| *** funzo_ has joined #tripleo | 01:38 | |
| *** funzo has quit IRC | 01:39 | |
| *** yamahata has joined #tripleo | 01:47 | |
| *** sdake_ has joined #tripleo | 01:57 | |
| *** pcrews has joined #tripleo | 02:07 | |
| *** sdake_ has quit IRC | 02:14 | |
| *** pcrews has quit IRC | 02:25 | |
| *** julim has quit IRC | 02:46 | |
| *** pradeep has joined #tripleo | 02:47 | |
| *** pradeep1 has joined #tripleo | 02:59 | |
| *** pradeep has quit IRC | 03:01 | |
| *** pradeep has joined #tripleo | 03:30 | |
| *** pradeep1 has quit IRC | 03:31 | |
| *** sdake_ has joined #tripleo | 03:58 | |
| *** untriaged-bot has joined #tripleo | 04:00 | |
| untriaged-bot | Untriaged bugs so far: | 04:00 |
|---|---|---|
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391218 | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387061 | 04:00 |
| uvirtbot | Launchpad bug 1391218 in tripleo "Configuration Dependent on HAProxy" [Undecided,In progress] | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393643 | 04:00 |
| uvirtbot | Launchpad bug 1387061 in tripleo "deploy under cloud failure due to pacemaker " [Undecided,New] | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391412 | 04:00 |
| uvirtbot | Launchpad bug 1393643 in tripleo "haproxy start failed loop with 'Starting proxy horizon: cannot bind socket'" [Undecided,New] | 04:00 |
| uvirtbot | Launchpad bug 1391412 in tripleo "devtest deploy overcloud successfully,but there have one baremetal in Maintenance state" [Undecided,New] | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387487 | 04:00 |
| uvirtbot | Launchpad bug 1387487 in tripleo "tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" [Undecided,New] | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393304 | 04:00 |
| uvirtbot | Launchpad bug 1393304 in tripleo "create-nodes failure for customised pool" [Undecided,New] | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1391656 | 04:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393298 | 04:00 |
| uvirtbot | Launchpad bug 1391656 in diskimage-builder "There is no way to customize the installation of the bootloader" [Undecided,New] | 04:00 |
| *** untriaged-bot has quit IRC | 04:00 | |
| uvirtbot | Launchpad bug 1393298 in tripleo "keepalived VRRP_Script(chk_haproxy) timed out" [Undecided,In progress] | 04:00 |
| *** sdake_ has quit IRC | 04:04 | |
| *** pradeep has quit IRC | 04:10 | |
| *** pensu has joined #tripleo | 04:14 | |
| *** vinsh has joined #tripleo | 04:33 | |
| *** vinsh has quit IRC | 04:35 | |
| *** pensu has quit IRC | 04:41 | |
| *** edmund has quit IRC | 04:58 | |
| *** pradeep has joined #tripleo | 05:10 | |
| *** rushiagr_away is now known as rushiagr | 05:46 | |
| *** sirushti has left #tripleo | 06:09 | |
| *** rushiagr is now known as rushiagr_away | 06:30 | |
| *** pradeep1 has joined #tripleo | 06:36 | |
| *** avozza is now known as zz_avozza | 06:36 | |
| *** pradeep has quit IRC | 06:37 | |
| tchaypo | In the unlikely event that anyone is looking for me - I’ll be away all day today | 06:41 |
| tchaypo | but i should be pingable here | 06:42 |
| *** zz_avozza is now known as avozza | 07:17 | |
| *** pradeep1 has quit IRC | 07:28 | |
| *** pradeep has joined #tripleo | 07:28 | |
| *** yamahata has quit IRC | 07:33 | |
| *** pradeep1 has joined #tripleo | 07:39 | |
| *** pradeep has quit IRC | 07:39 | |
| *** wuhg has joined #tripleo | 07:45 | |
| *** panda has quit IRC | 07:45 | |
| *** panda has joined #tripleo | 07:46 | |
| *** pradeep has joined #tripleo | 07:48 | |
| *** pradeep1 has quit IRC | 07:51 | |
| *** pradeep1 has joined #tripleo | 07:57 | |
| *** subscope_ has joined #tripleo | 07:58 | |
| *** pradeep has quit IRC | 08:00 | |
| *** avozza is now known as zz_avozza | 08:20 | |
| *** dshulyak_ has joined #tripleo | 08:30 | |
| *** dshulyak_ has quit IRC | 08:32 | |
| *** kpavel has quit IRC | 08:41 | |
| *** dshulyak_ has joined #tripleo | 08:42 | |
| *** kpavel has joined #tripleo | 08:47 | |
| *** zz_avozza is now known as avozza | 09:01 | |
| *** subscope_ has quit IRC | 09:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/tuskar: Updated from global requirements https://review.openstack.org/134385 | 09:41 |
| *** untriaged-bot has joined #tripleo | 10:00 | |
