| *** jralph_ has quit IRC | 00:00 | |
| *** sacharya has quit IRC | 00:00 | |
| *** sacharya has joined #openstack-ironic | 00:00 | |
| *** jralph has joined #openstack-ironic | 00:02 | |
| *** cdearborn has quit IRC | 00:17 | |
| *** jerrygb has quit IRC | 00:18 | |
| *** rbudden has quit IRC | 00:21 | |
| *** baoli has quit IRC | 00:26 | |
| *** mtan_____ has joined #openstack-ironic | 00:28 | |
| *** [1]rpioso has quit IRC | 00:28 | |
| *** mtanino has quit IRC | 00:29 | |
| *** jerrygb has joined #openstack-ironic | 00:34 | |
| *** hoangcx has joined #openstack-ironic | 00:41 | |
| *** Sukhdev has joined #openstack-ironic | 00:52 | |
| *** SerenaFeng has joined #openstack-ironic | 00:53 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/420835 | 00:55 |
|---|---|---|
| *** cdearborn has joined #openstack-ironic | 01:02 | |
| *** phuongnh has joined #openstack-ironic | 01:03 | |
| *** hw_wutianwei has joined #openstack-ironic | 01:10 | |
| *** jerrygb has quit IRC | 01:13 | |
| *** amotoki has quit IRC | 01:19 | |
| *** amotoki has joined #openstack-ironic | 01:19 | |
| *** trananhkma has quit IRC | 01:24 | |
| *** Sukhdev has quit IRC | 01:29 | |
| *** SerenaFeng has quit IRC | 01:34 | |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add soft reboot/poweroff power states. https://review.openstack.org/247904 | 01:48 |
| *** mtan_____ has quit IRC | 01:57 | |
| *** sacharya has quit IRC | 01:59 | |
| *** cdearborn has quit IRC | 02:00 | |
| *** sacharya has joined #openstack-ironic | 02:00 | |
| *** tiendc has joined #openstack-ironic | 02:02 | |
| *** jerrygb has joined #openstack-ironic | 02:14 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal node cmds https://review.openstack.org/357639 | 02:18 |
| *** cdearborn has joined #openstack-ironic | 02:20 | |
| *** gcb has joined #openstack-ironic | 02:36 | |
| openstackgerrit | Taku Izumi proposed openstack/ironic-python-agent: Add support for finalize_rescue https://review.openstack.org/417351 | 02:50 |
| *** amotoki has quit IRC | 02:55 | |
| *** yuanying has quit IRC | 03:01 | |
| *** yuanying has joined #openstack-ironic | 03:01 | |
| *** yuanying has quit IRC | 03:01 | |
| *** yuanying has joined #openstack-ironic | 03:02 | |
| *** xuao has joined #openstack-ironic | 03:03 | |
| *** yuanying has quit IRC | 03:06 | |
| *** amotoki has joined #openstack-ironic | 03:13 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 03:26 | |
| *** Nisha_ has joined #openstack-ironic | 03:27 | |
| *** vikrant has joined #openstack-ironic | 03:29 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add nodes tagging support - objs https://review.openstack.org/233357 | 03:29 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add DB layer support for node tags filter https://review.openstack.org/272624 | 03:30 |
| *** Nisha_Agarwal has quit IRC | 03:31 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 03:31 | |
| *** links has joined #openstack-ironic | 03:32 | |
| *** Nisha_Agarwal has quit IRC | 03:32 | |
| *** cdearborn has quit IRC | 03:33 | |
| *** Nisha_ has quit IRC | 03:34 | |
| *** vikrant is now known as vikrant|brb | 03:44 | |
| *** amotoki has quit IRC | 03:48 | |
| *** amotoki has joined #openstack-ironic | 03:51 | |
| *** vikrant|brb is now known as vikrant | 03:58 | |
| *** cfarquhar has quit IRC | 04:00 | |
| *** yuanying has joined #openstack-ironic | 04:02 | |
| *** vsaienko has joined #openstack-ironic | 04:05 | |
| *** amotoki has quit IRC | 04:05 | |
| *** Syed__ has quit IRC | 04:25 | |
| *** jrist has quit IRC | 04:27 | |
| *** wrouesnel has quit IRC | 04:31 | |
| *** vsaienko has quit IRC | 04:31 | |
| *** udesale has joined #openstack-ironic | 04:35 | |
| *** wrouesnel has joined #openstack-ironic | 04:39 | |
| *** jrist has joined #openstack-ironic | 04:39 | |
| *** Sukhdev has joined #openstack-ironic | 04:45 | |
