*** fysa has quit IRC | 00:02 | |
*** fysa has joined #openstack | 00:02 | |
*** _vinay has quit IRC | 00:02 | |
*** aliguori has joined #openstack | 00:04 | |
*** MotoMilind has quit IRC | 00:08 | |
*** hadrian has joined #openstack | 00:14 | |
*** rds__ has quit IRC | 00:17 | |
*** pmullaney has joined #openstack | 00:19 | |
creiht | jaypipes: it has nothing to do with launchpad, just seems very hackish to me | 00:34 |
---|---|---|
creiht | and I don't like the spurious warnings and errors on every setup | 00:34 |
creiht | I'm just trying to find a better solution | 00:35 |
creiht | and comparing to what most other people do | 00:35 |
creiht | jaypipes: http://wiki.maemo.org/How_to_Internationalize_python_apps | 00:36 |
creiht | is an example of hacking up setup.py | 00:36 |
creiht | I'm also not fond of adding other dependencies unless they are absolutely needed | 00:37 |
creiht | plus the fact that for setup now we are mixing setuptools and some variant of distutils | 00:38 |
creiht | so while I'm not always fond of checking in binaries, what disadvantages do you see there are in checking in the .mo files | 00:39 |
openstackjenkins | Project nova build #790: SUCCESS in 2 min 32 sec: http://jenkins.openstack.org/job/nova/790/ | 00:39 |
openstackjenkins | Tarmac: Fixes issues with describe instances due to improperly set metadata. | 00:39 |
openstackjenkins | ?* Removes image['properties']['type'] | 00:39 |
openstackjenkins | ?* Uses image['container_format'] to key display of type and create ec2 ids. | 00:39 |
openstackjenkins | ?* Defaults to 'ami' if container_format cannot be deduced. This allows | 00:39 |
openstackjenkins | ???bare images to show up in describe instances and be launched even though they | 00:39 |
openstackjenkins | ???are not officially in 'ami' format. | 00:39 |
openstackjenkins | ?* Changes nova-manage register to set proper container format | 00:39 |
openstackjenkins | ?* Fixes tests | 00:39 |
openstackjenkins | ?* Fixes _do_get_kernel_and_ramdisk in openstack api to only try to get them if it is a true 'ami' | 00:39 |
openstackjenkins | * Replaces 'owner_id' with 'project_id' since that is expected by glance code | 00:39 |
openstackjenkins | * Moves the filtering scheme to base image service so all services filter the same way | 00:39 |
creiht | especically if we can find some way to automate it when merges come in for .pos | 00:39 |
creiht | All I want to do is work towards a better solution | 00:41 |
*** adiantum has quit IRC | 00:46 | |
*** adiantum has joined #openstack | 00:50 | |
eday | creiht: can't we just generate .mo's on pkg build/install? I hear you on not wanting to introduce the extra distutils dep, but that doesn't mean we need to checkin generated files | 00:54 |
JordanRinke | so anyone in here explain the difference between glance add, and glance-upload to me? | 00:55 |
*** adjohn has joined #openstack | 00:56 | |
*** adiantum has quit IRC | 00:58 | |
*** dovetaildan has quit IRC | 01:03 | |
*** dovetaildan has joined #openstack | 01:04 | |
*** dovetaildan has quit IRC | 01:04 | |
*** dovetaildan has joined #openstack | 01:04 | |
creiht | eday: the main problem is that distutils doesn't make it easy to interject some random script in the build phase | 01:04 |
*** MarkAtwood has joined #openstack | 01:04 | |
creiht | that's why the need for the hackery | 01:05 |
creiht | if you want to build them at build/install phase | 01:05 |
creiht | there doesn't seem to be an ideal solution, unfortunately | 01:06 |
*** MarkAtwood has left #openstack | 01:07 | |
creiht | eday: I'm also not fond of the output that the distutils.extras produces either: | 01:08 |
creiht | http://paste.openstack.org/show/1130/ | 01:08 |
creiht | look for ERROR and WARNING | 01:08 |
kpepple | can anyone put a second look on bug 750777 ? i have two approve but need another nova-core member to slip in under the gammafreeze | 01:10 |
uvirtbot | Launchpad bug 750777 in nova "Metadata values in api responses should be strings." [Medium,In progress] https://launchpad.net/bugs/750777 | 01:10 |
*** flashn_ has quit IRC | 01:10 | |
*** flashn has joined #openstack | 01:10 | |
*** adiantum has joined #openstack | 01:11 | |
*** mrkodiak has left #openstack | 01:13 | |
eday | creiht: if we can't hack setup.py, then we can always just require two commands to build the pkgs too (babel + setup.py) | 01:15 |
creiht | eday: which I would be fine with, but mtaylor doesn't like that | 01:16 |
eday | creiht: I think he'll like it better than checking in .mo files :) | 01:17 |
creiht | hah | 01:18 |
*** btemple has joined #openstack | 01:18 | |
btemple | hello | 01:18 |
btemple | who | 01:19 |
btemple | hello | 01:20 |
btemple | is anyone here? | 01:20 |
JordanRinke | yes | 01:20 |
winston-d | yum | 01:20 |
winston-d | yup | 01:20 |
creiht | perhaps | 01:20 |
btemple | I have a question. I have a server on a local-only network (between only my client machine and itself, not internet) and was wondering how to install Openstack Compute on it. | 01:21 |
JordanRinke | I believe there are single node install docs out on nova.openstack.org and wiki.openstack.org | 01:22 |
winston-d | you need internet because you'd like to use Ubuntu repository to do installation? or? | 01:22 |
btemple | is there any way to do it without internet... and what virtualization tech is needed | 01:23 |
kpepple | btemple: you can use lxc, kvm, xen, etc for virtualization. assuming you are running a x64 linux box, KVM is probably the easiest | 01:24 |
winston-d | if you'd use xen as hypervisor, not hardware VT is needed ('cos you can use para-virtualization); if you'd use KVM, KVM requires Intel VT-x or AMD-V | 01:24 |
btemple | how can i configure with xen and no internet... i can manually download packages and xfer them if needed | 01:25 |
btemple | the server' | 01:25 |
btemple | s blank | 01:25 |
winston-d | btemple: u have already a xen up & running, right? | 01:26 |
btemple | no it's blank | 01:26 |
btemple | but i can install xen first | 01:26 |
kpepple | btemple: you don't have internet at all ? or just the server doesn't have internet ? | 01:26 |
jaypipes | JordanRinke: glance-upload was a tool added early in the development of Glance as a bit of a thin wrapper around the Glance client lib for uploading images to Glance that follow the EC2 style of construction (bundling ramdisk/kernel into an AMI). The glance CLI client is more general. glance add simply allows you to add an image to Glance, nothing more... glance-upload does a couple things to add specific properties like | 01:26 |
jaypipes | connecting the kernel_id and ramdisk_id image identifiers to the machine image... | 01:27 |
btemple | the server doesn't... my client does when i connect to my other network | 01:27 |
kpepple | btemple: if the client machine has internet, pull down an iso of ubuntu 10.10 and install it on the server. then download the nova source code (bzr, packages or tarball) and ftp it over the server. install then as normal. | 01:27 |
winston-d | aha, for choose a hypervisor at first and then choose one linux distribution that has built-in support for hypervisor. | 01:28 |
btemple | yeah, i'm about to install xen... then do i just go from source | 01:28 |
winston-d | or, you can just setup a proxy on your client and then your server will be able to install all the stuff Xen/KVM/OpenStack from internet | 01:29 |
JordanRinke | jaypipes: so if you wanted to replicate glance-upload using glance add, you would add each image and then set the metadata association? | 01:29 |
winston-d | btemple: for hypervisor, i'd suggest choose a pre-built package against source. | 01:29 |
btemple | i might actually do that with the proxy... any tips on configuration? I need to share from a wireless to a wired card.ok | 01:30 |
winston-d | unless you have some special needs | 01:30 |
btemple | binary on hypervisor it is | 01:30 |
*** dendro-afk is now known as dendrobates | 01:30 | |
*** adiantum has quit IRC | 01:31 | |
btemple | for the proxying... how do i do that | 01:31 |
btemple | from wireless to wired | 01:31 |
winston-d | btemple : well, i don't currently have any BKM for proxy setup, but it'd be easy to find one w/ google | 01:31 |
winston-d | is ur client a Linux machine? | 01:31 |
btemple | yes | 01:32 |
winston-d | maybe you can try NAT with iptables | 01:32 |
JordanRinke | jaypipes: I think what confused wayne and I was the doc shows doing a glance add on a .tar.gz which is similar to how ec2 takes a bundle | 01:33 |
winston-d | something like: iptables -t nat -A POSTROUTING -o eth1 -j MASQUERADE | 01:33 |
jaypipes | JordanRinke: yes, that is correct. | 01:33 |
winston-d | eth1 is the interface has internet access. | 01:33 |
btemple | for me it's wlan0, but i get the point | 01:33 |
jaypipes | JordanRinke: so, we're trying *not* to have Glance's API or CLI interfaces get beholden to the euca2ools/EC2 way of doing things... trying to stay agnostic on that front, which may be where the confusion has come in. | 01:33 |
*** btemple has quit IRC | 01:34 | |
JordanRinke | jaypipes: I think so, so glance add will unpack the .tar.gz but just register the file as a single image | 01:34 |
winston-d | oh... did i just make btemple disconnected? | 01:34 |
winston-d | with iptable rules? | 01:35 |
JordanRinke | how does it handle multiple files being in the tar? just curious | 01:35 |
*** adiantum has joined #openstack | 01:35 | |
jaypipes | JordanRinke: glance add doesn't unpack anything, no. | 01:38 |
*** aliguori has quit IRC | 01:38 | |
jaypipes | JordanRinke: Glance stores whatever you give to it, nothing more, nothing less. It doesn't peek into what you give it to try to understand what it is you're giving to it... | 01:38 |
jaypipes | JordanRinke: though that might change in the future, right now Glance stays quite dumb on the subject. ;) | 01:39 |
jaypipes | JordanRinke: for the case of the euca way of bundling and then uploading the bundle, you will want to either use glance-upload the way Wayne was doing, or alternately, you can do repeated calls to glance add, adding each "piece" of the image, storing various image properties (like image_type, kernel_id, ramdisk_id, etc) for each piece of the image (that is basically what glance-upload does.. | 01:41 |