| untriaged-bot | Untriaged bugs so far: | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391218 | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387061 | 10:00 |
| uvirtbot | Launchpad bug 1391218 in tripleo "Configuration Dependent on HAProxy" [Undecided,In progress] | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393643 | 10:00 |
| uvirtbot | Launchpad bug 1387061 in tripleo "deploy under cloud failure due to pacemaker " [Undecided,New] | 10:00 |
| uvirtbot | Launchpad bug 1393643 in tripleo "haproxy start failed loop with 'Starting proxy horizon: cannot bind socket'" [Undecided,New] | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391412 | 10:00 |
| uvirtbot | Launchpad bug 1391412 in tripleo "devtest deploy overcloud successfully,but there have one baremetal in Maintenance state" [Undecided,New] | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387487 | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393304 | 10:00 |
| uvirtbot | Launchpad bug 1387487 in tripleo "tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" [Undecided,New] | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1391656 | 10:00 |
| uvirtbot | Launchpad bug 1393304 in tripleo "create-nodes failure for customised pool" [Undecided,New] | 10:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393298 | 10:00 |
| uvirtbot | Launchpad bug 1391656 in diskimage-builder "There is no way to customize the installation of the bootloader" [Undecided,New] | 10:00 |
| uvirtbot | Launchpad bug 1393298 in tripleo "keepalived VRRP_Script(chk_haproxy) timed out" [Undecided,In progress] | 10:00 |
| *** untriaged-bot has quit IRC | 10:00 | |
| *** lifeless has quit IRC | 10:07 | |
| *** lifeless has joined #tripleo | 10:19 | |
| *** avozza is now known as zz_avozza | 10:21 | |
| *** dshulyak_ has quit IRC | 10:24 | |
| *** dshulyak_ has joined #tripleo | 10:45 | |
| *** zz_avozza is now known as avozza | 10:47 | |
| *** rlandy has joined #tripleo | 10:58 | |
| *** zaro has quit IRC | 11:24 | |
| *** lifeless has quit IRC | 12:14 | |
| *** lifeless has joined #tripleo | 12:29 | |
| *** avozza is now known as zz_avozza | 12:36 | |
| *** pradeep has joined #tripleo | 12:39 | |
| *** pradeep1 has joined #tripleo | 13:03 | |
| *** yamahata has joined #tripleo | 13:04 | |
| *** pradeep has quit IRC | 13:06 | |
| *** bvandenh has joined #tripleo | 13:07 | |
| *** pradeep has joined #tripleo | 13:14 | |
| *** dshulyak_ has quit IRC | 13:15 | |
| *** pradeep1 has quit IRC | 13:15 | |
| *** pradeep1 has joined #tripleo | 13:18 | |
| *** pradeep has quit IRC | 13:19 | |
| *** pradeep has joined #tripleo | 13:21 | |
| *** pradeep1 has quit IRC | 13:22 | |
| *** dshulyak_ has joined #tripleo | 13:22 | |
| *** zz_avozza is now known as avozza | 13:28 | |
| *** pradeep1 has joined #tripleo | 13:28 | |
| *** pradeep has quit IRC | 13:29 | |
| *** pradeep has joined #tripleo | 13:35 | |
| *** pradeep1 has quit IRC | 13:36 | |
| *** pradeep1 has joined #tripleo | 13:38 | |
| *** pradeep has quit IRC | 13:40 | |
| *** pradeep has joined #tripleo | 13:40 | |
| *** pradeep1 has quit IRC | 13:42 | |
| *** pradeep2 has joined #tripleo | 13:44 | |
| *** pradeep has quit IRC | 13:45 | |
| *** pradeep has joined #tripleo | 13:48 | |
| *** pradeep2 has quit IRC | 13:49 | |
| *** bvandenh has quit IRC | 13:49 | |
| *** pradeep1 has joined #tripleo | 13:50 | |
| *** avozza is now known as zz_avozza | 13:51 | |
| *** pradeep has quit IRC | 13:53 | |
| *** dshulyak_ has quit IRC | 13:59 | |
| *** dshulyak_ has joined #tripleo | 14:09 | |
| *** pradeep has joined #tripleo | 14:24 | |
| *** pradeep1 has quit IRC | 14:26 | |
| *** dshulyak_ has quit IRC | 14:35 | |
| *** pradeep1 has joined #tripleo | 14:41 | |
| *** pradeep has quit IRC | 14:43 | |
| *** pradeep has joined #tripleo | 14:54 | |
| *** pradeep1 has quit IRC | 14:56 | |
| SpamapS | greghaynes: I will send a final call for "speak now or forever hold your peace" on Monday | 15:09 |
| *** pensu has joined #tripleo | 15:18 | |
| *** zz_avozza is now known as avozza | 15:21 | |
| *** yamahata has quit IRC | 15:33 | |
| *** avozza is now known as zz_avozza | 15:46 | |
| *** kpavel has quit IRC | 15:49 | |
