*** dkingshott has quit IRC | 00:04 | |
*** markvoelker has quit IRC | 00:07 | |
*** VW has joined #openstack-operators | 00:11 | |
*** dmsimard is now known as dmsimard_away | 00:11 | |
*** VW has quit IRC | 00:11 | |
*** VW has joined #openstack-operators | 00:12 | |
*** signed8bit is now known as signed8bit_ZZZzz | 00:46 | |
*** markvoelker has joined #openstack-operators | 00:52 | |
*** david-lyle is now known as david-lyle_afk | 00:57 | |
*** markvoelker has quit IRC | 01:00 | |
*** VW has quit IRC | 01:09 | |
*** VW has joined #openstack-operators | 01:09 | |
*** signed8bit_ZZZzz is now known as signed8bit | 01:13 | |
*** signed8bit is now known as signed8bit_ZZZzz | 01:13 | |
*** signed8bit_ZZZzz is now known as signed8bit | 01:13 | |
*** markvoelker has joined #openstack-operators | 01:19 | |
*** SimonChung1 has quit IRC | 01:31 | |
*** Piet has quit IRC | 01:33 | |
*** VW has quit IRC | 01:42 | |
*** VW has joined #openstack-operators | 01:43 | |
*** VW has quit IRC | 01:44 | |
*** Marga_ has quit IRC | 01:51 | |
*** Marga_ has joined #openstack-operators | 01:52 | |
*** alop has quit IRC | 01:57 | |
*** Marga_ has quit IRC | 02:07 | |
*** VW has joined #openstack-operators | 02:18 | |
*** VW has quit IRC | 02:42 | |
*** VW has joined #openstack-operators | 02:43 | |
*** signed8bit has quit IRC | 03:28 | |
*** harlowja is now known as harlowja_away | 03:28 | |
*** ToMiles has joined #openstack-operators | 03:32 | |
*** ToMiles has quit IRC | 03:37 | |
*** david-lyle_afk has quit IRC | 05:11 | |
*** david-lyle_afk has joined #openstack-operators | 05:11 | |
*** harlowja_away has quit IRC | 05:19 | |
*** ToMiles has joined #openstack-operators | 05:32 | |
*** ToMiles has quit IRC | 05:37 | |
*** Marga_ has joined #openstack-operators | 05:53 | |
*** Marga_ has quit IRC | 05:53 | |
*** Marga_ has joined #openstack-operators | 05:54 | |
*** markvoelker has quit IRC | 06:08 | |
*** blairo has joined #openstack-operators | 06:27 | |
*** blair has quit IRC | 06:29 | |
*** blairo has quit IRC | 06:31 | |
*** markvoelker has joined #openstack-operators | 06:39 | |
*** markvoelker has quit IRC | 06:43 | |
*** zz_avozza is now known as avozza | 06:52 | |
*** avozza is now known as zz_avozza | 06:56 | |
*** VW has quit IRC | 07:00 | |
*** VW has joined #openstack-operators | 07:01 | |
*** VW has quit IRC | 07:15 | |
*** VW has joined #openstack-operators | 07:16 | |
*** VW has quit IRC | 07:23 | |
*** VW has joined #openstack-operators | 07:23 | |
*** VW has quit IRC | 07:29 | |
*** VW has joined #openstack-operators | 07:29 | |
*** Miouge has joined #openstack-operators | 07:35 | |
*** belmoreira has joined #openstack-operators | 07:36 | |
*** VW has quit IRC | 07:37 | |
*** VW has joined #openstack-operators | 07:38 | |
*** markvoelker has joined #openstack-operators | 07:39 | |
*** markvoelker has quit IRC | 07:44 | |
*** VW has quit IRC | 07:46 | |
*** VW has joined #openstack-operators | 07:47 | |
*** VW has quit IRC | 07:49 | |
*** VW has joined #openstack-operators | 07:50 | |
*** VW has quit IRC | 07:52 | |
*** VW has joined #openstack-operators | 07:53 | |
*** zz_avozza is now known as avozza | 07:55 | |
*** VW has quit IRC | 07:58 | |
*** VW has joined #openstack-operators | 07:59 | |
*** VW has quit IRC | 07:59 | |
*** VW has joined #openstack-operators | 08:00 | |
*** Miouge_ has joined #openstack-operators | 08:09 | |
*** Miouge has quit IRC | 08:11 | |
*** Miouge_ is now known as Miouge | 08:11 | |
*** VW has quit IRC | 08:11 | |
*** VW has joined #openstack-operators | 08:12 | |
*** VW has quit IRC | 08:18 | |
*** VW has joined #openstack-operators | 08:19 | |