jaypipes | JordanRinke: that make a bit more sense? | 01:41 |
JordanRinke | that does, what doesn't make sense is the glance doc online, shows doing a glance add with a .tar.gz | 01:42 |
JordanRinke | if glance doesn't process the packed files, that doesn't make much sense to me | 01:43 |
jaypipes | JordanRinke: could you explain a bit more what you mean by "process the packed files"? trying to better understand what glance should be doing... | 01:45 |
*** adiantum has quit IRC | 01:47 | |
jaypipes | JordanRinke: I think perhaps there is some confusion that the glance/glance-upload tools are supposed to be replacements for euca-bundle-image and euca-upload-bundle? That definitely is not the case... | 01:47 |
JordanRinke | jaypipes: just curious if it untars the .tar.gz | 01:47 |
jaypipes | JordanRinke: no, it does not... right now, there's been no reason to untar the tar.gz. Also, assuming tar.gz kinda assumes only the AMI-like image management. Glance is not AMI-specific, remember. | 01:48 |
JordanRinke | right, so i think the example in the doc might be what is causing the problem | 01:49 |
jaypipes | JordanRinke: ok, that's definitely something we can look at editing. which specific doc page are you referring to? | 01:49 |
JordanRinke | jaypipes: http://glance.openstack.org/glance.html | 01:49 |
JordanRinke | look at the add command example | 01:49 |
*** adiantum has joined #openstack | 01:50 | |
jaypipes | JordanRinke: ah, I see. so, perhaps we should change the extension tar.gz to something else... | 01:51 |
JordanRinke | yeah just .img or something | 01:51 |
*** pmullaney has quit IRC | 01:52 | |
JordanRinke | the .tar.gz makes you assume it is going to at least unpack it, if not also process it | 01:52 |
jaypipes | JordanRinke: or just note on that page something like "Glance doesn't know what is contained in your image. You must tell Glance what you are storing by using the container_format and disk_format fields and any additional custom image properties"? | 01:53 |
JordanRinke | that and change the extension, giving it a compressed format in the doc really makes you think it will unpack it | 01:55 |
*** dprince has joined #openstack | 01:55 | |
jaypipes | JordanRinke: k, I'll file a bug on that. | 01:55 |
JordanRinke | as is that example wouldn't work | 01:55 |
uvirtbot | New bug: #754215 in openstack-dashboard "Connection error message on instances page impedes on user experience." [Undecided,New] https://launchpad.net/bugs/754215 | 01:56 |
JordanRinke | even with image flags etc, that would still just download a packed file and try to mount it as an image which obviously doesnt work | 01:56 |
jaypipes | JordanRinke: understood. | 01:56 |
JordanRinke | jaypipes: cool so that makes sense then, the way glance add works | 01:56 |
jaypipes | JordanRinke: feel free to add any additional comments to https://bugs.launchpad.net/glance/+bug/754221 | 01:59 |
uvirtbot | Launchpad bug 754221 in glance "Rework documentation on glance add to be more clear" [Medium,Triaged] | 01:59 |
uvirtbot | New bug: #754218 in nova "Errors when deleting default route in ensure_bridge" [Undecided,In progress] https://launchpad.net/bugs/754218 | 02:01 |
JordanRinke | https://bugs.launchpad.net/glance/+bug/754221 | 02:01 |
uvirtbot | Launchpad bug 754221 in glance "Rework documentation on glance add to be more clear" [Medium,Triaged] | 02:01 |
JordanRinke | err | 02:01 |
*** adiantum has quit IRC | 02:02 | |
JordanRinke | jaypipes: thanks dude, much appreciated for the detail and patience ;) | 02:03 |
jaypipes | JordanRinke: any time. | 02:04 |
jaypipes | JordanRinke: very much appreciate the feedback. | 02:04 |
JordanRinke | we are very actively using glance now so you will def hear from us more :-D | 02:06 |
uvirtbot | New bug: #754221 in glance "Rework documentation on glance add to be more clear" [Medium,Triaged] https://launchpad.net/bugs/754221 | 02:06 |
[1]RickB17 | redbo: are you there? I have am hoping you could look at (http://paste.openstack.org/show/1131/) and see if you can see something that i'm doing wrong while attempting to use cloudfuse | 02:06 |
*** adiantum has joined #openstack | 02:08 | |
*** mahadev has quit IRC | 02:08 | |
*** mray has joined #openstack | 02:11 | |
*** bcwaldon has quit IRC | 02:11 | |
*** Ryan_Lane has quit IRC | 02:14 | |
*** adiantum has quit IRC | 02:16 | |
redbo | [1]RickB17: can you throw a -d on that cloudfuse command to get some debuggy goodness? | 02:17 |
*** adiantum has joined #openstack | 02:22 | |
[1]RickB17 | sure thing | 02:24 |
[1]RickB17 | wasn't aware of that | 02:24 |
*** ChameleonSys has joined #openstack | 02:24 | |
[1]RickB17 | i'll do it now | 02:24 |
[1]RickB17 | http://paste.openstack.org/show/1132/ | 02:24 |
[1]RickB17 | looks like it's not using my auth_url, but going to rackspace | 02:25 |
redbo | op.. authurl all one word | 02:27 |
*** mahadev has joined #openstack | 02:27 | |
[1]RickB17 | ahhh i'll give it ago | 02:27 |
[1]RickB17 | i was going off of another example that must be incorrect | 02:27 |
[1]RickB17 | that was it | 02:28 |
[1]RickB17 | your the man! | 02:28 |
redbo | [1]RickB17: cool | 02:28 |
[1]RickB17 | not sure how i did that | 02:29 |
[1]RickB17 | sorry about that, didn't mean to bother you for a typo... | 02:29 |
redbo | no problem | 02:30 |
*** dprince has quit IRC | 02:30 | |
creiht | [1]RickB17: you made his day by calling him the man :) | 02:30 |
*** mahadev has quit IRC | 02:32 | |
redbo | It's almost like being accepted. | 02:33 |
*** benbenhappy has joined #openstack | 02:38 | |
*** adiantum has quit IRC | 02:39 | |
*** adiantum has joined #openstack | 02:45 | |
*** mahadev has joined #openstack | 02:50 | |
uvirtbot | New bug: #754242 in nova "euca-describe-image-attribute -l fails w/ UnknownError:" [Undecided,In progress] https://launchpad.net/bugs/754242 | 02:51 |
*** fysa has quit IRC | 02:57 | |
*** adiantum has quit IRC | 03:01 | |
*** adiantum has joined #openstack | 03:07 | |
*** AimanA is now known as HouseAway | 03:23 | |
*** adiantum has quit IRC | 03:29 | |
*** mray has quit IRC | 03:30 | |
*** adiantum has joined #openstack | 03:33 | |
*** benbenhappy has quit IRC | 03:42 | |
*** adiantum has quit IRC | 03:46 | |
*** mray has joined #openstack | 03:48 | |
*** adiantum has joined #openstack | 03:58 | |
*** fysa has joined #openstack | 04:03 | |
*** ibatanov has joined #openstack | 04:08 | |
koolhead17 | hi all | 04:09 |
dsockwell | greetings | 04:11 |
koolhead17 | hello dsockwell | 04:12 |
JordanRinke | HELLO! | 04:16 |
*** mahadev has quit IRC | 04:19 | |
dsockwell | i've been looking at openstack's SAN support, and it seems pretty iscsi oriented. Does there exist some FC functionality i'm missing? | 04:20 |
dsockwell | if there's interest in developing some, I have a toy FC switch and openindiana(solaris) box | 04:21 |
*** mahadev has joined #openstack | 04:26 | |
*** adiantum has quit IRC | 04:27 | |
*** adiantum has joined #openstack | 04:40 | |
*** f4m8_ is now known as f4m8 | 04:43 | |
*** mray has quit IRC | 04:44 | |
*** mray1 has joined #openstack | 04:49 | |
*** adjohn has quit IRC | 04:53 | |
*** adjohn has joined #openstack | 04:54 | |
*** adjohn has quit IRC | 04:58 | |
*** adiantum has quit IRC | 05:01 | |
*** dendrobates is now known as dendro-afk | 05:05 | |
*** adiantum has joined #openstack | 05:07 | |
*** lionel has quit IRC | 05:13 | |
*** lionel has joined #openstack | 05:14 | |
*** [1]RickB17 has quit IRC | 05:16 | |
koolhead17 | annegentle, | 05:18 |
*** jeffjapan has quit IRC | 05:21 | |
*** ibatanov has quit IRC | 05:22 | |
*** jeffjapan has joined #openstack | 05:25 | |
*** kashyap has joined #openstack | 05:27 | |
*** mray1 has quit IRC | 05:31 | |
*** CloudChris has joined #openstack | 05:44 | |
*** guigui has joined #openstack | 05:50 | |
HugoKuo__ | hello | 05:53 |
HugoKuo__ | I have a concept question with nova-network.............. | 05:55 |
*** benbenhappy has joined #openstack | 05:55 | |
HugoKuo__ | there's only one NIC on my nova-network host , once nova-network ran up , it will create a bridge interface for flat_netowork. | 05:56 |
*** ramkrsna has joined #openstack | 05:56 | |
*** ramkrsna has joined #openstack | 05:56 | |
HugoKuo__ | eth0 is 192.168.1.2 , nova-network assign a ip 192.168.2.1 for br100.... I know 192.168.2.1 is for instance's gateway.... | 05:57 |
HugoKuo__ | but while I check brctl show , nova-netowork did not add eth0 to br100....... | 05:58 |
*** hadrian has quit IRC | 05:58 | |
HugoKuo__ | did I lose something | 05:58 |
HugoKuo__ | any flag configuration ?............... | 05:58 |
*** hadrian has joined #openstack | 05:58 | |
dsockwell | HugoKuo__: sorry, am not in a position to help you out. You may have to wait a while for someone to show up who can. | 05:59 |
HugoKuo__ | dsockwell , thanks | 06:00 |
*** adiantum has quit IRC | 06:01 | |
*** hadrian has quit IRC | 06:01 | |
*** kashyap has quit IRC | 06:02 | |
*** jeffjapan has quit IRC | 06:02 | |
*** slyphon has quit IRC | 06:03 | |
*** eskp has quit IRC | 06:04 | |
*** allsystemsarego has joined #openstack | 06:05 | |
*** allsystemsarego has joined #openstack | 06:05 | |
*** vernhart1 has quit IRC | 06:05 | |
*** eskp has joined #openstack | 06:09 | |
*** mahadev has quit IRC | 06:09 | |
winston-d | HugoKuo__ : r u using glance? | 06:13 |
*** adiantum has joined #openstack | 06:13 | |
*** eskp has quit IRC | 06:13 | |
HugoKuo__ | no , I just nfs share image from objectstore to compute node... | 06:14 |
HugoKuo__ | for now | 06:14 |
HugoKuo__ | I'll use glance later. | 06:14 |
winston-d | i see. thanks | 06:14 |
HugoKuo__ | facing a rediculous problem now , | 06:14 |
HugoKuo__ | haha | 06:14 |
HugoKuo__ | I think that must bcz I'm not qualify enough :\ | 06:15 |
winston-d | it means u need more time, that's all | 06:15 |
HugoKuo__ | r u good in bridge ? | 06:16 |
winston-d | what's the problem? I'll see what i can do | 06:18 |
koolhead17 | hi winston-d HugoKuo__ :) | 06:19 |
HugoKuo__ | bonjour koolhead | 06:19 |
*** mahadev has joined #openstack | 06:19 | |
winston-d | hi koolhead17 | 06:19 |