| *** bvandenh has joined #tripleo | 15:56 | |
| *** untriaged-bot has joined #tripleo | 16:00 | |
| untriaged-bot | Untriaged bugs so far: | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391218 | 16:00 |
| uvirtbot | Launchpad bug 1391218 in tripleo "Configuration Dependent on HAProxy" [Undecided,In progress] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387061 | 16:00 |
| uvirtbot | Launchpad bug 1387061 in tripleo "deploy under cloud failure due to pacemaker " [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393643 | 16:00 |
| uvirtbot | Launchpad bug 1393643 in tripleo "haproxy start failed loop with 'Starting proxy horizon: cannot bind socket'" [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391412 | 16:00 |
| uvirtbot | Launchpad bug 1391412 in tripleo "devtest deploy overcloud successfully,but there have one baremetal in Maintenance state" [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387487 | 16:00 |
| uvirtbot | Launchpad bug 1387487 in tripleo "tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393304 | 16:00 |
| uvirtbot | Launchpad bug 1393304 in tripleo "create-nodes failure for customised pool" [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1391656 | 16:00 |
| uvirtbot | Launchpad bug 1391656 in diskimage-builder "There is no way to customize the installation of the bootloader" [Undecided,New] | 16:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393298 | 16:00 |
| uvirtbot | Launchpad bug 1393298 in tripleo "keepalived VRRP_Script(chk_haproxy) timed out" [Undecided,In progress] | 16:00 |
| *** untriaged-bot has quit IRC | 16:00 | |
| *** pradeep1 has joined #tripleo | 16:09 | |
| *** pradeep has quit IRC | 16:12 | |
| *** pensu has quit IRC | 16:15 | |
| *** vinsh has joined #tripleo | 16:15 | |
| *** vinsh has quit IRC | 16:15 | |
| *** zz_avozza is now known as avozza | 16:37 | |
| *** russell_h has quit IRC | 16:51 | |
| *** russell_h has joined #tripleo | 16:51 | |
| *** sdake_ has joined #tripleo | 16:56 | |
| *** JameLk has joined #tripleo | 17:02 | |
| *** avozza is now known as zz_avozza | 17:04 | |
| *** linkai has quit IRC | 17:05 | |
| *** wuhg has quit IRC | 17:13 | |
| *** pradeep has joined #tripleo | 17:22 | |
| *** pradeep1 has quit IRC | 17:24 | |
| mordred | o hai | 17:29 |
| mordred | can I suggest respectfully that ./elements/rpm-distro/pre-install.d/00-allow-heat-admin-sudo does not belong in rpm-distro ? | 17:30 |
| mordred | as it is very much a thing you care about if you are doing tripleo, but not at all if you are using dib to build general images | 17:30 |
| *** zz_avozza is now known as avozza | 17:40 | |
| *** dshulyak_ has joined #tripleo | 17:51 | |
| *** bvandenh has quit IRC | 17:58 | |
| mordred | zomg. | 18:04 |
| mordred | how do I use redhat-common without installing redhat-lsb-core? | 18:04 |
| mordred | because O M G I do not want to install it | 18:04 |
| mordred | really? cups? ghostscript? mesa? | 18:06 |
| mordred | (I mean, I think my question is - is there anyway, short of me just forking that element, for me to declare ina different element that I do not want to add that package?) | 18:08 |
| *** michchap_ has joined #tripleo | 18:19 | |
| greghaynes | mordred: +1 - we could probably just move that into tripleo-image-elements entirely | 18:19 |
| openstackgerrit | Monty Taylor proposed openstack/diskimage-builder: Remove heat-admin code from rpm-distro element https://review.openstack.org/136640 | 18:20 |
| openstackgerrit | Monty Taylor proposed openstack/diskimage-builder: Remove use of sudo from yum pre and post elements https://review.openstack.org/136641 | 18:20 |
| openstackgerrit | Monty Taylor proposed openstack/diskimage-builder: Add minimal ubuntu and centos base elements https://review.openstack.org/136642 | 18:20 |
| mordred | greghaynes: great. glad you think so ... | 18:20 |
| greghaynes | looks like the heat element is really what needs it, so maybe in there | 18:20 |
| *** michchap has quit IRC | 18:20 | |
| mordred | greghaynes: the minimal ubuntu/centos elements there may show what I'm trying to accomplish, although so far they're not finished | 18:20 |
| mordred | I also will not be offended if tripleo does not want those in dib directly | 18:21 |