*** avozza is now known as zz_avozza | 08:26 | |
*** VW has quit IRC | 08:27 | |
*** VW has joined #openstack-operators | 08:27 | |
*** VW has quit IRC | 08:32 | |
*** racedo_ has joined #openstack-operators | 08:37 | |
*** racedo_ is now known as racedo | 08:38 | |
*** markvoelker has joined #openstack-operators | 08:40 | |
*** markvoelker has quit IRC | 08:45 | |
*** matrohon has joined #openstack-operators | 08:54 | |
*** bvandenh has joined #openstack-operators | 09:00 | |
*** belmoreira has quit IRC | 09:21 | |
*** belmoreira has joined #openstack-operators | 09:22 | |
*** Miouge has quit IRC | 09:25 | |
*** derekh has joined #openstack-operators | 09:27 | |
*** Miouge has joined #openstack-operators | 09:32 | |
*** markvoelker has joined #openstack-operators | 09:41 | |
*** markvoelker has quit IRC | 09:47 | |
*** pasquier-s has quit IRC | 10:21 | |
*** pasquier-s has joined #openstack-operators | 10:36 | |
*** markvoelker has joined #openstack-operators | 10:43 | |
*** markvoelker has quit IRC | 10:48 | |
*** belmoreira has quit IRC | 10:48 | |
*** VW has joined #openstack-operators | 10:58 | |
*** VW has quit IRC | 11:03 | |
*** markvoelker has joined #openstack-operators | 11:44 | |
*** markvoelker has quit IRC | 11:49 | |
*** pasquier-s has quit IRC | 11:58 | |
*** david-lyle_afk is now known as david-lyle | 12:01 | |
*** Gala-G has joined #openstack-operators | 12:08 | |
*** pasquier-s has joined #openstack-operators | 12:13 | |
*** mux_ has joined #openstack-operators | 12:15 | |
*** subscope has quit IRC | 12:19 | |
*** subscope has joined #openstack-operators | 12:20 | |
*** fsaad has quit IRC | 12:20 | |
*** mux has quit IRC | 12:20 | |
*** sbadia has quit IRC | 12:20 | |
*** sbadia has joined #openstack-operators | 12:22 | |
*** fsaad has joined #openstack-operators | 12:28 | |
*** david-lyle is now known as david-lyle_afk | 12:36 | |
*** david-lyle_afk is now known as david-lyle | 12:36 | |
*** Gala-G12 has joined #openstack-operators | 12:39 | |
*** Gala-G has quit IRC | 12:42 | |
*** zz_avozza is now known as avozza | 12:43 | |
*** david-lyle is now known as david-lyle_afk | 12:44 | |
*** david-lyle_afk is now known as david-lyle | 12:45 | |
*** racedo has quit IRC | 12:45 | |
*** markvoelker has joined #openstack-operators | 12:45 | |
*** markvoelker has quit IRC | 12:49 | |
*** pboros has joined #openstack-operators | 13:06 | |
*** racedo has joined #openstack-operators | 13:07 | |
*** racedo has quit IRC | 13:07 | |
*** racedo has joined #openstack-operators | 13:07 | |
*** Miouge has quit IRC | 13:09 | |
*** dmsimard_away is now known as dmsimard | 13:09 | |
*** dmsimard is now known as dmsimard_away | 13:13 | |
*** markvoelker has joined #openstack-operators | 13:20 | |
*** belmoreira has joined #openstack-operators | 14:11 | |
*** reed has joined #openstack-operators | 14:14 | |
*** Marga_ has quit IRC | 14:17 | |
*** dmsimard_away is now known as dmsimard | 14:18 | |
*** VW_ has joined #openstack-operators | 14:24 | |
*** zerda has joined #openstack-operators | 14:42 | |
zerda | Hello. How do I disallow ephemeral storage completely in Icehouse? I mean, is there a way easier than patching horizon/nova? | 14:45 |
---|---|---|
*** Ctina has joined #openstack-operators | 14:47 | |
j05h | zerda: no hard drives | 14:49 |
j05h | sorry…woke on the wrong side of the snark this morning. | 14:50 |
mfisch | The openstack naming choices are all pretty cool | 14:53 |
*** derekh has quit IRC | 14:53 | |
mfisch | I just wish Lizard was "Lizard King" | 14:53 |
zerda | j05h, that's ok. Actually, I'd like not to distract end users with possibility to use that, not just forbid ephemeral storage completely | 14:54 |