*** CloudChris has left #openstack | 06:19 | |
koolhead17 | how have you been guys | 06:20 |
winston-d | im ok. thx | 06:21 |
ttx | vishy: o/ | 06:21 |
*** eskp has joined #openstack | 06:21 | |
*** adiantum has quit IRC | 06:22 | |
*** rchavik has quit IRC | 06:22 | |
HugoKuo__ | learning bridge............ | 06:23 |
*** joearnold has quit IRC | 06:24 | |
*** kashyap has joined #openstack | 06:24 | |
winston-d | anyone here has successfully enable glance as image service? | 06:24 |
HugoKuo__ | I hope that I can test glance later , if I solved my problem .....I would love to test that with you | 06:26 |
koolhead17 | winston-d, you were testing nova on redhat box i remember | 06:27 |
winston-d | koolhead17 : yeah, right | 06:27 |
*** DigitalFlux has quit IRC | 06:27 | |
*** mahadev has quit IRC | 06:28 | |
*** adiantum has joined #openstack | 06:29 | |
*** DigitalFlux has joined #openstack | 06:30 | |
*** DigitalFlux has joined #openstack | 06:30 | |
*** jeffjapan has joined #openstack | 06:30 | |
*** vernhart has joined #openstack | 06:41 | |
*** lionel has quit IRC | 06:57 | |
*** lionel has joined #openstack | 06:58 | |
*** mahadev has joined #openstack | 06:58 | |
*** adiantum has quit IRC | 06:59 | |
*** adiantum has joined #openstack | 07:00 | |
*** mahadev has quit IRC | 07:03 | |
*** rcc has joined #openstack | 07:05 | |
*** kashyap has quit IRC | 07:06 | |
*** DigitalFlux has quit IRC | 07:07 | |
*** flopflip has quit IRC | 07:09 | |
*** flopflip has joined #openstack | 07:10 | |
*** littleidea has joined #openstack | 07:11 | |
*** lionel has quit IRC | 07:11 | |
*** adiantum has quit IRC | 07:14 | |
littleidea | is there a document that describes the openstack api for nova? I know there is ongoing work, but is there a description of the RESTful resources and the format of what each side expects? | 07:15 |
*** lionel has joined #openstack | 07:16 | |
*** kashyap has joined #openstack | 07:16 | |
ironcamel | littleidea: yes | 07:17 |
littleidea | ironcamel: link? | 07:18 |
ironcamel | let me see if i can find it ... i have a pdf of the api | 07:19 |
*** adiantum has joined #openstack | 07:19 | |
ironcamel | littleidea: http://www.rackspace.com/cloud/cloud_hosting_products/servers/api/ | 07:24 |
ironcamel | click on API Developer Guide | 07:24 |
*** omidhdl has joined #openstack | 07:25 | |
*** adiantum has quit IRC | 07:30 | |
*** nacx has joined #openstack | 07:30 | |
*** kashyap_ has joined #openstack | 07:31 | |
*** kashyap has quit IRC | 07:34 | |
littleidea | ironcamel: danke, I found what I wanted. | 07:35 |
*** onlany has joined #openstack | 07:38 | |
*** adiantum has joined #openstack | 07:43 | |
*** ccustine has quit IRC | 07:47 | |
HugoKuo__ | iptables -t nat -A PREROUTING -d 169.254.169.254/32 -p tcp -m tcp --dport 80 -j DNAT --to-destination 192.168.1.1:8773 | 07:52 |
HugoKuo__ | I add this rule into iptables | 07:52 |
HugoKuo__ | I'm sure curl 192.168.1.1:8773 is working , | 07:53 |
HugoKuo__ | Why that "curl 169.254.169.254" do not work ..... is this rule incorrect ? | 07:53 |
*** daveiw has joined #openstack | 07:54 | |
*** adiantum has quit IRC | 07:58 | |
*** dirkx has joined #openstack | 07:58 | |
*** adiantum has joined #openstack | 08:02 | |
openstackjenkins | Project nova build #791: SUCCESS in 2 min 32 sec: http://jenkins.openstack.org/job/nova/791/ | 08:04 |
openstackjenkins | Tarmac: Implement a mechanism to enforce a configurable quota limit for image metadata (properties) within the OS API image metadata controller. | 08:04 |
*** zul has quit IRC | 08:06 | |
*** miclorb_ has quit IRC | 08:07 | |
openstackjenkins | Project nova build #792: SUCCESS in 2 min 33 sec: http://jenkins.openstack.org/job/nova/792/ | 08:11 |
openstackjenkins | Tarmac: corrects incorrect openstack api responses for metadata (numeric/string conversion issue) and image format status (not uppercase) | 08:11 |
*** reldan has joined #openstack | 08:15 | |
*** aloga has quit IRC | 08:17 | |
*** littleidea has quit IRC | 08:20 | |
*** kashyap_ has quit IRC | 08:23 | |
*** bkkrw has joined #openstack | 08:30 | |
*** adiantum has quit IRC | 08:31 | |
*** reldan has quit IRC | 08:32 | |
*** zul has joined #openstack | 08:33 | |
*** adiantum has joined #openstack | 08:34 | |
*** rw4c has joined #openstack | 08:37 | |
*** jeffjapan has quit IRC | 08:38 | |
*** aloga has joined #openstack | 08:41 | |
*** befreax has joined #openstack | 08:45 | |
*** adiantum has quit IRC | 08:45 | |
*** adiantum has joined #openstack | 08:51 | |
*** dirkx has quit IRC | 08:59 | |
*** soosfarm_ has quit IRC | 09:00 | |
*** soosfarm has joined #openstack | 09:02 | |
*** rchavik has joined #openstack | 09:05 | |
*** kashyap has joined #openstack | 09:07 | |
*** adiantum has quit IRC | 09:11 | |
*** adiantum has joined #openstack | 09:14 | |
*** ramkrsna has quit IRC | 09:20 | |
*** nagyz has joined #openstack | 09:22 | |
nagyz | heya | 09:23 |
*** ramkrsna has joined #openstack | 09:23 | |
openstackjenkins | Project nova build #793: SUCCESS in 2 min 27 sec: http://jenkins.openstack.org/job/nova/793/ | 09:24 |
openstackjenkins | Tarmac: Update the describe_image_attribute and modify_image_attribute functions | 09:24 |
openstackjenkins | in the EC2 API so they use the top level 'is_public' attribute of image | 09:24 |
openstackjenkins | objects. This brings these functions in line with the base image service. | 09:24 |
openstackjenkins | Added missing EC2 API unit tests for describing and modifying image attributes. | 09:24 |
HugoKuo__ | failed to get metadata :\ | 09:27 |
*** adiantum has quit IRC | 09:28 | |
*** dirkx has joined #openstack | 09:28 | |
*** citral has joined #openstack | 09:32 | |
*** benbenhappy has left #openstack | 09:32 | |
*** dirkx has quit IRC | 09:35 | |
*** dirkx has joined #openstack | 09:35 | |
*** masumotok has quit IRC | 09:35 | |
*** arun_ has joined #openstack | 09:52 | |
*** reldan has joined #openstack | 09:54 | |
*** adiantum has joined #openstack | 09:58 | |
*** dirkx has quit IRC | 09:59 | |
*** befreax has quit IRC | 10:00 | |
*** adiantum has quit IRC | 10:02 | |
*** MarcMorata has joined #openstack | 10:03 | |
*** adiantum has joined #openstack | 10:08 | |
*** adiantum has quit IRC | 10:13 | |
*** rds__ has joined #openstack | 10:15 | |
ttx | ok, let's freeze this. | 10:15 |
*** arun_ has quit IRC | 10:16 | |
*** arun_ has joined #openstack | 10:17 | |
dsockwell | halte | 10:20 |
dsockwell | hammerzeit | 10:20 |
*** zul has quit IRC | 10:20 | |
*** zul has joined #openstack | 10:23 | |
ttx | GammaFreeze now in effect... https://lists.launchpad.net/openstack/msg01739.html | 10:25 |
* ttx lunches | 10:25 | |
*** adiantum has joined #openstack | 10:26 | |
* koolhead17 looks around | 10:26 | |
*** adiantum has quit IRC | 10:34 | |
*** asksol_ is now known as asksol | 10:51 | |
uvirtbot | New bug: #754509 in nova "terminating instance causes nova-compute hang" [Undecided,New] https://launchpad.net/bugs/754509 | 11:01 |
HugoKuo__ | https://answers.launchpad.net/nova/+question/152089 | 11:27 |
HugoKuo__ | That's a usually issue . I always stuck over there. don't know how to resolve it until now.... | 11:28 |
HugoKuo__ | If anyone would like to tell me some clues , it's thanksful.... | 11:29 |
ttx | soren: I have the nova translations branch up for merging at: https://code.launchpad.net/~ttx/nova/cactus-gamma-translations/+merge/56898 | 11:31 |
*** drico has joined #openstack | 11:33 | |
soren | ttx: Should I just set it to Approved? | 11:36 |
ttx | soren: that's a good qeustion. If it was a straight merge from lp:~nova-core/nova/translations, I'd say yes, go ahead | 11:38 |
ttx | soren: but it's not, so you should review I only touch po/*... | 11:39 |
ttx | and approve :) | 11:39 |
*** RobertLaptop has quit IRC | 11:39 | |
soren | ttx: We should codify that translation imports are exempt from the two-core-approvals rule. | 11:40 |
*** ctennis has quit IRC | 11:40 | |
soren | Just for good measure. | 11:40 |
ttx | soren: indeed. | 11:40 |
ttx | soren: we also need to fix lp:~nova-core/nova/translations fpr stright imports | 11:40 |
ttx | straight | 11:40 |
soren | Eh? | 11:41 |
ttx | lp:~nova-core/nova/translations currently doesn't merge cleanly into lp:nova, hence me fixing conflicts in my branch | 11:41 |
ttx | both branches added po/. | 11:41 |
ttx | soren: we could also wait for a fix for that, and import translations afterwards. | 11:42 |
soren | Too late. | 11:42 |
ttx | heh | 11:42 |
ttx | soren: i'll do Jenkins-based tagging when that hits and builds. | 11:44 |
* soren goes to fetch caffeine | 11:45 | |
HugoKuo__ | so tired , I need take a rest on my weekend . If any suggestion of my issue , plz leave comment over this question https://answers.launchpad.net/nova/+question/152089 | 11:49 |
HugoKuo__ | guys have a nice weekend | 11:50 |
*** HugoKuo__ has quit IRC | 11:50 | |
sandywalsh_ | I wrote some docs yesterday, but how to do I merge it if there is no bug? Should I create a bug for it? | 11:53 |
ttx | sandywalsh_: yes, you can file a wishlist bug for it | 11:53 |
sandywalsh_ | thx ttx | 11:54 |
nagyz | sandywalsh_, would you have some time to talk about rabbitmq and command flow in general? | 11:54 |
sandywalsh_ | nagyz, sure, gimme 5 and I'll be free | 11:54 |
nagyz | sure, take your time | 11:56 |
sandywalsh_ | nagyz ... k, free ... 'sup? | 11:57 |
nagyz | so, I'm trying to understand how could I query all compute nodes | 11:58 |
nagyz | basically I'd like to send all of them a command and get all the replies, and expose that as an xml in the API | 11:58 |
nagyz | I've figured out how I can give back an xml :) | 11:58 |
nagyz | when someone does for example an euca-instances-run with the EC2 API, how is it determined which compute node should do it? | 11:59 |
openstackjenkins | Project nova build #794: SUCCESS in 2 min 28 sec: http://jenkins.openstack.org/job/nova/794/ | 11:59 |
openstackjenkins | Tarmac: Import translations from Launchpad. | 11:59 |
nagyz | if its on a topic, I guess all of them gets the message | 11:59 |
sandywalsh_ | nagyz, ok ... sec, lemme gather some links | 11:59 |
sandywalsh_ | nagyz, they all get the message if you publish on the fan-out exchange, otherwise only 1 worker will pick off the message | 12:00 |