| greghaynes | ooo | 18:23 |
| greghaynes | dib seems like a fine place for that | 18:23 |
| * greghaynes wonders why we were using sudo in the first place | 18:23 | |
| greghaynes | the commit looks like just an oversight | 18:24 |
| openstackgerrit | Monty Taylor proposed openstack/tripleo-image-elements: Add head sudo enablement from dib https://review.openstack.org/136643 | 18:30 |
| mordred | greghaynes: ^^ there's the corresponding commit | 18:30 |
| *** pcrews has joined #tripleo | 18:51 | |
| *** jang1 has joined #tripleo | 18:53 | |
| *** dshulyak_ has quit IRC | 18:54 | |
| *** pcrews has quit IRC | 19:02 | |
| *** pradeep has quit IRC | 19:11 | |
| *** david-lyle_afk has quit IRC | 19:11 | |
| *** avozza is now known as zz_avozza | 19:24 | |
| *** zz_avozza is now known as avozza | 19:27 | |
| *** avozza is now known as zz_avozza | 19:38 | |
| *** sdake_ has quit IRC | 20:04 | |
| *** kpavel has joined #tripleo | 20:07 | |
| *** penick has joined #tripleo | 20:10 | |
| *** Guest83569 is now known as gael | 20:16 | |
| *** gael is now known as Guest22010 | 20:17 | |
| *** Guest22010 is now known as gchamoul | 20:17 | |
| *** penick has quit IRC | 20:24 | |
| *** zz_avozza is now known as avozza | 20:41 | |
| *** bvandenh has joined #tripleo | 20:49 | |
| *** zigo has quit IRC | 20:59 | |
| *** zigo has joined #tripleo | 21:00 | |
| *** pcrews has joined #tripleo | 21:04 | |
| *** marun has joined #tripleo | 21:08 | |
| *** kpavel has quit IRC | 21:09 | |
| *** bvandenh has quit IRC | 21:09 | |
| *** pcrews has quit IRC | 21:10 | |
| *** goneri has joined #tripleo | 21:26 | |
| *** bvandenh has joined #tripleo | 21:38 | |
| *** bvandenh has quit IRC | 21:47 | |
| *** rlandy has quit IRC | 21:54 | |
| *** untriaged-bot has joined #tripleo | 22:00 | |
| untriaged-bot | Untriaged bugs so far: | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391218 | 22:00 |
| uvirtbot | Launchpad bug 1391218 in tripleo "Configuration Dependent on HAProxy" [Undecided,In progress] | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387061 | 22:00 |
| uvirtbot | Launchpad bug 1387061 in tripleo "deploy under cloud failure due to pacemaker " [Undecided,New] | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393643 | 22:00 |
| uvirtbot | Launchpad bug 1393643 in tripleo "haproxy start failed loop with 'Starting proxy horizon: cannot bind socket'" [Undecided,New] | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1391412 | 22:00 |
| uvirtbot | Launchpad bug 1391412 in tripleo "devtest deploy overcloud successfully,but there have one baremetal in Maintenance state" [Undecided,New] | 22:00 |
| *** goneri has quit IRC | 22:00 | |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1387487 | 22:00 |
| uvirtbot | Launchpad bug 1387487 in tripleo "tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" [Undecided,New] | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393304 | 22:00 |
| uvirtbot | Launchpad bug 1393304 in tripleo "create-nodes failure for customised pool" [Undecided,New] | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/diskimage-builder/+bug/1391656 | 22:00 |
| untriaged-bot | https://bugs.launchpad.net/tripleo/+bug/1393298 | 22:00 |
| uvirtbot | Launchpad bug 1391656 in diskimage-builder "There is no way to customize the installation of the bootloader" [Undecided,New] | 22:00 |
| uvirtbot | Launchpad bug 1393298 in tripleo "keepalived VRRP_Script(chk_haproxy) timed out" [Undecided,In progress] | 22:00 |
| *** untriaged-bot has quit IRC | 22:00 | |
| *** noslzzp has joined #tripleo | 22:35 | |
| *** marun has quit IRC | 22:36 | |
| *** marun has joined #tripleo | 22:38 | |
| *** marun has quit IRC | 22:46 | |
| mordred | lifeless, greghaynes: so - any chance either of you can explain 15-remove-grub to me? | 22:48 |
| mordred | specifically, ./elements/redhat-common/pre-install.d/15-remove-grub | 22:48 |
| mordred | ESPECIALLY considering that grub2 is in ./elements/redhat-common/pre-install.d/package-installs-redhat-common | 22:49 |
| greghaynes | mordred: its in there as a - (meaning remove) | 22:50 |
| greghaynes | and IIRC removal happens at the end of phase | 22:51 |
| mordred | greghaynes: I thought - was "install this but remove it at the end of the phase" | 22:51 |