*** signed8bit has joined #openstack-operators | 14:54 | |
j05h | zerda: why no ephemeral? | 14:59 |
j05h | (i like ephemeral :D) | 14:59 |
*** derekh has joined #openstack-operators | 15:01 | |
*** signed8bit is now known as signed8bit_ZZZzz | 15:03 | |
zerda | j05h, well, we have Cinder based on some SAN, and compute hosts really have no disks (only USB drives as root devices), and I want to avoid situation when someone creates a VM on that USB drive ^) | 15:03 |
j05h | probably a good thing to do. | 15:05 |
zerda | yep. I looked at Horizon code briefly, and seems that there is no way to remove "ephemeral" options from the "create instance" screen | 15:08 |
j05h | provide only boot from volume flavors i believe. http://docs.openstack.org/admin-guide-cloud/content/section_storage-and-openstack-compute.html | 15:10 |
*** Gala-G12 has quit IRC | 15:13 | |
*** signed8bit_ZZZzz is now known as signed8bit | 15:17 | |
zerda | j05h, hmm, sounds like a solution. Will try this tomorrow, thank you | 15:18 |
j05h | zerda: sure, no problem. hope it works for you. | 15:23 |
*** avozza is now known as zz_avozza | 15:23 | |
*** VW_ has quit IRC | 15:33 | |
*** Ctina has quit IRC | 15:38 | |
*** Ctina has joined #openstack-operators | 15:39 | |
*** matrohon has quit IRC | 15:50 | |
*** rzerda has joined #openstack-operators | 15:56 | |
*** zerda has quit IRC | 15:58 | |
*** VW_ has joined #openstack-operators | 16:00 | |
Ctina | With our deployment, we've added two panels to the dashboard. They show up and display just fine but the links on the navigation panel aren't consistent, one session it'll have panel A first and the next session it'll have panel B first. The docs hint that there's a way to set ordering, has anyone set ordering for their customized panels? | 16:02 |
*** Marga_ has joined #openstack-operators | 16:03 | |
*** alop has joined #openstack-operators | 16:24 | |
*** Gala-G has joined #openstack-operators | 16:30 | |
Ctina | are there people here using PKI tokens and heat? | 16:31 |
*** Marga_ has quit IRC | 16:39 | |
*** mdorman has joined #openstack-operators | 16:43 | |
*** belmoreira has quit IRC | 16:52 | |
rzerda | j05h, suddenly there is a tiny option named max_local_block_devices in nova.conf, which, if set to zero, forbids any local=ephemeral storage for the node. | 16:56 |
*** mdorman has quit IRC | 17:11 | |
j05h | rzerda: nice! | 17:14 |
*** mdorman has joined #openstack-operators | 17:15 | |
*** signed8bit is now known as signed8bit_ZZZzz | 17:20 | |
*** matrohon has joined #openstack-operators | 17:20 | |
*** Marga_ has joined #openstack-operators | 17:21 | |
*** bvandenh has quit IRC | 17:30 | |
*** Piet has joined #openstack-operators | 17:30 | |
*** rzerda has quit IRC | 17:35 | |
*** SimonChung has joined #openstack-operators | 17:43 | |
*** Gala-G has quit IRC | 17:51 | |
*** VW_ has quit IRC | 17:59 | |
*** Gala-G has joined #openstack-operators | 18:00 | |
*** derekh has quit IRC | 18:02 | |
*** SimonChung1 has joined #openstack-operators | 18:02 | |
*** SimonChung has quit IRC | 18:03 | |
*** SimonChung1 has quit IRC | 18:03 | |
*** SimonChung2 has joined #openstack-operators | 18:03 | |
*** SimonChung has joined #openstack-operators | 18:04 | |
*** SimonChung2 has quit IRC | 18:04 | |
*** Gala-G has quit IRC | 18:09 | |
*** harlowja has joined #openstack-operators | 18:11 | |
*** signed8bit_ZZZzz is now known as signed8bit | 18:17 | |
*** SimonChung1 has joined #openstack-operators | 18:29 | |
*** SimonChung has quit IRC | 18:29 | |
*** SimonChung has joined #openstack-operators | 18:29 | |