| *** amotoki has joined #openstack-ironic | 04:47 | |
| *** wrouesnel has quit IRC | 04:55 | |
| *** yuanying has quit IRC | 04:55 | |
| *** yuanying has joined #openstack-ironic | 05:03 | |
| *** wrouesnel has joined #openstack-ironic | 05:05 | |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/420835 | 05:05 |
| *** jerrygb has quit IRC | 05:08 | |
| *** wrouesnel has quit IRC | 05:17 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add DB layer support for node tags filter https://review.openstack.org/272624 | 05:26 |
| *** wrouesnel has joined #openstack-ironic | 05:27 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal port cmds https://review.openstack.org/375971 | 05:33 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Enhance pxe_ilo driver to do inband-inspection https://review.openstack.org/343577 | 05:33 |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal port cmds https://review.openstack.org/375971 | 05:36 |
| *** yuanying has quit IRC | 05:46 | |
| *** yuanying has joined #openstack-ironic | 05:47 | |
| *** SerenaFeng has joined #openstack-ironic | 05:47 | |
| *** yuanying has quit IRC | 05:51 | |
| openstackgerrit | Aparna proposed openstack/proliantutils: Add `erase_devices` as a clean step https://review.openstack.org/421081 | 05:55 |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal chassis cmds https://review.openstack.org/375972 | 05:55 |
| *** wrouesnel has quit IRC | 06:00 | |
| *** wrouesnel has joined #openstack-ironic | 06:01 | |
| *** pc_m has quit IRC | 06:03 | |
| *** e0ne has quit IRC | 06:04 | |
| *** rbartal has joined #openstack-ironic | 06:09 | |
| *** rama_y has quit IRC | 06:15 | |
| *** amotoki has quit IRC | 06:16 | |
| *** vmud213 has joined #openstack-ironic | 06:21 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal driver,portgroup cmds https://review.openstack.org/421087 | 06:23 |
| *** rbudden has joined #openstack-ironic | 06:23 | |
| *** SerenaFeng has quit IRC | 06:23 | |
| *** SerenaFeng has joined #openstack-ironic | 06:28 | |
| *** sacharya has quit IRC | 06:28 | |
| *** sacharya has joined #openstack-ironic | 06:29 | |
| *** Sukhdev has quit IRC | 06:31 | |
| *** mjura has joined #openstack-ironic | 06:38 | |
| *** amotoki has joined #openstack-ironic | 06:43 | |
| *** Sukhdev has joined #openstack-ironic | 06:43 | |
| *** amotoki has quit IRC | 06:49 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal chassis cmds https://review.openstack.org/375972 | 06:50 |
| *** ohamada has joined #openstack-ironic | 06:56 | |
| *** rbartal has quit IRC | 06:57 | |
| *** moshele has joined #openstack-ironic | 06:59 | |
| *** dsneddon has quit IRC | 06:59 | |
| *** dsneddon has joined #openstack-ironic | 06:59 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal driver,portgroup cmds https://review.openstack.org/421087 | 07:00 |
| *** pc_m has joined #openstack-ironic | 07:02 | |
| *** rbudden has quit IRC | 07:02 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic: Ironic devstack portgroup support https://review.openstack.org/381743 | 07:04 |
| *** tuanluong has joined #openstack-ironic | 07:09 | |
| *** rbartal has joined #openstack-ironic | 07:10 | |
| *** tesseract has joined #openstack-ironic | 07:16 | |
| *** yuanying has joined #openstack-ironic | 07:23 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-python-agent: Use sh instead of bash when installing grub https://review.openstack.org/421102 | 07:27 |
| *** ChubYann has quit IRC | 07:32 | |
| xek | morning Ironic! | 07:37 |
| xek | mariojv, I think your proposal makes sense, I'll make the change | 07:37 |
| tuanluong | morning ironic | 07:41 |
| vsaienk0 | morning Ironic! | 07:41 |
| *** Sukhdev has quit IRC | 07:53 | |
| *** aparnav has joined #openstack-ironic | 07:55 | |
| openstackgerrit | jiang wei proposed openstack/python-ironicclient: Support i18n for baremetal driver,portgroup cmds https://review.openstack.org/421087 | 07:58 |
