | |
kokyu | I thought you meant the name of the hypervisor | 23:11 |
*** cmagina_ has quit IRC | 23:12 | |
kokyu | I usually check via `ps -ef | grep kvm` and if it outputs kvm processes, I know it's using kvm (instead of qemu or alike) | 23:12 |
*** zby_ has joined #openstack | 23:12 | |
*** vitiho has quit IRC | 23:13 | |
*** Leseb has quit IRC | 23:13 | |
*** zby_ has quit IRC | 23:13 | |
*** vitiho has joined #openstack | 23:13 | |
*** dtroyer_zzz is now known as dtroyer | 23:14 | |
*** andreas__ has joined #openstack | 23:14 | |
*** ahasenack has quit IRC | 23:14 | |
*** edygarcia has quit IRC | 23:14 | |
*** pmezard has quit IRC | 23:15 | |
*** andreas__ is now known as ahasenack | 23:15 | |
*** dtroyer is now known as dtroyer_zzz | 23:16 | |
monst | kokyu: thanks, looks like its working | 23:17 |
monst | i tweeted to stackops to see how to update the cpu_allocation_ratio | 23:18 |
monst | hopefully they get back | 23:18 |
*** reed has joined #openstack | 23:19 | |
*** swift-india has quit IRC | 23:19 | |
*** swift-india has joined #openstack | 23:20 | |
*** Pebbles` has quit IRC | 23:26 | |
*** monst has quit IRC | 23:27 | |
*** kokyu has quit IRC | 23:27 | |
*** rgoodwin is now known as rgoodwin_away | 23:28 | |
*** rgoodwin_away is now known as rgoodwin | 23:30 | |
*** ahasenack has quit IRC | 23:30 | |
*** melmoth has joined #openstack | 23:33 | |
*** natea has quit IRC | 23:35 | |
RamJett | Where is a good place to read about the networking part of nova. I think my network part should be easy ' I can do it in libvirt in 5 minutes' but I do not know all these term they are using in the nova.conf "flat_network .. floating .." I've been looking and the docs and just can't get my head around what part of the network is being setup. I is real hard to follow docs when they say 'This is what you add for floating .. ' I do not know wha | 23:36 |
RamJett | t the terms them self mean. | 23:36 |
*** jj0hns0n has quit IRC | 23:45 | |
*** lborda has quit IRC | 23:47 | |
*** swift-india has quit IRC | 23:48 | |
*** vincentricci has quit IRC | 23:48 | |
*** swift-india has joined #openstack | 23:48 | |
*** jj0hns0n has joined #openstack | 23:48 | |
*** dtroyer_zzz is now known as dtroyer | 23:49 | |
*** melmoth has quit IRC | 23:51 | |
*** shang has quit IRC | 23:52 | |
*** reed has quit IRC | 23:52 | |
*** ahasenack has joined #openstack | 23:53 | |
*** bbcmicrocomputer has quit IRC | 23:54 | |
*** koolhead17 has quit IRC | 23:55 | |
salgado | mnewby, ok, just let me know what needs doing and I'll do it :) | 23:56 |
salgado | (oops, wrong channel. was talking about https://review.openstack.org/#/c/7019/) | 23:57 |
*** dtroyer is now known as dtroyer_zzz | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!