*** SimonChung2 has joined #openstack-operators | 18:30 | |
*** Marga_ has quit IRC | 18:30 | |
*** SimonChung3 has joined #openstack-operators | 18:30 | |
*** SimonChung2 has quit IRC | 18:31 | |
*** Marga_ has joined #openstack-operators | 18:31 | |
*** SimonChung1 has quit IRC | 18:31 | |
*** SimonChung has quit IRC | 18:31 | |
*** signed8bit is now known as signed8bit_ZZZzz | 18:33 | |
*** signed8bit_ZZZzz is now known as signed8bit | 18:35 | |
xavpaice | zerda: without disk on the compute nodes, what happens when people suspend a VM? Do you have some NFS type volume for that? | 18:39 |
dvorak | it'd all be on the cinder volume, no? or you mean the memory for the instance | 18:47 |
jlk | suspend goes to ram no? | 18:48 |
jlk | I guess there are two ways to suspend kernels normally | 18:48 |
*** SimonChung3 has quit IRC | 18:48 | |
dvorak | I'd expect if I suspend and instance that the memory would go to persistent storage | 18:48 |
dvorak | but I have no idea if it actually does :) | 18:49 |
jlk | yeah I was thinking of laptop suspends | 18:49 |
dvorak | I'd guess it goes to /var/lib/nova/instances | 18:49 |
jlk | which go to memory usually instead of disks | 18:49 |
dvorak | which may or may not go to shared storage | 18:49 |
dvorak | this says it goes to disk - http://docs.openstack.org/user-guide/content/suspend_resume.html | 18:50 |
jlk | wonder then if it would go to the boot volume | 18:50 |
dvorak | uh, but it also says it doesn't free any memory, which seems crazy | 18:50 |
jlk | well | 18:50 |
dvorak | maybe it doesn't free it from a scheduler standpoint | 18:50 |
jlk | it may not free | 18:50 |
jlk | right | 18:50 |
jlk | it keeps it in reserve so that the instance could be resumed | 18:50 |
dvorak | that just seems weird to me. seems like finding a place with enough resources to resume it would be an issue you'd worry about when you wanted to resume it | 18:51 |
jlk | yeah, but you can't easily move the local disk | 18:52 |
jlk | if you suspend an instance on Hypervisor A, you can't resume it on Hypervisor B | 18:52 |
jlk | if local disk is involved | 18:52 |
dvorak | unless you have shared storage | 18:52 |
dvorak | nod | 18:52 |
dvorak | I've only ever really done suspend/resume on vmware, where shared storage is assumed | 18:53 |
xavpaice | we've not tried with shared storage, but yeah, was referring to suspend putting the memory on disk | 18:53 |
dvorak | actually, that's not true, share storage isn't assumed | 18:53 |
dvorak | you just can't resume it unless the storage you suspended it to is available | 18:53 |
* xavpaice wonders if there's an option to disable suspend | 18:53 | |
dvorak | seems like you couild with policy.json | 18:53 |
dvorak | not sure if it'll disappear from the horizon ui though :) | 18:54 |
xavpaice | well, that would at least stop people from being able to fill up the disk on a hypervisor | 18:54 |
dvorak | node | 18:54 |
dvorak | err, nod :) | 18:54 |
*** Gala-G has joined #openstack-operators | 18:54 | |
dvorak | but since you have everything on cinder volumes, you can do live-migration, right? | 18:55 |
xavpaice | works for us, cinder volumes on ceph storage | 18:55 |
dvorak | ah, ok. do you do ephemeral on ceph also? | 18:56 |
xavpaice | we don't have ephemeral really, unless you count what happens when they boot from an image | 18:57 |
dvorak | yeah, that's what I was thinking | 18:57 |
xavpaice | ceph is our only storage - even for object store at this stage | 18:57 |
dvorak | we can do boot from image, or boot from volume with image | 18:57 |
*** Marga_ has quit IRC | 18:58 | |
*** Marga_ has joined #openstack-operators | 18:58 | |
xavpaice | same. We've got opened up the ability for swap etc | 18:58 |