nagyz | so rabbitmq does round-robin on queue subscribers? | 12:00 |
sandywalsh_ | nagyz, for topic exchanges, yes | 12:00 |
sandywalsh_ | nagyz, in nova/scheduler/api.py look at the update_service_capabilities() method | 12:02 |
sandywalsh_ | nagyz, that's used by the compute/network/volume services to publish back to *all* scheduler services using fanout | 12:03 |
sandywalsh_ | nagyz, you would need a similar method to publish to all of the compute services | 12:03 |
*** coreym has joined #openstack | 12:03 | |
nagyz | uhm, makes sense. | 12:03 |
nagyz | how would I gather the responses? | 12:03 |
sandywalsh_ | nagyz, ah, that's the trick, fanout is one-way | 12:04 |
nagyz | exactly my problem :) | 12:05 |
sandywalsh_ | nagyz, so in that case, you wouldn't use the fanout exchange :) You'd have to do a sync call to each compute node. You could do that in parallel using a greenpool. | 12:05 |
nagyz | is there an example function of doing such a sync call? | 12:06 |
sandywalsh_ | nagyz, let me get you an example of using the greenpool | 12:06 |
nagyz | great :) | 12:06 |
uvirtbot | New bug: #754552 in nova "No docs for Zones" [Wishlist,New] https://launchpad.net/bugs/754552 | 12:06 |
sandywalsh_ | nagyz, nova/scheduler/zone_manager.py::_poll_zones() does this | 12:06 |
soren | nagyz: I tried rewinding trunk back to before I did the CA split.. "nova-manage project zipfile" fails for me in the same, expected way for me as it does now. | 12:06 |
soren | nagyz: ..until I run nova-api. | 12:07 |
nagyz | soren yep, if I run nova-api first, it works | 12:07 |
sandywalsh_ | nagyz, but this isn't going a sync rpc call, it's doing a sync nova-client call, but the idea is the same. | 12:07 |
soren | nagyz: ...but you said that didn't used to be necessary. | 12:07 |
nagyz | but it did work before as per the wiki description | 12:07 |
ttx | soren: ok, so... tagging. | 12:07 |
soren | nagyz: Right? | 12:07 |
nagyz | yes. | 12:07 |
nagyz | sandywalsh_, ty, I'll look into it | 12:07 |
soren | nagyz: ..and that's what I'm saying now: When I try with a revision from before my split, it doesn't work for me before I run nova-api. | 12:07 |
ttx | soren: could you hold my hand while I do it ? I'll document in parallel | 12:08 |
soren | ttx: I've never tried it, actually. I'm as curious as you are :) | 12:08 |
nagyz | soren, did you run the CA generation script by hand as per the wiki? | 12:08 |
soren | nagyz: Ah. Err.. No. | 12:08 |
ttx | soren: tss, at least you know the magic URL. | 12:08 |
soren | ttx: Oh, right, yes. | 12:08 |
nagyz | soren, read the wiki page I've linked in the bugreport; the bug appeared following those instructions :) | 12:08 |
*** citral1 has joined #openstack | 12:09 | |
soren | nagyz: a) There are lots of changes to the wiki each and every day. It's not exactly a static reference. | 12:10 |
nagyz | let me recheck the page now, if it's still in the same shape | 12:10 |
soren | nagyz: b) If you want actually help solve this, as small a test case as possible would be very helpful. | 12:10 |
nagyz | you said I should have never run the CA generation script in the first place, right? | 12:10 |
soren | ttx: Which revision are you wanting to tag? | 12:11 |
ttx | 965. | 12:11 |
soren | nagyz: Yeah. nova-api does that. | 12:11 |
nagyz | I can't really give a smaller testcase than in the wiki: fetch trunk, run genrootca.sh, setup nova.conf, then I run nova-manage commands, and project fill fail | 12:11 |
nagyz | so either I shouldn't run the script in the first place, in which case someone needs to update the wiki page | 12:12 |
nagyz | and the bug is an incorrect one | 12:12 |
nagyz | or it should work this way too | 12:12 |
soren | ttx: Ok, so you find the "nova" job run that corresponds to that commit. | 12:12 |
*** citral has quit IRC | 12:12 | |
soren | nagyz: Does it depend on using mysql? | 12:12 |
nagyz | havent tried it without, but I would get not, since it's not related to that at all | 12:13 |
soren | ttx: http://jenkins.openstack.org/job/nova/794/ in our case. | 12:13 |
soren | nagyz: Exactly. So that's a whole bunch of stuff that doesn't need to happen to reproduce it. | 12:13 |
ttx | oh, there is tag here | 12:13 |
nagyz | true enough | 12:13 |
ttx | soren: ack, clicked on Tags. Got an intimidating list of very similar tags. | 12:14 |
soren | But anyways, genrootca.sh doesn't work right now. Sort of. | 12:14 |
soren | It works when nova-api runs it, because it knows which directory it should be run from. | 12:14 |
soren | ttx: Hm.... Yes. That's rather spectacular. | 12:15 |
soren | ttx: Pick one. Let's see :) | 12:15 |
soren | ttx: What's the tag, you're adding? | 12:16 |
ttx | "2011.2gamma1" | 12:16 |
soren | Alright. Go for it. | 12:16 |
soren | If it goes badly, I'll revert it. | 12:16 |
ttx | Tagging 965 (revid: tarmac-20110408115352-xw8kcoiftt7udezx) to 2011.2gamma1 $ bzr tag -r tarmac-20110408115352-xw8kcoiftt7udezx -d lp:nova 2011.2gamma1 | 12:16 |
ttx | Created tag 2011.2gamma1 | 12:16 |
soren | 2011.2gamma1 965 | 12:16 |
soren | Yup. | 12:17 |
soren | \o/ | 12:17 |
ttx | cool. | 12:17 |
soren | That was easy :) | 12:17 |
*** ahale has quit IRC | 12:21 | |
ttx | soren: the whole thing is easier to do while not fighting with flu, amazingly. | 12:22 |
soren | Imagine that :) | 12:22 |
*** coreym has quit IRC | 12:23 | |
*** citral1 has quit IRC | 12:23 | |
* soren is not entirely happy about https://code.launchpad.net/~justin-fathomdb/nova/where-has-my-instance-gone/+merge/55450 having landed | 12:24 | |
*** dendro-afk is now known as dendrobates | 12:25 | |
ttx | Cactus Nova Gamma milestone released at https://launchpad.net/nova/cactus/cactus-gamma | 12:27 |
* ttx moves on to Glance | 12:27 | |
ttx | soren: there is nothing like a "Glance" job in Jenkins where you can access tagging ? | 12:30 |
soren | There should be. | 12:30 |
soren | Uh.. It's *right* there. | 12:31 |
ttx | oh, glance-tarball it is | 12:31 |
soren | No. | 12:31 |
soren | Glance. | 12:31 |
soren | Just glance. | 12:31 |
soren | ?!? | 12:31 |
ttx | ah. doesn't appear under "Glance" | 12:31 |
ttx | got it in the all view. | 12:31 |
ttx | false alarm. | 12:31 |
* soren fixes | 12:31 | |
soren | There. | 12:32 |
ttx | soren: wow: http://jenkins.openstack.org/job/glance/96/tagBuild/? | 12:33 |
ttx | it's not an intimidating list now, it's scary. | 12:33 |
ttx | soren: I guess I should pick one of the 104's in there... | 12:34 |
soren | Hang on. | 12:36 |
*** [1]RickB17 has joined #openstack | 12:37 | |
soren | ttx: Yeah, you're looking for tarmac-20110406203257-flg9tgnyg0ka7evk | 12:38 |
soren | ttx: Got it? | 12:40 |
ttx | yep | 12:41 |
ttx | tagging... | 12:41 |
ttx | Created tag 2011.2gamma1 | 12:42 |
soren | What about Swift? | 12:42 |
*** DigitalFlux has joined #openstack | 12:42 | |
*** DigitalFlux has joined #openstack | 12:42 | |
soren | The use a different versioning scheme still, don't they? | 12:42 |
ttx | soren: right, should be tagged 1.3gamma1 | 12:42 |
ttx | will check their tree for existing tags though | 12:42 |
ttx | just in case somethign else was used already. | 12:43 |
soren | Jenkins shows those as well. | 12:43 |
ttx | Cactus Glance Gamma milestone released at https://launchpad.net/glance/cactus/cactus-gamma | 12:45 |
*** dendrobates is now known as dendro-afk | 12:46 | |
ttx | soren: for swift I'll wait a few more minutes, see if we can fix the translations first | 12:47 |
*** dendro-afk is now known as dendrobates | 12:47 | |
*** coreym has joined #openstack | 12:47 | |
*** kashyap has quit IRC | 12:50 | |
westmaas | termie: is this good now? https://code.launchpad.net/~rackspace-titan/nova/server_metadata_quotas/+merge/55603 | 12:50 |
*** bcwaldon has joined #openstack | 12:51 | |
*** tildedave has quit IRC | 12:53 | |
ttx | creiht, notmyname, jaypipes: let me know if I should release gamma milestone of Swift now or wait to get the i18n fixes first, if they are just around the corner | 12:53 |
*** drico has quit IRC | 12:54 | |
*** drico has joined #openstack | 12:55 | |
BK_man | got a stupid question: how to add a public IP address to Nova instances by default? (Emulate Public IaaS) | 12:55 |
*** tildedave has joined #openstack | 12:55 | |
BK_man | we are using VlanManager and multi-host configuration | 12:56 |
*** h0cin has joined #openstack | 12:56 | |
*** h0cin has joined #openstack | 12:56 | |
*** rw4c has quit IRC | 12:59 | |
creiht | ttx: I would say go ahead and make the gamma release | 13:02 |
*** aliguori has joined #openstack | 13:02 | |
ttx | creiht: ok, we can fix it for RC. | 13:03 |
*** ppetraki has joined #openstack | 13:04 | |
creiht | I'm not sure why it is that critical for release | 13:04 |
creiht | There are no translations currently | 13:05 |
creiht | so even if we had it all figured out, it wouldn't make a difference | 13:05 |
ttx | oh, I didn't know that. | 13:05 |
ttx | creiht: I guess exercising the framework wouldn't hurt. | 13:05 |
creiht | It seems to have never been hooked up in launchpad | 13:05 |
*** pmullaney has joined #openstack | 13:06 | |
ttx | ok, it's up. Sending email. | 13:07 |
dsockwell | i see nova-volume has a SAN driver for iscsi on Solaris. Could that be expanded to FC on Solaris? It's the same framework, but I understand it's lots easier to test iSCSI than FC. | 13:08 |
dsockwell | ethernet switches being more common than fibre channel ones, that is | 13:11 |
*** kakella has joined #openstack | 13:17 | |
creiht | how do I register a blueprint for the design summit that was previously created? | 13:18 |
*** dprince has joined #openstack | 13:19 | |
ttx | creiht: "Propose for sprint" and select design summit diablo | 13:20 |
creiht | ahh ok | 13:20 |
creiht | thx | 13:20 |
*** gondoi has joined #openstack | 13:25 | |
*** ibatanov has joined #openstack | 13:27 | |
*** aryan has quit IRC | 13:29 | |
*** Zangetsu has joined #openstack | 13:30 | |
*** f4m8 is now known as f4m8_ | 13:31 | |
*** mray has joined #openstack | 13:31 | |
*** hadrian has joined #openstack | 13:34 | |
*** Harbringer has joined #openstack | 13:36 | |
*** aryan has joined #openstack | 13:43 | |
*** slyphon has joined #openstack | 13:46 | |
*** kakella has quit IRC | 13:46 | |
*** kakella has joined #openstack | 13:47 | |
*** ctennis has joined #openstack | 13:48 | |
*** ctennis has joined #openstack | 13:48 | |
*** imsplitbit has joined #openstack | 13:49 | |
*** rchavik has quit IRC | 13:54 | |
*** rw4c has joined #openstack | 13:54 | |
*** Harbringer has quit IRC | 13:54 | |
jaypipes | ttx, creiht: uhm, of course there won't be an translations until its set up properly on Launchpad... | 13:57 |