| mordred | yup. I'm pretty sure there is just a bug here | 22:52 |
| greghaynes | yea, that looks correct | 22:52 |
| mordred | because whathappens is package-installs installs grub, then remove grub removes it | 22:52 |
| mordred | then the - in the list causes package-uninstall to try to uninstall it | 22:52 |
| greghaynes | slagle: ^ | 22:52 |
| mordred | which does not work | 22:52 |
| mordred | because it already got uninstalled | 22:52 |
| mordred | I'm guessing that perhaps some other element accidentally causes grub2 to get installed which is why this hasn't been seen before | 22:54 |
| mordred | but I have no proof of that - only a guess because I happen to be building thigns with a massively small list of elements | 22:55 |
| lifeless | f you're working on a rinsed image | 22:55 |
| lifeless | then yeah | 22:55 |
| lifeless | we worked with the big fedora images | 22:55 |
| mordred | I am indeed | 22:55 |
| mordred | I'm darned close to getting it to work though :) | 22:55 |
| lifeless | 'big' being relative, but they aren't micro | 22:55 |
| mordred | (in case folks aren't also reading -infra - I'm working on elements to produce stable base images for our long lived servers, since our cloud providers do strange things) | 22:56 |
| lifeless | orly | 22:57 |
| lifeless | they do?! | 22:57 |
| mordred | zomg | 22:57 |
| greghaynes | https://review.openstack.org/#/c/108762/11/elements/redhat-common/pre-install.d/15-remove-grub | 22:57 |
| greghaynes | Seems like this discussion has been had before :) | 22:57 |
| mordred | don't EVEN get me started on the strange things | 22:57 |
| mordred | greghaynes: you were right in your comment | 22:58 |
| mordred | or - you were right that there is a problem | 22:58 |
| *** jang1 has quit IRC | 22:59 | |
| greghaynes | so yea, I think we want to just remove that -grub2 in that package-installs file | 22:59 |
| mordred | yah | 23:00 |
| mordred | also | 23:00 |
| mordred | install-packages -m base vlan_package iscsi_package | 23:00 |
| mordred | in base | 23:00 |
| mordred | seem like they should be in a tripleo element somewhere | 23:00 |
| * mordred adds -n to his invocation | 23:02 | |
| greghaynes | one thing thats bugging me is I cant find anywhere else we install grub2 in pre-install.d phase | 23:04 |
| greghaynes | but it could just be a dependency of some other package | 23:04 |
| mordred | lifeless: so - amongst the fun things are that rackspace adds in-instance agents that we have no use for | 23:04 |
| lifeless | greghaynes: the base image has it installed | 23:04 |
| lifeless | greghaynes: as a bootloader | 23:04 |
| lifeless | greghaynes: IIRC | 23:04 |
| greghaynes | Yes, but then we remove it with yum | 23:04 |
| mordred | lifeless: and HP decides that they should spin new images from time to time and upload them with new names while changing the name of the old image | 23:05 |
| greghaynes | im just not sure how we werent running into this issue with double remove already... | 23:05 |
| greghaynes | Gotta keep the users on their toes | 23:05 |
| mordred | lifeless: and then, also, I'm looking at making base infra images with our public key baked in so that we don't have to deal with cloud-init or keypair-extention or config drive or ec2-metadata | 23:07 |
| mordred | BECAUSE - if I do that, then I don't have to deal with the ubuntu user on ubuntu and the cloud-user on centos and the whatever fedora does | 23:07 |
| mordred | which is just bonghits in the first place | 23:08 |
| clarkb | fedora is fedora | 23:08 |
| mordred | there it is | 23:08 |
| clarkb | its massive bong hits | 23:08 |
| mordred | so much bonghits | 23:08 |
| mordred | clarkb: I'm very close | 23:08 |
| lifeless | cloud-init can override the username, FWIW | 23:08 |
| mordred | clarkb: although there are portions of the script that are going to make you laugh | 23:08 |
| mordred | lifeless: I want to delete cloud-init | 23:09 |
| lifeless | so you can use cloud-init *and* have infra as the default user everywhere, if you want | 23:09 |
| mordred | lifeless: I find cloud-init to be very annoying to work with - but on the other hand, you have written me this wonderful disk-image-builder tool | 23:09 |