| *** dmellado has quit IRC | 08:02 | |
| *** dmellado has joined #openstack-ironic | 08:04 | |
| mkrai | Hi Ironic Team | 08:09 |
| mkrai | I have an Ironic setup with devstack | 08:10 |
| mkrai | I tried to add a new hardware to ironic's node-list | 08:11 |
| mkrai | It was added and I was able to powe on/off using ironic CLI | 08:11 |
| mkrai | But it failed while deploying with error "Failed to prepare to deploy. Error: No VIFs found for node 2cd3edca-1261-4ba6-8b7e-fc6037959901 when attempting to update DHCP BOOT options." | 08:12 |
| mkrai | Can anyone please help? | 08:12 |
| mkrai | I have also tried booting via Nova and it failed with "no valid host" and seeing the logs I found out that the hypervisor-stats show 0 resources | 08:14 |
| *** amoralej|off is now known as amoralej | 08:15 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Use TENANT_VIF_KEY constant everywhere https://review.openstack.org/421121 | 08:17 |
| *** trown|outtypewww has quit IRC | 08:21 | |
| *** trown has joined #openstack-ironic | 08:22 | |
| openstackgerrit | Aparna proposed openstack/proliantutils: Add `erase_devices` into `get_clean_step` method https://review.openstack.org/421081 | 08:26 |
| *** ralonsoh has joined #openstack-ironic | 08:30 | |
| *** amotoki has joined #openstack-ironic | 08:34 | |
| *** athomas has joined #openstack-ironic | 08:35 | |
| *** e0ne has joined #openstack-ironic | 08:48 | |
| *** priteau has joined #openstack-ironic | 08:48 | |
| *** yibo has joined #openstack-ironic | 08:55 | |
| *** zzzeek has quit IRC | 09:00 | |
| *** zzzeek has joined #openstack-ironic | 09:00 | |
| *** tuanluong has quit IRC | 09:03 | |
| *** tuanluong has joined #openstack-ironic | 09:04 | |
| *** rama_y has joined #openstack-ironic | 09:12 | |
| *** dtantsur|afk is now known as dtantsur | 09:13 | |
| dtantsur | morning Ironic | 09:13 |
| mkrai | Gm dtantsur | 09:15 |
| mkrai | I have a question regarding ironic | 09:16 |
| mkrai | I have posted the questions ^^ | 09:16 |
| *** mgoddard has joined #openstack-ironic | 09:16 | |
| mkrai | Can you please have a look if you have time? | 09:16 |
| *** tiendc has quit IRC | 09:23 | |
| *** tiendc has joined #openstack-ironic | 09:23 | |
| tuanluong | morning Ironic, dtantsur mkrai | 09:24 |
| dtantsur | mkrai, you should boot ironic nodes with nova for the beginning. please refer to http://docs.openstack.org/developer/ironic/deploy/troubleshooting.html#nova-returns-no-valid-host-was-found-error | 09:25 |
| dtantsur | morning tuanluong | 09:25 |
| openstackgerrit | Merged openstack/proliantutils: Add `erase_devices` into `get_clean_step` method https://review.openstack.org/421081 | 09:26 |
| mkrai | dtantsur: Thanks. I tried booting via nova also | 09:30 |
| mkrai | Let me go through the link | 09:30 |
| *** derekh has joined #openstack-ironic | 09:38 | |
| mkrai | dtantsur: The issue is the hypervisor stats shows ram:0 disk:0 etc so nova-scheduler returns 0 hosts | 09:38 |
| mkrai | How is the ironic node stats reported to nova? | 09:39 |
| dtantsur | mkrai, periodically, via polling. check other bullet points, you may not have enough "available" nodes, maybe their properties are not populated correctly, etc | 09:40 |
| mkrai | Ok nova polls ironic or is it ironic sending the data to nova? | 09:41 |
| dtantsur | nova polls ironic | 09:43 |
| mkrai | Ok thanks | 09:43 |
| *** lucas-afk is now known as lucasagomes | 09:45 | |
| lucasagomes | morning all | 09:45 |
| *** rama_y has quit IRC | 09:49 | |
| vdrok | good morning ironic! xek tuanluong mkrai dtantsur lucasagomes | 09:52 |
| mkrai | dtantsur: nova baremetal flavor show capabilities:boot_option=local | 09:53 |
| mkrai | How do we add this to node.properties | 09:53 |
| dtantsur | then it has to match your nodes | 09:53 |