*** mux_ is now known as mux | 18:58 | |
xavpaice | s/got/not/ | 18:59 |
dvorak | you have people that want swap? | 19:00 |
xavpaice | not just yet :) | 19:00 |
*** j05h1 has joined #openstack-operators | 19:01 | |
*** j05h has quit IRC | 19:01 | |
*** j05h2 has joined #openstack-operators | 19:01 | |
*** j05h1 has quit IRC | 19:05 | |
dvorak | weird, trying to get designate to connect to rabbitmq, and rabbit keeps insisting that the user/pass is wrong | 19:06 |
*** signed8bit is now known as signed8bit_ZZZzz | 19:08 | |
dvorak | ah, that's because it's wrong :) | 19:09 |
*** VW_ has joined #openstack-operators | 19:10 | |
*** signed8bit_ZZZzz is now known as signed8bit | 19:11 | |
*** SimonChung has joined #openstack-operators | 19:13 | |
*** SimonChung has quit IRC | 19:22 | |
*** Gala-G has quit IRC | 19:28 | |
*** SimonChung has joined #openstack-operators | 19:38 | |
*** Marga_ has quit IRC | 19:38 | |
*** Marga_ has joined #openstack-operators | 19:39 | |
*** harlowja is now known as harlowja_away | 19:50 | |
*** signed8bit is now known as signed8bit_ZZZzz | 19:54 | |
*** signed8bit_ZZZzz is now known as signed8bit | 19:56 | |
*** SimonChung has quit IRC | 19:58 | |
*** VW_ has quit IRC | 19:59 | |
*** VW_ has joined #openstack-operators | 20:02 | |
*** SimonChung has joined #openstack-operators | 20:17 | |
*** SimonChung1 has joined #openstack-operators | 20:17 | |
*** SimonChung has quit IRC | 20:17 | |
*** harlowja_away is now known as harlowja | 20:33 | |
*** matrohon has quit IRC | 20:48 | |
*** blair has joined #openstack-operators | 20:58 | |
*** pboros has quit IRC | 21:03 | |
*** Ctina has quit IRC | 21:03 | |
*** SimonChung has joined #openstack-operators | 21:07 | |
*** SimonChung1 has quit IRC | 21:07 | |
*** SimonChung has quit IRC | 21:11 | |
*** SimonChung has joined #openstack-operators | 21:19 | |
*** Piet has quit IRC | 21:33 | |
*** Marga_ has quit IRC | 21:33 | |
*** Marga_ has joined #openstack-operators | 21:34 | |
*** blairo has joined #openstack-operators | 21:38 | |
*** blair has quit IRC | 21:41 | |
*** belmoreira has joined #openstack-operators | 21:48 | |
dvorak | anyone using giftwrap? | 21:59 |
jlk | dvorak: we are (bluebox), in some preproduction stuff. | 22:02 |
dvorak | you guys wrote it, no? ;) | 22:03 |
jlk | yeah, we just don't have it in productino yet | 22:03 |
dvorak | you're not filling me with confidence :) | 22:03 |
jlk | we are doing havana -> juno first, then switching out how we lay stuff down | 22:03 |
dvorak | I really hate debian packaging, so I'm looking into it. I ran into one or two weird problems, but I manged to get it to make a deb | 22:03 |
jlk | gotcha. | 22:04 |
dvorak | how are you guys handling cli commands and such, since you have to activate the environment? | 22:04 |
jlk | Personally, I really don't like operating system packaging for something like this | 22:04 |
jlk | we're using the alternatives system | 22:04 |
jlk | to install the runtime executable stuff into a $PATH | 22:04 |
dvorak | well, I've done a ton of rpm packaging in the past, and I just assumed that debian packaging would be roughly comparable. it's dramatically more complicate for reaosns I can't relaly understand | 22:04 |
jlk | that's the part that isn't merged yet | 22:04 |
jlk | oh yes | 22:05 |
xavpaice | dvorak: would appreciate feedback on https://github.com/xavpaice/openstack-packages re debian packaging | 22:05 |
jlk | I used to work for Red Hat, I know all about rpms. Debian/Ubuntu is just weird to me, but it's what I have to work with | 22:05 |