| lifeless | hah | 23:09 |
| mordred | so I will use it to make wonderful disk images | 23:09 |
| lifeless | so cloud-init-resizeroot stuff and the early network setup I find useful | 23:10 |
| lifeless | but | 23:10 |
| lifeless | if you want it diferent, cool. More use cases == more awesomes | 23:10 |
| mordred | yup | 23:10 |
| mordred | this is one of the reasons I'm word vomiting in here - I'm exercising strange code paths :) | 23:10 |
| lifeless | mordred: btw | 23:10 |
| lifeless | mordred: have you heard of a plan to move openstack away from eventlet to asyncio ? | 23:11 |
| mordred | lifeless: I have heard constant grumbling and intent | 23:11 |
| lifeless | mordred: I know /some/ folk are interested in such a thing, but what I'm referring to is an agreed on decision | 23:11 |
| mordred | lifeless: I'd say "plan" might be a bit strong | 23:11 |
| clarkb | trollius | 23:11 |
| lifeless | similar to the move from twisted to eventlet actual thing that happened | 23:11 |
| lifeless | (which I wasn't around for so only know by hearsay) | 23:11 |
| mordred | lifeless: I think dhellmann is the holder of the plan on that though | 23:12 |
| lifeless | surely its a tc thing | 23:12 |
| lifeless | since its going to involve a rewrite of every project | 23:12 |
| mordred | not the plan | 23:12 |
| lifeless | the agreement | 23:12 |
| lifeless | not the plan | 23:12 |
| clarkb | its a per project thing right? | 23:12 |
| mordred | I believe the TC are all in general agreement that we'd like to move away from eventlet | 23:12 |
| mordred | because we have to to get to python3 | 23:12 |
| mordred | I believe a concrete plan is still seen as a thing that someone will figure out at some point in the future | 23:13 |
| mordred | because everytime I ask it's always "well, some day we'll ..." | 23:13 |
| lifeless | eventlet has python3 | 23:13 |
| mordred | oh! when did that happen? | 23:13 |
| lifeless | the patches are being merged at the moment | 23:13 |
| mordred | gotcha | 23:13 |
| mordred | that might remove the pressure on that topic then | 23:13 |
| mordred | python3 was the driving factor | 23:14 |
| lifeless | an eventlet committer was in the python3 session for oslo in Paris | 23:14 |
| *** avozza is now known as zz_avozza | 23:14 | |
| lifeless | mordred: see my mail in 'Re: [openstack-dev] [oslo] Add a new aiogreen executor for Oslo Messaging' [just sent] | 23:14 |
| lifeless | mordred: please correct me as needed! | 23:14 |
| mordred | lifeless: it seems that you know at least 100x more than I do on the specifics here | 23:15 |
| lifeless | yes, but I'm not, thankfully, on the TC | 23:15 |
| lifeless | or I wouldn't be doing WTH I want right now | 23:15 |
| lifeless | I'd be accountable and responsible :) | 23:16 |
| *** chlong has joined #tripleo | 23:17 | |
| mordred | lifeless: ah - so the reasons we moved away from twisted were: | 23:20 |
| mordred | a) flame war | 23:20 |
| mordred | b) twisted is hard for people to understand | 23:21 |
| mordred | I'm not saying those are good or correct reasons, merely that they are the raesons (b was the real reason, but finding a conclusion to a was also important) | 23:22 |
| mordred | twisted v. eventlet, xml v. json and git v. bzr are the three largest flame wars in our history, fwiw | 23:22 |
| lifeless | so asyncio is going to be approx the same as twisted for understanding | 23:23 |
| lifeless | IMNSHO | 23:23 |
| mordred | also - it's worth noting taht it wasn't purely a "move away from twisted" - nova was in twisted, swift was in eventlet | 23:23 |
| lifeless | interesting | 23:23 |
| mordred | so half of the project was in one, and half of the project was in the other | 23:23 |
| lifeless | FWIW many folk don't understand eventlet at all | 23:23 |
| lifeless | :) | 23:23 |
| mordred | yah | 23:23 |
| mordred | I'm one of them | 23:23 |
| clarkb | asyncio is very similar to eventlet | 23:23 |
| lifeless | just keep saying trampoline | 23:24 |
| lifeless | clarkb: oh? | 23:24 |
| mordred | I'm really not trolling when I say we should ditch all this crap and do explicit threads | 23:24 |
| clarkb | at least when compared to twisted | 23:24 |
| lifeless | mordred: or processes :) | 23:24 |
| mordred | lifeless: sure | 23:24 |