*** nacx has quit IRC | 14:00 | |
*** Harbringer has joined #openstack | 14:01 | |
creiht | jaypipes: I was just making the point that there are currently no translations for swift, so I don't see it as absolutely critical for the i18n fixes to get in for the cactus release | 14:01 |
uvirtbot | New bug: #754651 in nova "bzr ignore the top level CA dir" [Undecided,In progress] https://launchpad.net/bugs/754651 | 14:01 |
jaypipes | creiht: oh, I understand. :) I was just making the point that there won't *be* any translations until the i18n is fixed. ;) | 14:02 |
creiht | jaypipes: I don't understand that, as my understanding is that you just have to point to the .pot file from launchpad, which was my understanding | 14:03 |
creiht | which was has been there for a while now | 14:03 |
*** nacx has joined #openstack | 14:04 | |
*** koolhead17 has quit IRC | 14:04 | |
creiht | just because the auto-building part isn't working, doesn't mean that people can't start translating | 14:04 |
jaypipes | creiht: I see. Well, yes, that is true enough. I'll point the import/export to a branch and see what happens. Not sure how much traction that will get since people can't actually install their translations, but you have a point. | 14:05 |
creiht | jaypipes: people can install their translations, they just have to do it manually ;) | 14:05 |
*** Harbringer has quit IRC | 14:05 | |
jaypipes | creiht: true nuf. | 14:05 |
doude | Hi all, in the LiveMigration wiki, the sample configuration uses a NFSv4 share FS. Is it necessary or NFSv3 is sufficient ? | 14:08 |
*** omidhdl has left #openstack | 14:09 | |
notmyname | creiht: I got pulled in a few directions by kids this morning. thanks for answering that stuff | 14:11 |
jaypipes | oh ffs | 14:13 |
jaypipes | creiht: I'm a damn idiot. | 14:13 |
* jaypipes wait's for the "yeah, we knew that..." | 14:14 | |
jaypipes | creiht: I had already set up the translation export branch but had it set to "Import None" instead of "Import Translation templates"... | 14:15 |
*** RobertLaptop has joined #openstack | 14:15 | |
creiht | lol | 14:15 |
ttx | None is good for you. | 14:15 |
jaypipes | creiht: flipped that switch. hopefully we'll see the Rosetta system kick into gear now... | 14:16 |
creiht | cool | 14:16 |
jaypipes | creiht: sorry for all this... gah I feel like Chris Farley in Tommy Boy. | 14:16 |
*** blamar has quit IRC | 14:18 | |
creiht | jaypipes: no worries | 14:18 |
*** MotoMilind has joined #openstack | 14:19 | |
ttx | jaypipes: any bug in Glance you want to push on the RC list ? | 14:20 |
jaypipes | ttx: https://code.launchpad.net/~dan-prince/glance/fix_update_props/+merge/56865 | 14:21 |
ttx | jaypipes: ack | 14:22 |
*** kashyap has joined #openstack | 14:22 | |
dprince | ttx: I would suggest this as an RC list bug for nova: | 14:23 |
dprince | ttx: https://code.launchpad.net/~dan-prince/nova/ignore_route_del_errors | 14:23 |
jaypipes | dprince: morning. | 14:23 |
dprince | ttx: very safe fix in my oppinion. | 14:24 |
ttx | dprince: right, it's on my list of "potential ones", looking now | 14:24 |
dprince | jaypipes: morning sir. | 14:24 |
ttx | dprince: done | 14:25 |
dprince | ttx: That one is more important to me since it breaks the dev/testing environments we use on team titan. | 14:25 |
dprince | ttx: great. Thanks. | 14:25 |
*** MotoMilind has quit IRC | 14:25 | |
jaypipes | ttx: if vishy is ok with it, I would vote this one for RC, too: https://code.launchpad.net/~sandy-walsh/nova/zone-docs/+merge/56924 | 14:30 |
ttx | jaypipes: it's rc already. | 14:31 |
jaypipes | ttx: ah, sorry. | 14:31 |
* ttx hugs http://wiki.openstack.org/reviewslist/ | 14:31 | |
*** ibatanov has quit IRC | 14:33 | |
*** mahadev has joined #openstack | 14:34 | |
*** daveiw has left #openstack | 14:34 | |
*** nelson has quit IRC | 14:35 | |
*** nelson has joined #openstack | 14:35 | |
*** ibatanov has joined #openstack | 14:39 | |
uvirtbot | New bug: #754687 in nova "Change the default 'lock_path' flag value" [Undecided,New] https://launchpad.net/bugs/754687 | 14:41 |
jaypipes | creiht: swift translation template has now been successfully imported by rosetta. | 14:42 |
*** ramkrsna has quit IRC | 14:47 | |
*** blamar has joined #openstack | 14:48 | |
*** zaitcev has joined #openstack | 14:50 | |
*** dragondm has joined #openstack | 14:55 | |
*** mahadev has quit IRC | 14:56 | |
*** guigui has quit IRC | 15:00 | |
*** befreax has joined #openstack | 15:03 | |
*** mahadev has joined #openstack | 15:03 | |
*** mgoldmann has joined #openstack | 15:04 | |
*** mgoldmann has joined #openstack | 15:04 | |
*** JuanPerez has quit IRC | 15:08 | |
*** deshantm_laptop has joined #openstack | 15:11 | |
*** vernhart has quit IRC | 15:12 | |
*** mgoldmann has quit IRC | 15:22 | |
*** rnirmal has joined #openstack | 15:25 | |
BK_man | can somebody give me a sample URL for accessing noVNC console? | 15:26 |
*** mray has quit IRC | 15:26 | |
* BK_man got following from stack command: {u'url': u'http://127.0.0.1:6080/vnc_auto.html?token=4a8beb76-a909-4a98-aaa7-e0167ec61eb5&host=hostignore&port=portignore'} | 15:26 | |
BK_man | is that correct? | 15:27 |
*** JuanPerez has joined #openstack | 15:28 | |
*** mray has joined #openstack | 15:29 | |
*** mray has joined #openstack | 15:29 | |
*** dendrobates is now known as dendro-afk | 15:33 | |
*** j05h has left #openstack | 15:39 | |
*** drico has quit IRC | 15:42 | |
*** drico has joined #openstack | 15:43 | |
*** throughnothing has joined #openstack | 15:43 | |
*** j05h has joined #openstack | 15:44 | |
*** mahadev has quit IRC | 15:48 | |
*** bcwaldon has quit IRC | 15:49 | |
jk0 | soren: mind having a quick look at https://code.launchpad.net/~ken-pepple/nova/lp753416/+merge/56862 ? | 15:49 |
jk0 | soren: wanted you to look before marking Approved | 15:49 |
*** j05h has quit IRC | 15:54 | |
*** j05h has joined #openstack | 15:54 | |
*** smeier00 has joined #openstack | 15:55 | |
*** smeier00 has left #openstack | 15:55 | |
*** rds__ has quit IRC | 15:58 | |
*** deshantm_laptop has quit IRC | 15:59 | |
*** elasticdog has quit IRC | 15:59 | |
*** mray has quit IRC | 15:59 | |
*** elasticdog has joined #openstack | 16:00 | |
*** elasticdog has joined #openstack | 16:00 | |
*** joearnold has joined #openstack | 16:02 | |
*** j05h has quit IRC | 16:02 | |
*** elasticdog has joined #openstack | 16:02 | |
*** j05h has joined #openstack | 16:02 | |
openstackjenkins | Project nova build #795: SUCCESS in 2 min 31 sec: http://jenkins.openstack.org/job/nova/795/ | 16:04 |
openstackjenkins | Tarmac: Ignore errors when deleting the default route in the ensure_bridge function. | 16:04 |
*** johnpur has joined #openstack | 16:04 | |
*** ChanServ sets mode: +v johnpur | 16:04 | |
*** mray has joined #openstack | 16:05 | |
*** bcwaldon has joined #openstack | 16:10 | |
*** mahadev has joined #openstack | 16:12 | |
openstackjenkins | Project nova build #796: SUCCESS in 2 min 34 sec: http://jenkins.openstack.org/job/nova/796/ | 16:14 |
openstackjenkins | Tarmac: This branch fixes https://bugs.launchpad.net/nova/+bug/751242. | 16:14 |
openstackjenkins | According to the above URL, the solution is 'migrating' instances never be terminated. | 16:14 |
uvirtbot | Launchpad bug 751242 in nova "migrating instance can be terminated, and data inconsistency occurs" [Low,In progress] | 16:14 |
openstackjenkins | To do this, nova.compute.api.delete() should be modified. | 16:14 |
*** nacx has quit IRC | 16:16 | |
*** jernst has joined #openstack | 16:17 | |
jernst | co | 16:17 |
*** jernst has left #openstack | 16:17 | |
*** jernst has joined #openstack | 16:17 | |
jernst | Newbie question if you don't mind: right now OpenStack is compute and storage it seems. Is anybody working on anything beyond that? | 16:20 |
*** befreax has quit IRC | 16:24 | |
*** adjohn has joined #openstack | 16:24 | |
*** dendro-afk is now known as dendrobates | 16:27 | |
*** nagyz has quit IRC | 16:29 | |
*** MarcMorata has quit IRC | 16:29 | |
[1]RickB17 | redbo: is there any reason that a mounted cloudfuse directory cannot be accessed via a samba share? | 16:30 |
*** bkkrw has quit IRC | 16:31 | |
dprince | ttx: This might be another good one to fix for cactus: | 16:33 |
dprince | ttx: https://code.launchpad.net/~dan-prince/nova/fix_gen_x509_redhat/+merge/56982 | 16:33 |
*** kashyap has quit IRC | 16:34 | |
*** rcc has quit IRC | 16:40 | |
*** dspano has joined #openstack | 16:40 | |
jernst | Newbie question if you don't mind: right now member:OpenStack is compute and storage it seems. Is anybody working on anything beyond that? | 16:44 |
*** mray has quit IRC | 16:44 | |
*** mray has joined #openstack | 16:45 | |
[1]RickB17 | jernst: I'm new to the community as well. What would be an example of "beyond" be? | 16:49 |
*** kashyap has joined #openstack | 16:50 | |
jernst | Well, take Amazon as an analogy. Openstack compute roughly equals EC2, storage S3, but AWS has a lot more tabs lately. | 16:52 |
jernst | Like load balancing, relational database service, payments etc. | 16:53 |
[1]RickB17 | jernst: gotcha. I don't know if thats in their plans. Someone here might have an idea though. I was just curious what you ment. | 16:55 |
* nhm_ gives up on vlan networking, tries flatmanager | 16:55 | |
nhm_ | jernst: new as well. I imagine right now they are mostly just working on getting everything working well. Lots of bugfixing still going on right now. | 16:56 |
jernst | What I'm really trying to find out is: what is OpenStack? Where does it end? Who, if any, does the stuff that needs to get done beyond OpenStack's end? | 16:57 |
nhm_ | jernst: they seem to follow a pretty agile devel process, so I imagine those kinds of things are still a bit nebulous as they focus on the immediate 3 month release cycle goals. | 16:58 |