jlk | for stuff like openstack, where we want to lay down multiple versions and whatnot, I'd rather just use tarr'd up venvs | 22:05 |
jlk | and skip the whole package repo stuff. venvs on a http source are very easy to manage | 22:06 |
dvorak | jlk: maybe I'm missing something, don't I have to source the activate script before I can run something the package installs? | 22:06 |
dvorak | or is there some magic I'm unaware of | 22:06 |
jlk | not the way alternatives works | 22:06 |
jlk | it makes a symlink | 22:06 |
dvorak | xavpaice: reading :) | 22:07 |
jlk | so the app is launched from within the venv | 22:07 |
jlk | so the venv gets loaded | 22:07 |
dvorak | interesting. that doesn't sound too bad | 22:07 |
dvorak | are you bothering to purge old packages? | 22:07 |
jlk | not as of yet , but we will at some point | 22:07 |
jlk | they don't take up a LOT of space, but they take up some. | 22:07 |
jlk | We did something similar at Rackspace, but there it was just venvs in a tarball. We'd purge old ones as a background job | 22:08 |
dvorak | yeah, we're doing somethign like that for horizon and monasca | 22:08 |
dvorak | but I think the giftwrap approach would probably be faster when building dev environments and probably more reproducible | 22:09 |
jlk | not sure how | 22:09 |
jlk | we had venv's in a tarball, so it was just wget and extract | 22:09 |
jlk | and off you go | 22:09 |
dvorak | nono, I mean compared to the way we do it for horizon and monasca | 22:09 |
jlk | should be a bit quicker than fetching apt metadata, fetching apt package, resolving deps, installing | 22:09 |
jlk | ah | 22:09 |
dvorak | we have ansible or puppet building the venv and installing all the stuff in it | 22:09 |
dvorak | so why are you guys using packages instead of tarballs then? | 22:10 |
jlk | wasn't my decision | 22:11 |
dvorak | fair enough | 22:11 |
jlk | if it were, we'd be using tarballs of venvs on CentOS7 | 22:11 |
dvorak | part of the reason packages are attractive to me is that I'm hoping I'll be able to continue to use the stackforge puppet modules without much change | 22:11 |
jlk | ah. Yeah, I'm no fan of puppet. We're all Ansible here. We publish those too | 22:12 |
dvorak | I don't think we'll be switching even if I agreed :) | 22:12 |
jlk | puppet is terrible come upgrade time when you need to handle stops, migrations, and starts in very precise order | 22:12 |
dvorak | we use ansible for that :) | 22:12 |
dvorak | w/puppet | 22:13 |
jlk | haha, RAX did that too, until I got them off of that | 22:13 |
jlk | managing the puppet manifests was too high of a barrier | 22:13 |
dvorak | the internet does most of that for me :) | 22:13 |
jlk | the ansible playbooks were much easier to work with and got a lot more people participating in them | 22:13 |
*** Ctina has joined #openstack-operators | 22:14 | |
*** Ctina has quit IRC | 22:19 | |
*** SimonChung1 has joined #openstack-operators | 22:43 | |
*** SimonChung has quit IRC | 22:43 | |
*** blairo has quit IRC | 22:51 | |
*** jaypipes has quit IRC | 23:00 | |
*** j05h2 has quit IRC | 23:02 | |
*** blair has joined #openstack-operators | 23:04 | |
*** j05h has joined #openstack-operators | 23:06 | |
*** signed8bit has quit IRC | 23:07 | |
*** SimonChung has joined #openstack-operators | 23:08 | |
*** SimonChung1 has quit IRC | 23:08 | |
*** blair has quit IRC | 23:09 | |
*** SimonChung has quit IRC | 23:09 | |
*** SimonChung1 has joined #openstack-operators | 23:09 | |
*** belmoreira has quit IRC | 23:20 | |
*** reed has quit IRC | 23:24 | |
*** markvoelker has quit IRC | 23:56 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!