| mordred | lifeless: either one work for me | 23:24 |
| lifeless | clarkb: I don't get that at all. | 23:24 |
| lifeless | clarkb: convince me? | 23:24 |
| clarkb | so similar twisted devs complain about both in the same bucket | 23:24 |
| lifeless | how so? | 23:25 |
| clarkb | unlike twisted you dont deal in futures with a massive loop | 23:25 |
| lifeless | I am concerned that I've fundamentally missed some thing about eventlet | 23:25 |
| clarkb | instead you pass control between green threads | 23:25 |
| clarkb | so the abstraction is similar to eventlet | 23:26 |
| clarkb | but eventlet does a bit more hand holding aiui | 23:26 |
| lifeless | clarkb: huh | 23:26 |
| lifeless | clarkb: I don't agree with anything you said | 23:26 |
| clarkb | ok :) | 23:26 |
| lifeless | clarkb: or rather it doesn't agree with my model of reality | 23:26 |
| lifeless | eventlet [as we use it] causes *every* blocking call in Python to trampoline out of a greenthread that has been patched into the interpreter | 23:27 |
| lifeless | asyncio is bog standard generator co-routines | 23:28 |
| lifeless | twisted has a decorator that turns such coroutines into regular twisted deferreds | 23:29 |
| lifeless | I forget its name | 23:29 |
| lifeless | [not asyncio specific ones - the exact calls etc are different] | 23:29 |
| clarkb | sure but control is explicit in asyncio and you have a similar abstraction of these thread like things | 23:29 |
| clarkb | with twisted you have futures and you throw them at arun loop to be executed | 23:29 |
| clarkb | its a very differwnt mental model | 23:29 |
| lifeless | you have Futures in asyncio to | 23:30 |
| lifeless | and asyncio is based on a reactor which pushes data into handlers as needed | 23:30 |
| clarkb | but that isnt how you program in it aiui | 23:31 |
| lifeless | sure it is | 23:31 |
| clarkb | the underlying implementation isnt super important | 23:31 |
| clarkb | how I make code go vroom on it is | 23:31 |
| lifeless | https://docs.python.org/3/library/asyncio-eventloop.html#hello-world-with-call-soon | 23:31 |
| lifeless | http://stackoverflow.com/questions/14751138/http-echo-server-and-client-twisted | 23:32 |
| lifeless | http://eventlet.net/doc/examples.html | 23:32 |
| lifeless | which one is the odd one out? | 23:32 |
| mordred | ZOMG. WHY DOES THE git PACKAGE ON REDHAT DEPEND ON A BILLION PERL LIBRARIES?????? | 23:33 |
| clarkb | mordred git is much perl | 23:33 |
| clarkb | lifeless all of the examples I had seen were explicit coroutine flow control | 23:33 |
| clarkb | so asyncio is reaching to both camps? | 23:33 |
| mordred | clarkb: how about libgnome-keyring? | 23:33 |
| mordred | clarkb, lifeless: yes - I believe asyncio is trying to provide love to both programming models | 23:34 |
| clarkb | woo python and one way to do things | 23:35 |
| mordred | well piddle. the package I wanted to exclude that redhat-common pulls in gets pulled in by puppet anyway | 23:35 |
| mordred | ho-hum | 23:35 |
| lifeless | clarkb: an explicit design goal of asyncio is to be a unifying thing | 23:38 |
| lifeless | clarkb: but eventlets /point/ is the trampoline | 23:38 |
| lifeless | clarkb: thats what makes it so different to both asyncio an twisted, as neither have that | 23:39 |
| lifeless | IMO | 23:39 |
| clarkb | thats an implementation detail. for me the programming model is more interesting and looks like you can be like twisted or eventlet with asyncio | 23:40 |
| lifeless | clarkb: but that *is* the programming model in the OpenStack context. | 23:41 |
| clarkb | people dont have trouble with trampolines. they have trouble with futures and coroutines and threads | 23:41 |
| lifeless | clarkb: we write 'threads' | 23:41 |
| lifeless | clarkb: you don't write 'threads' in asyncio or in Twisted. | 23:42 |
| lifeless | clarkb: you write futures and coroutines. In both. | 23:42 |
| mordred | lifeless: where do we write threads? | 23:44 |
| *** enterprisedc_ has joined #tripleo | 23:45 | |
| lifeless | mordred: 99% of the code in e.g. nova is synchronous blocking code. | 23:45 |
| lifeless | mordred: (which hands off to another green thread magically) | 23:45 |