*** Ryan_Lane has joined #openstack | 17:02 | |
jernst | Well, I guess either they don't know or they won't tell ;-) | 17:04 |
johnpur | jernst: There is nothing secret going on... there are a number of projects that are bubbling under the surface that will likely be proposed to OpenStack in the future | 17:09 |
johnpur | Check the archives for "Burrow", a simple queueing project that eday is building | 17:10 |
johnpur | At the upcoming design summit there will be a lot of discussion about network as a service | 17:10 |
johnpur | Also, we expect that block storage (ala EBS) will be a topic of discussion as well | 17:11 |
*** zul has quit IRC | 17:11 | |
johnpur | Other community members are working on load balancing solutions and database as a service (ala RDS) | 17:11 |
*** arun_ has quit IRC | 17:12 | |
johnpur | Lots of activity, as you can see! | 17:12 |
jernst | So these other "community members", are they doing this as part of the OpenStack project, our outside? (Where?) | 17:17 |
eday | jernst: At least for burrow (queue/event system) I'm doing it uner openstack, but it's not proposed for an official openstack project yet. That will happen once there is something to evaluate | 17:18 |
kpepple | jernst: there's a wiki page listing the "other" projects ... let me pull it up | 17:18 |
kpepple | jernst: see http://wiki.openstack.org/Projects?highlight=%28create%29%7C%28project%29 | 17:20 |
jernst | kpepple: thanks for the link, that's interesting. | 17:21 |
kpepple | jernst: np | 17:21 |
johnpur | jernst: a combination of the two... some have existing projects they will open source and propose to OpenStack, others will be built from day 1 in the open as OpenStack projects | 17:23 |
*** arun_ has joined #openstack | 17:24 | |
*** arun_ has joined #openstack | 17:24 | |
*** zigo-_- has joined #openstack | 17:24 | |
eday | soren: ? | 17:24 |
*** zul has joined #openstack | 17:25 | |
jernst | johnpur: So the idea is that community members will write something (design, preferably code) when they see a need for something, publish it for review, and then the project policy board (?) decides whether to make it an official project? | 17:25 |
*** reldan has quit IRC | 17:25 | |
kpepple | jernst: http://wiki.openstack.org/Governance/Approved/NewProjectProcess | 17:26 |
eday | jernst: Pretty much. I also wrote up something about what I'm doing with burrow, the queue service: http://oddments.org/?p=591 | 17:27 |
*** masumotok has joined #openstack | 17:27 | |
*** dendrobates is now known as dendro-afk | 17:28 | |
*** tester23 has joined #openstack | 17:29 | |
*** arun_ has quit IRC | 17:29 | |
tester23 | hi all, have a bit of a n00b question | 17:30 |
kpepple | tester23: those are the best kind | 17:30 |
tester23 | whenever I try to run nova-manage, it'll tell me ImportError: No module named paste | 17:30 |
tester23 | am using the current nova trunk | 17:31 |
kpepple | tester23: did you install the pip-required modules ... specifically paste ? | 17:31 |
*** dendro-afk is now known as dendrobates | 17:33 | |
*** reldan has joined #openstack | 17:34 | |
*** masumotok_ has joined #openstack | 17:34 | |
tester23 | that fixed it :) | 17:34 |
tester23 | It's not listed in http://wiki.openstack.org/InstallFromSource | 17:35 |
eday | soren: when you reemerge, MPs 56862 and 56821 are waiting for your blessing (tied to cactus-rc bugs) | 17:35 |
kpepple | tester23: did you install from the package (python-paste) or just with a "pip install paste" ? i'll redline the wiki page | 17:36 |
tester23 | package | 17:36 |
*** masumotok has quit IRC | 17:36 | |
tester23 | I dont have pip | 17:36 |
kpepple | tester23: so you used "sudo apt-get install python-paste" ? | 17:36 |
tester23 | yep and python-pastedeploy | 17:37 |
tester23 | but now its telling me im missing novaclient | 17:37 |
kpepple | tester23: i think we moved nova-client to a separate package recently | 17:37 |
tester23 | which i guess i'll need pip for | 17:37 |
tester23 | to install | 17:38 |
kpepple | tester23: check and make sure that we don't have a ubuntu package first | 17:38 |
kpepple | tester23: okay, we don't | 17:39 |
tester23 | so did a apt-get install pip | 17:40 |
tester23 | and a pip install novaclient | 17:40 |
tester23 | pip install python-novaclient that is | 17:41 |
*** adiantum has joined #openstack | 17:42 | |
*** clauden_ has joined #openstack | 17:43 | |
BK_man | what is a right way to upload an instance image to Nova at the moment? I'm using glance and euca-upload-bundle failed | 17:43 |
*** arun_ has joined #openstack | 17:44 | |
*** jernst has quit IRC | 17:44 | |
BK_man | this is what I got: http://paste.openstack.org/show/1146/ | 17:45 |
*** mgoldmann has joined #openstack | 17:45 | |
eday | mtaylor: hey! how's that scp target going? :) | 17:46 |
*** adiantum has quit IRC | 17:47 | |
dprince | BK_man: uec-publish-tarball? | 17:47 |
dprince | BK_man: you can also just use glance on the command line. | 17:48 |
BK_man | dprince: that's my script: http://paste.openstack.org/show/1147/ | 17:48 |
*** openstackjenkins has quit IRC | 17:49 | |
BK_man | dprince: it was working today's morning. Seems that last builds of Nova changed a bit | 17:49 |
*** openstackjenkins has joined #openstack | 17:49 | |
mtaylor | eday: done | 17:50 |
*** arun_ has quit IRC | 17:50 | |
eday | mtaylor: awesome, thanks! testing now... | 17:51 |
openstackjenkins | Project burrow build #1: SUCCESS in 12 sec: http://jenkins.openstack.org/job/burrow/1/ | 17:51 |
eday | mtaylor: so, it scp'd somewhere, but http://burrow.openstack.org/ doesn't have the docs | 17:51 |
dprince | BK_man: looking... | 17:52 |
*** adiantum has joined #openstack | 17:53 | |
*** mgoldmann has quit IRC | 17:53 | |
BK_man | dprince: lines 24&25 are debug lines. line 23 failed with previous paste | 17:53 |
mtaylor | eday: ok. refresh your browser (I didn't have the vhost set up) | 17:54 |
*** imsplitbit has quit IRC | 17:54 | |
eday | mtaylor: \o/ | 17:54 |
*** dendrobates is now known as dendro-afk | 17:55 | |
*** RJD22 is now known as RJD22|away | 17:55 | |
eday | mtaylor: thanks, I promise not to bug you for at least another hour now :) | 17:55 |
mtaylor | eday: hehe. bug any time! | 17:55 |
*** lionel has quit IRC | 17:55 | |
mtaylor | :) | 17:55 |
*** adjohn has quit IRC | 17:55 | |
*** lionel has joined #openstack | 17:56 | |
*** johnpur has quit IRC | 17:57 | |
*** RJD22|away is now known as RJD22 | 18:01 | |
coreym | who is able to change the howto wiki? | 18:01 |
*** vernhart has joined #openstack | 18:02 | |
*** masumotok_ has quit IRC | 18:02 | |
kpepple | coreym: wiki.openstack.org ? | 18:02 |
*** clauden_ has quit IRC | 18:05 | |
*** lionel has quit IRC | 18:05 | |
*** lionel has joined #openstack | 18:05 | |
*** RJD22 is now known as RJD22|away | 18:06 | |
coreym | kpepple: swift.openstack.org/development_saio.html | 18:07 |
coreym | maybe not the wiki i guess | 18:07 |
kpepple | coreym: that's the docs, not the wiki | 18:07 |
coreym | sure | 18:07 |
coreym | i want to suggest some changes | 18:08 |
kpepple | coreym: to change that you need to pull the openstack-manuals repo | 18:08 |
dprince | BK_man: I'm getting errors uploading images as well. Still looking into it... | 18:08 |
*** RJD22|away is now known as RJD22 | 18:08 | |
tester23 | hey ken... so some packages seem to be missing on that wiki page | 18:08 |
*** throughnothing has quit IRC | 18:08 | |
BK_man | dprince: thanks for mentioning. I'm now switching to nova.image.local.LocalImageService to check | 18:08 |
kpepple | tester23: which wiki page ? | 18:09 |
tester23 | paste, novaclient and a bunch more | 18:09 |
tester23 | http://wiki.openstack.org/InstallFromSource| | 18:09 |
openstackjenkins | Project nova build #797: SUCCESS in 2 min 27 sec: http://jenkins.openstack.org/job/nova/797/ | 18:09 |
openstackjenkins | Tarmac: Added a patch for python eventlet, when using install_venv.py (see FAQ # 1485) | 18:09 |
coreym | kpepple: i dont really want to change it myself, I want to discuss with someone in the know first. i dont feel like i have the expertise yet | 18:09 |
kpepple | tester23: i just added the paste and paste-deploy stuff | 18:09 |
kpepple | coreym: ping annegentle ... she heads the docs | 18:10 |
coreym | kpepple: thx | 18:10 |
tester23 | cheers... theres a couple more python modules missing though | 18:11 |
tester23 | ie migrate | 18:11 |
tester23 | Not sure what the exact packages are but ended up copypasting the whole bunch from nova.sh | 18:11 |
kpepple | tester23: hmmm ... yeah, i don't really agree with how they are installing here. the easier way (assuming you are going the source route), is to just install pip and then do a "pip install -r tools/pip-requires" which would pull everything down at once | 18:12 |
tester23 | yep lets put that in the wiki instead | 18:14 |
tester23 | at least that works :) | 18:14 |
*** arun_ has joined #openstack | 18:15 | |
tester23 | kpepple: and pip also needs swig to install M2Crypto | 18:16 |
kpepple | tester23: it used to be that way ... but apparently people wanted to go another way :( i'll ping annegentle about it | 18:17 |
*** dprince has quit IRC | 18:21 | |
*** BK_man has quit IRC | 18:24 | |
eday | Any nova-core folks want to review: https://code.launchpad.net/~sandy-walsh/nova/zone-docs/+merge/56924 ? | 18:26 |
*** dendro-afk is now known as dendrobates | 18:29 | |
coreym | any SWIFT AIO person willing to discuss ~/bin/resetswift ? | 18:33 |
uvirtbot | New bug: #754900 in nova "Nova-manage network delete does not delete from fixed_ips" [Undecided,New] https://launchpad.net/bugs/754900 | 18:36 |
*** adiantum has quit IRC | 18:43 | |
*** lionel has quit IRC | 18:44 | |
*** lionel has joined #openstack | 18:44 | |
*** adiantum has joined #openstack | 18:45 | |
*** Harbringer has joined #openstack | 18:48 | |
uvirtbot | New bug: #754908 in nova "uec-publish-tarball hangs when uploading bundle/kernel" [Undecided,New] https://launchpad.net/bugs/754908 | 18:52 |
justinsb | Has anyone found any good instructions for how to set up networking? | 18:52 |
*** dendrobates is now known as dendro-afk | 18:54 | |
*** reldan has quit IRC | 19:01 | |
*** zk has joined #openstack | 19:05 | |
*** zaitcev has quit IRC | 19:05 | |
uvirtbot | New bug: #754924 in nova "When creating a network for FlatDHCPManager, flat_network_dhcp_start must be set first" [Undecided,New] https://launchpad.net/bugs/754924 | 19:06 |
nhm_ | justinb: not really, but I'm moving away from vlan networking to just flat networking since vlan networking seems to be screwing up routing in mysterious ways. | 19:07 |
*** jonkelly has joined #openstack | 19:08 | |
justinsb | nhm_: Thanks - I've been told that FlatDHCPManager is the best, but I'm struggling to get it to work | 19:08 |
nhm_ | justinb: I haven't tried flat dhcp yet. I'm think straight flat since it apperas to try to do the least automatically. | 19:09 |
nhm_ | justinb: I figure that gives me a better chance of making it work. ;) | 19:09 |
justinsb | nhm_: Good luck to you! | 19:10 |
*** hazmat has quit IRC | 19:12 | |
nhm_ | justinb: you too. If you getf flatdhcp working in a multi-node configuration I'd love to see your configuration. :) | 19:12 |
*** adiantum has quit IRC | 19:12 | |
*** kakella has left #openstack | 19:13 | |
*** bkkrw has joined #openstack | 19:15 | |
*** kakella has joined #openstack | 19:16 | |
uvirtbot | New bug: #754944 in nova "self.interfaces_xml is assigned in LibvirtConnection but never used" [Undecided,New] https://launchpad.net/bugs/754944 | 19:21 |
*** kashyap has quit IRC | 19:21 | |
*** johnpur has joined #openstack | 19:22 | |
*** ChanServ sets mode: +v johnpur | 19:22 | |
*** adiantum has joined #openstack | 19:22 | |
*** hazmat has joined #openstack | 19:24 | |
*** mftx has joined #openstack | 19:25 | |
*** dubsquared has joined #openstack | 19:25 | |
*** ibatanov has quit IRC | 19:26 | |
dubsquared | Having an issue with HyperV, anyone seen this? http://paste.openstack.org/show/1149/ | 19:26 |
dubsquared | vish? :D | 19:26 |
*** shentonfreude has joined #openstack | 19:28 | |
justinsb | dubsquared: 2011-04-08 12:23:21,799 DEBUG nova.utils [-] Inner Exception: No module named no | 19:30 |
justinsb | vaclient from (pid=1836) import_class c:\openstack\nova\nova\utils.py:65 | 19:30 |
justinsb | dubsquared: Is novaclient installed and accessible? | 19:30 |
dubsquared | running from source | 19:30 |
*** adiantum has quit IRC | 19:31 | |
uvirtbot | New bug: #754950 in nova "When files are injected into a disk image, they should have a comment e.g. "Injected during boot by the OpenStack cloud"" [Undecided,New] https://launchpad.net/bugs/754950 | 19:31 |
justinsb | I've seen some weird stuff when an import fails, Python just gives an unrelated error message | 19:31 |
*** vernhart has quit IRC | 19:34 | |
dubsquared | yeah…i guess this doc was written for bexar | 19:35 |
dubsquared | and i just pulled down latest trunk | 19:35 |
uvirtbot | New bug: #754959 in nova "Compute node hanging while updating MySQL DB on instance shutdown" [Undecided,New] https://launchpad.net/bugs/754959 | 19:36 |
*** vernhart has joined #openstack | 19:36 | |
blamar | justinsb, any chance that bug is related to https://bugs.launchpad.net/nova/+bug/754509 ? | 19:37 |
uvirtbot | Launchpad bug 754509 in nova "terminating instance causes nova-compute hang" [Undecided,New] | 19:37 |
justinsb | blamar: Good call. I think it's probably the same bug | 19:39 |
justinsb | blamar: My guess is that we broke the SQL Alchemy model | 19:40 |
justinsb | blamar: I can probably guess which MP it was as well - probably the instance_type_id one, because I don't think we otherwise changed the model recently | 19:40 |
blamar | justinsb: yeah, still trying to figure out how the code worked in the first place :) | 19:41 |
justinsb | blamar: Oh - can you see anything wrong with it? My stack trace is deep in the bowels of SA | 19:41 |
uvirtbot | New bug: #754963 in nova "nova-manage doesn't handle commands that are prefixes of other commands" [Undecided,New] https://launchpad.net/bugs/754963 | 19:42 |
justinsb | blamar: Sorry - just saw you already had a patch up... reading it now | 19:42 |
blamar | justinsb: np, just about to MP if you want me to tag you | 19:43 |
*** omidhdl has joined #openstack | 19:43 | |
justinsb | blamar: I don't really get the fix though? | 19:43 |
blamar | justinsb: SHUTDOWN is the state when an instance is in the process of shutting down, SHUTOFF is when it's actually off. | 19:43 |
blamar | justinsb: so basically there is a loop which never exists becuase it doesn't catch the server in SHUTDOWN, because it quickly goes to SHUTOFF | 19:44 |
*** adiantum has joined #openstack | 19:44 | |
*** zigo-_- has quit IRC | 19:44 | |
blamar | s/exists/exits | 19:44 |
justinsb | blamar: Ah - while True there :-) | 19:44 |
blamar | yeah :( | 19:44 |
justinsb | Two suggestions: Put a message when you LOG.debug the exception | 19:45 |
*** ctennis has quit IRC | 19:45 | |
justinsb | And maybe put a comment next to SHUTDOWN vs SHUTOFF in power_states so we can document this | 19:45 |
blamar | justinsb, That log was just because I absolutely *loathe* except Exception: | 19:45 |
blamar | but yeah, I'll try to create an error message | 19:45 |
justinsb | As there's a lot of confusion about SHUTDOWN vs SHUTOFF | 19:45 |
blamar | silly naming IMO | 19:45 |
*** dendro-afk is now known as dendrobates | 19:45 | |
justinsb | blamar: Absolutely agree with you on catching exceptions, and also about catching them silently | 19:46 |
justinsb | I guess that explains why sometimes Control-C didn't seem to work the first time :-) | 19:46 |
justinsb | But sure, put me as a reviewer. I'll pull the patch into my tree and verify that it fixes the problem for me also. | 19:46 |
justinsb | Awesome time between bug and fix if so :-) | 19:47 |
blamar | k, should be up in a few | 19:47 |
*** dirkx has joined #openstack | 19:48 | |
*** adiantum has quit IRC | 19:56 | |
justinsb | blamar: Going to file a bug over that loop hammering the DB as well... | 19:56 |
justinsb | blamar: As in, "I am going to..." | 19:56 |
blamar | justinsb: Ugh, yeah it's not too pretty. Oiy that loop. | 19:57 |
blamar | justinsb: functional fix first, then better fix later though | 19:57 |
justinsb | blamar: Exactly - filing a bug so we don't lose it. | 19:58 |
blamar | :) | 19:58 |
justinsb | Blamar: Filed Bug #754983 | 19:59 |
uvirtbot | Launchpad bug 754983 in nova "Instance shutdown with libvirt hammers the DB" [Undecided,New] https://launchpad.net/bugs/754983 | 19:59 |
*** adiantum has joined #openstack | 20:01 | |
uvirtbot | New bug: #754981 in nova "nova-compute on HyperV will not start using latest trunk" [Undecided,New] https://launchpad.net/bugs/754981 | 20:02 |
*** reldan has joined #openstack | 20:03 | |
*** reldan has quit IRC | 20:05 | |
*** reldan has joined #openstack | 20:05 | |
uvirtbot | New bug: #754983 in nova "Instance shutdown with libvirt hammers the DB" [Undecided,New] https://launchpad.net/bugs/754983 | 20:06 |
*** dendrobates is now known as dendro-afk | 20:08 | |
*** mray has quit IRC | 20:10 | |
*** Harbringer has quit IRC | 20:12 | |
*** adiantum has quit IRC | 20:12 | |
dsockwell | justinsb: would it be reasonable to expand the solaris SAN driver to export volumes over fibre channel? | 20:14 |
*** bkkrw has quit IRC | 20:16 | |
justinsb | dsockwell: Don't see why not! The only issue is that volume provisioning and the transport are currently linked, so you'd have to write a different function for the mounting, and it would be inside a class that is currently called "IscsiDriver" or something like that! | 20:17 |
*** adiantum has joined #openstack | 20:17 | |
justinsb | dsockwell: But I think the thing to do is to implement something, and then we can refactor as needed | 20:17 |
*** tester23 has quit IRC | 20:22 | |
dsockwell | i took a glance at IscsiDriver, it's rather specific to iscsi. I haven't gone so far as to look at VolumeDriver, though. | 20:23 |
*** adiantum has quit IRC | 20:24 | |
*** dirkx has quit IRC | 20:25 | |
*** kakella has left #openstack | 20:26 | |
*** bkkrw has joined #openstack | 20:29 | |
devcamcar | i'd love to see this branch land: https://code.launchpad.net/~vishvananda/nova/has_role_cache/+merge/56449 | 20:29 |
devcamcar | greatly improves performance for ldap backends | 20:29 |
zykes- | http://twistedmatrix.com/documents/current/web/howto/ | 20:29 |
zykes- | http://twistedmatrix.com/documents/current/web/howto/ | 20:29 |
*** bcwaldon has quit IRC | 20:29 | |
zykes- | wrong paste :7 | 20:30 |
*** stewart has quit IRC | 20:31 | |
*** adiantum has joined #openstack | 20:37 | |
*** mray has joined #openstack | 20:39 | |
soren | eday: On it. | 20:41 |
*** mray has quit IRC | 20:49 | |
vishy | soren: are you familiar with conntrack + snatting? | 20:49 |
vishy | soren: I'm seeing something a little odd | 20:49 |
soren | vishy: pray tell | 20:50 |
vishy | soren: so I have a nova install using 3 machines, they all have ips on the 192.168 range | 20:50 |
*** hggdh has quit IRC | 20:51 | |
vishy | the vms all have the standard 10.0 addresses | 20:51 |
vishy | 192.168.2.11 is running nova-network | 20:51 |
*** hggdh has joined #openstack | 20:51 | |
vishy | the vm is on 192.168.2.13 and has the ip 10.0.0.4 | 20:51 |
vishy | so if i send ping or curl or whatever from the vm | 20:52 |
*** omidhdl has left #openstack | 20:52 | |
vishy | it goes to the network host which is 192.168.2.11/10.0.0.1 | 20:52 |
vishy | then it hits the snat rule | 20:52 |
*** ccustine has joined #openstack | 20:53 | |
vishy | and gets snatted to 192.168.2.11 and the packet makes it out through the router gateway (192.168.2.1) | 20:53 |
soren | Got it. | 20:53 |
vishy | but the response comes back to 192.168.2.11 and never makes it back to 10.0.0.4 | 20:53 |
vishy | so for some reason the conntrack is not tracking all the way back | 20:54 |
vishy | SYN_RECV src=10.0.0.4 dst=74.125.224.48 sport=41779 dport=80 packets=1 bytes=60 src=74.125.224.48 dst=192.168.2.11 sport=80 dport=41779 packets=4 bytes=240 mark=0 secmark=0 use=2 | 20:54 |
vishy | note the dst in that rule... | 20:54 |
*** coreym is now known as CoreyM-afk | 20:55 | |
vishy | this is the first time I've tried to do a multimachine install with everything using the same interface (even the public route) and compute running on the network host as well | 20:55 |