| lifeless | mordred: all of the big long functions, for instance which call into libvirt, or poll ironic in a loop. | 23:46 |
| *** enterprisedc has quit IRC | 23:46 | |
| *** enterprisedc_ is now known as enterprisedc | 23:46 | |
| mordred | lifeless: wow. ok. so, I guess there is a really good reason I don't look in nova very often | 23:46 |
| lifeless | the programming model is one of threads. the implementation is one of intercepted syscalls and greenthreads and trampolines. | 23:46 |
| mordred | then why not actually just use threads if the programming model is threads? | 23:47 |
| lifeless | thats a good question :) | 23:47 |
| lifeless | But you see why I am concerned about asyncio- not because its bad(I rather like it) | 23:47 |
| lifeless | but because we moved away from a programming model that is not very different | 23:48 |
| lifeless | to asyncio's | 23:48 |
| lifeless | at least AIU asyncio, which I may not and I need some examples from clarkb showing what he thinks of as asyncio code | 23:48 |
| clarkb | the magical yield from | 23:48 |
| mordred | lifeless: but the fellow in the mailing list claims to have an eventlet compatible thing done on top of asyncio | 23:49 |
| clarkb | I see coroutines as similar to green threads | 23:49 |
| lifeless | btw the swift code also reads like threaded code | 23:49 |
| lifeless | mordred: yes, but thats transitional | 23:49 |
| clarkb | the major difference in control flow is explicit yielding | 23:49 |
| lifeless | mordred: read the detail: its a thunk to let eventlet keep working, then the goal is to migrate to direct asyncio and stop using eventlet at all | 23:49 |
| clarkb | so instead of trampolines you have street signs | 23:49 |
| lifeless | so it makes potential blocking points explicit | 23:50 |
| lifeless | and you have to not use any primitives that might themselves block without wrapping them with a real thread | 23:51 |
| lifeless | thats the delta right? | 23:51 |
| lifeless | right now we can't tell whats going to block, and we dont' need to wrap anything because $magic, even to the exent of threading.locals being context switched out | 23:51 |
| lifeless | extent | 23:52 |
| clarkb | ya | 23:52 |
| greghaynes | yea, you dont get to block in asyncio, its just tying blocking resources into an event loop via yield from | 23:52 |
| clarkb | eventlet is definielty more magic so not 1:1 | 23:52 |
| lifeless | so you have to do things like | 23:54 |
| lifeless | 'oh this might block, factor it into a co-routine' | 23:54 |
| mordred | greghaynes: I get a lot of these: WARNING: map-packages is deprecated. Please use the pkg-map element. | 23:54 |
| lifeless | which folk don't today. | 23:54 |
| mordred | greghaynes: I'd like to use the pkg-map element, but I think that's out of my control :) | 23:54 |
| greghaynes | mordred: Yep, do that | 23:54 |
| mordred | greghaynes: well, I'm not using either | 23:55 |
| greghaynes | mordred: ah | 23:55 |
| mordred | directly | 23:55 |
| lifeless | mordred: then ignore it | 23:55 |
| greghaynes | I guess it would be a good idea for us to clean those up at some point | 23:55 |
| mordred | greghaynes: ok - can you explain the intent of the grub install/uninstall magic? | 23:56 |
| mordred | greghaynes: because I just hit this: | 23:56 |
| mordred | ++ rpm -i /tmp/grub/grub2-2.02-0.2.10.el7.centos.1.x86_64.rpm | 23:56 |
| mordred | error: Failed dependencies: | 23:56 |
| mordred | dracut is needed by grub2-1:2.02-0.2.10.el7.centos.1.x86_64 | 23:56 |
| mordred | greghaynes: but I don't know why we install and then uninstall, then install-download, then copy for later, then rpm -i | 23:57 |
| mordred | so it's hard to debug what _should_ happen | 23:57 |
| greghaynes | I actually dont know the original reason we do it... I assumed it was because we wanted to force grub to reconfigure | 23:58 |
| mordred | lifeless: ^^ do you know? | 23:58 |
| lifeless | fedora, no sorry | 23:59 |
| lifeless | the broad picture is that | 23:59 |
| lifeless | we need to reconfigure because the defaults are poor | 23:59 |
| lifeless | and sometimes we need to install a bootblock (the vm element) | 23:59 |
| lifeless | which you probably are in the case of | 23:59 |
| mordred | yah | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!