*** allsystemsarego has quit IRC | 20:57 | |
* soren ponders | 20:59 | |
openstackjenkins | Project nova build #798: SUCCESS in 2 min 29 sec: http://jenkins.openstack.org/job/nova/798/ | 20:59 |
openstackjenkins | Tarmac: removes log command from nova-manage as it no longer worked in multi-log setup | 20:59 |
soren | vishy: Sorry, where is that output from? | 21:02 |
vishy | soren: conntrack -L | grep 10.0.0.4 | 21:03 |
*** mahadev has quit IRC | 21:04 | |
vishy | soren: btw, it works fine for instances actually on the 192.168.2.11 host | 21:06 |
*** RobertLaptop has quit IRC | 21:06 | |
*** ctennis has joined #openstack | 21:07 | |
vishy | hmm | 21:07 |
vishy | soren: although it has the same odd dst in the conntrack so maybe it isn't a conntrack issue | 21:08 |
*** mray has joined #openstack | 21:08 | |
*** shentonfreude has quit IRC | 21:09 | |
*** niksnut has quit IRC | 21:09 | |
*** adiantum has quit IRC | 21:09 | |
soren | vishy: Is "74.125.224.48" the one you consider odd? | 21:09 |
soren | vishy: I was assuming not. | 21:10 |
vishy | no that is fine | 21:11 |
soren | Ok, good. Just making sure we're ont the same page. :) | 21:11 |
*** dubsquared has quit IRC | 21:11 | |
vishy | I thought that the 192.168.2.11 was odd, thought it would put 10.0.0.4 there, but as i said it seems to work fine with vms on the same host | 21:11 |
*** dubsquared has joined #openstack | 21:12 | |
jonkelly | stupid question, but do you have ip_forward turned on? | 21:12 |
*** h0cin has quit IRC | 21:12 | |
jonkelly | </lurk> | 21:12 |
jonkelly | :P | 21:12 |
soren | vishy: Did you paste that conntract entry verbatim? Which kernel version is this? | 21:13 |
vishy | yes | 21:14 |
vishy | maverick kernel... | 21:14 |
soren | vishy: Yes, what? ip_forward or verbatim? | 21:14 |
vishy | verbatim | 21:14 |
soren | Ok. | 21:14 |
soren | Assuming you do have ip_forward enabled? | 21:14 |
vishy | yup | 21:15 |
*** dendro-afk is now known as dendrobates | 21:17 | |
*** adiantum has joined #openstack | 21:23 | |
*** bkkrw has quit IRC | 21:27 | |
*** mancdaz has joined #openstack | 21:27 | |
*** troytoman-away is now known as troytoman | 21:28 | |
*** rnirmal has quit IRC | 21:29 | |
*** rnirmal has joined #openstack | 21:31 | |
soren | vishy: I think I may have slightly overestimated my ability to decipher netfilter code this time of day. | 21:33 |
vishy | hehe :) | 21:33 |
vishy | it is too bad | 21:34 |
vishy | i'd really like multimachine flatdhcp to work with just one interface | 21:34 |
*** BK_man has joined #openstack | 21:34 | |
soren | I can't find any docs on the format of that output, so it's hard to figure out what to expect from those fields. | 21:34 |
jonkelly | could be a routing issue | 21:35 |
nhm_ | vishy: that whole aspect of flatdhcp still confuses me. | 21:35 |
jonkelly | iirc packets received from an unexpected vlan get silently dropped | 21:35 |
jonkelly | so if it's coming in on an unintended interface that can cause issues | 21:36 |
jonkelly | a bit rusty on my iptables, but i recall running into that before | 21:36 |
vishy | ahaha! | 21:36 |
vishy | i found magic | 21:36 |
soren | ahaha? | 21:36 |
vishy | it is going on br100 | 21:37 |
vishy | properly | 21:37 |
soren | As expected... right? | 21:37 |
vishy | gdi! | 21:37 |
vishy | i've run into this before, i just forgot | 21:37 |
vishy | the interface has to be promisc | 21:38 |
vishy | ! | 21:38 |
soren | "gdi"? | 21:38 |
soren | Why would it have to be promiscuous? | 21:39 |
jonkelly | if it's the same thing I'm thinking of, it can be fixed via iproute as well | 21:39 |
jonkelly | setting the interface to promiscuous would be another solution | 21:39 |
soren | Oh. "gdi". i get it. :) | 21:39 |
soren | jonkelly: Which thing are you thinking of? | 21:41 |
vishy | soren: so if i tcpdump on br100 | 21:41 |
*** ppetraki has quit IRC | 21:41 | |
vishy | packets go across to 10.0.0.4 properly | 21:41 |
vishy | stop the tcpdump packets stop | 21:41 |
soren | *chuckle* | 21:41 |
soren | Heisenbug. | 21:41 |
vishy | yeah | 21:42 |
*** bkkrw has joined #openstack | 21:42 | |
vishy | i remember this happening once before and if i set the interface promisc it works | 21:42 |
soren | I'm not sure why, though? | 21:42 |
vishy | interesting | 21:43 |
vishy | tcpdump on eth0 doesn't work | 21:43 |
vishy | has to be on the bridge | 21:43 |
soren | Yeah. | 21:43 |
nhm_ | is nova-networking required at all for flatmanager? | 21:43 |
vishy | yes | 21:44 |
vishy | well | 21:44 |
*** HouseAway has quit IRC | 21:44 | |
vishy | it doesn't do a whole lot :) | 21:44 |
soren | promiscuous mode should only affect whether or not a given interface passes ethernet frames with a foreign mac furhter up the network stack. Why would the MAC be different here? | 21:44 |
nhm_ | vishy: on the compute nodes or just the cloud controller? | 21:44 |
vishy | well the mac is on another machine | 21:44 |
vishy | nhm_ network just needs to run in one place in all modes | 21:44 |
vishy | so conroller | 21:44 |
soren | vishy: Whuh? | 21:44 |
vishy | bleh having two conversations | 21:45 |
vishy | soren: 10.0.0.4 is on another machine | 21:45 |
soren | Sure. | 21:45 |
nhm_ | vishy: thanks, I think one of the docs says to install it on the compute nodes too. | 21:45 |
vishy | nhm_: yeah that is wrong | 21:45 |
nhm_ | probably explains why it was breaking my routing. | 21:46 |
*** adiantum has quit IRC | 21:46 | |
*** niksnut has joined #openstack | 21:46 | |
soren | vishy: Oh, hang on... Which machine needs its nic in promisc mode? | 21:46 |
vishy | ip link set dev br100 promisc on | 21:46 |
vishy | booom works | 21:46 |
blamar | vishy: tcpdump, by default, puts the interface it's listening on into promisc fyi if you were curious | 21:46 |
vishy | yeah that is what i remember from last time | 21:46 |
vishy | so apparently it can't forward requests back across the same bridge that the packet came in on? | 21:48 |
vishy | so packet comes in on eth0/br100 and it won't forward back to br100 without promisc | 21:48 |
vishy | i don't know why but i'm wondering if i should make a patch for it anyway | 21:49 |
soren | vishy: Is it the network node that needs to be in promisc mode? | 21:49 |
vishy | yes | 21:49 |
soren | Fascinating. | 21:49 |
soren | So the topology to reproduce it: | 21:50 |
soren | One node A: br100 connected to eth0 and a tap device. | 21:50 |
soren | Err.. | 21:51 |
soren | Man, I'm tired! | 21:51 |
soren | :) | 21:51 |
vishy | both noda a and b have that | 21:51 |
vishy | noda a has two a public and private ip on the bridge | 21:52 |
soren | Well, the tap device on the network node is not really relevant. | 21:52 |
soren | Ok. | 21:52 |
*** adiantum has joined #openstack | 21:52 | |
openstackjenkins | Project nova build #799: SUCCESS in 2 min 32 sec: http://jenkins.openstack.org/job/nova/799/ | 21:54 |
openstackjenkins | Tarmac: Added RST file on using Zones. | 21:54 |
*** jonkelly has left #openstack | 21:54 | |
soren | vishy: I'm going to have to try and reproduce this on Monday. | 21:58 |
vishy | soren: gl :) | 21:59 |
*** CoreyM-afk has quit IRC | 22:05 | |
*** Zangetsu has quit IRC | 22:05 | |
*** adiantum has quit IRC | 22:07 | |
redbo | [1]RickB17: dunno, how'd it fail? | 22:07 |
*** gondoi has quit IRC | 22:09 | |
openstackjenkins | Project nova build #800: SUCCESS in 2 min 33 sec: http://jenkins.openstack.org/job/nova/800/ | 22:09 |
openstackjenkins | Tarmac: There is a race condition when a VDI is mounted and the device node is created. Sometimes (depending on the configuration of the Linux distribution) nova loses the race and will try to open the block device before it has been created in /dev. | 22:09 |
openstackjenkins | This change will poll waiting on the device node to be created before attempting to use it. | 22:09 |
*** rnirmal has quit IRC | 22:10 | |
*** shentonfreude has joined #openstack | 22:11 | |
*** adiantum has joined #openstack | 22:13 | |
*** mftx has quit IRC | 22:13 | |
*** dspano has quit IRC | 22:17 | |
*** BK_man has quit IRC | 22:21 | |
*** mray has quit IRC | 22:23 | |
*** kronous_ has quit IRC | 22:26 | |
*** kronous_ has joined #openstack | 22:28 | |
*** dubsquared has left #openstack | 22:29 | |
*** adiantum has quit IRC | 22:29 | |
*** troytoman is now known as troytoman-away | 22:33 | |
*** bkkrw has quit IRC | 22:34 | |
*** rw4c has quit IRC | 22:34 | |
*** [1]RickB17 has quit IRC | 22:36 | |
*** Zangetsu has joined #openstack | 22:39 | |
uvirtbot | New bug: #755089 in swift "swift3 middleware doesn't support signed URLs" [Undecided,New] https://launchpad.net/bugs/755089 | 22:42 |
*** adiantum has joined #openstack | 22:43 | |
*** burris has quit IRC | 22:51 | |
*** BK_man has joined #openstack | 22:59 | |
*** hggdh has quit IRC | 23:04 | |
*** aliguori has quit IRC | 23:07 | |
*** hggdh has joined #openstack | 23:09 | |
*** mancdaz has quit IRC | 23:11 | |
*** johnpur has quit IRC | 23:14 | |
*** reldan has quit IRC | 23:16 | |
*** reldan has joined #openstack | 23:19 | |
*** aliguori has joined #openstack | 23:22 | |
*** dragondm has quit IRC | 23:23 | |
*** magglass1 has quit IRC | 23:28 | |
*** magglass1 has joined #openstack | 23:30 | |
openstackjenkins | Project burrow build #2: SUCCESS in 13 sec: http://jenkins.openstack.org/job/burrow/2/ | 23:33 |
openstackjenkins | Tarmac: Added overview and protocol docs. | 23:33 |
*** nelson has quit IRC | 23:37 | |
*** mathrock has joined #openstack | 23:43 | |
*** adiantum has quit IRC | 23:48 | |
*** adiantum has joined #openstack | 23:48 | |
*** dendrobates is now known as dendro-afk | 23:49 | |
*** kostil has joined #openstack | 23:50 | |
*** dendro-afk is now known as dendrobates | 23:52 | |
JordanRinke | elo | 23:54 |
*** kostil[1] has joined #openstack | 23:54 | |
kostil[1] | but the installation and configuration of storage nodes is below the steps for configuring proxy, which confuses me. Should one install and configure storage nodes prior to configurating proxy and auth nodes? | 23:54 |
*** kostil has quit IRC | 23:54 | |
*** reldan has quit IRC | 23:55 | |
*** BK_man has quit IRC | 23:59 | |
*** kostil[1] has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!