| dtantsur | why do you have it in your flavor? | 09:53 |
| tuanluong | morning vdrok | 09:53 |
| dtantsur | morning lucasagomes, vdrok | 09:53 |
| mkrai | This doesn't match our pattern | 09:53 |
| mkrai | This is in nova's flavor | 09:53 |
| mkrai | Not in ironic node's properties | 09:54 |
| dtantsur | mkrai, how did it end up there? maybe you should remove it from the flavor? | 09:54 |
| mkrai | Ok | 09:54 |
| mkrai | I will remove it | 09:54 |
| tuanluong | mkrai, ironic node-update $NODE add properties/capabilities="{'boot_mode':'bios','boot_option':'local'}" | 09:54 |
| mkrai | tuanluong: thanks! | 09:55 |
| vdrok | mikal: you could also check the docs https://github.com/openstack/ironic/blob/6394c279ca298afc5887ac884df35bfa36c3dd9c/install-guide/source/include/local-boot-partition-images.rst#L33 | 09:55 |
| vdrok | ouch, sorry, mkrai | 09:55 |
| mkrai | vdrok: Thanks that is helpful | 09:56 |
| openstackgerrit | Nisha Agarwal proposed openstack/proliantutils: snmp inspection for disk size https://review.openstack.org/394858 | 09:57 |
| *** SerenaFeng has quit IRC | 09:57 | |
| *** xuao has quit IRC | 10:09 | |
| *** SerenaFeng has joined #openstack-ironic | 10:09 | |
| *** hoangcx has quit IRC | 10:10 | |
| *** vsaienko has joined #openstack-ironic | 10:10 | |
| *** phuongnh has quit IRC | 10:11 | |
| pas-ha | morning all :) | 10:13 |
| dtantsur | morning pas-ha | 10:13 |
| mkrai | dtantsur: it passed but the node-state is stuck n "wait call-back" | 10:17 |
| *** MattMan has quit IRC | 10:17 | |
| *** SerenaFeng has quit IRC | 10:18 | |
| *** MattMan has joined #openstack-ironic | 10:18 | |
| mkrai | vdrok: tuanluong ^^ | 10:23 |
| *** SerenaFeng has joined #openstack-ironic | 10:23 | |
| *** udesale has quit IRC | 10:26 | |
| *** amotoki has quit IRC | 10:27 | |
| *** udesale has joined #openstack-ironic | 10:27 | |
| vdrok | mkrai: by stuck, you mean the ironic python agent has not started up on node? | 10:27 |
| vdrok | morning pas-ha | 10:27 |
| *** sambetts|afk is now known as sambetts | 10:27 | |
| sambetts | Morning all | 10:27 |
| aarefiev | morning dtantsur, vdrok, pas-ha, sambetts | 10:28 |
| dtantsur | morning sambetts, aarefiev | 10:29 |
| vdrok | morning sambetts and aarefiev | 10:29 |
| mkrai | vdrok: The "provisioning state" is "wait call-back" for very long time period | 10:30 |
| vdrok | mkrai: could you check the node console? to see what's going on there | 10:30 |
| mkrai | What does this state mean? | 10:30 |
| sambetts | mkrai: sounds like your nodes aren't able to boot the ramdisk or the ramdisk is unable to call-back to Ironic to tell, us it has booted and ready | 10:31 |
| vdrok | mkrai: it means that ironic waits for callback from agent | 10:31 |
| *** vikrant has quit IRC | 10:31 | |
| mkrai | To add, I am on proxy network | 10:31 |
| mkrai | So do I need to add the node's IP to no_proxy list or something like that? | 10:32 |
| sambetts | mkrai: your nodes you are booting need to be able to talk to the Ironic API, and the Ironic conductor needs to be able to talk to your nodes | 10:32 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Devstack: don't rely on Ironic local vars https://review.openstack.org/420725 | 10:33 |
| *** aparnav has quit IRC | 10:36 | |
| *** vikrant has joined #openstack-ironic | 10:44 | |
| *** dtantsur is now known as dtantsur|bbl | 10:48 | |
| *** udesale has quit IRC | 10:48 | |
| mkrai | vdrok: what is the way to check whether console is supported on the h/w or not? | 10:48 |
| amarao | Hello, everyone. Before I file a bug: Does ironic+nova support for 'rebuild' command? I see it didn't removed old port and just created new (so there are two neutron baremetall ports now with same mac in different nets). Should I report this? | 10:48 |
| *** udesale has joined #openstack-ironic | 10:49 | |
| mkrai | I am new to ironic so my questions might be weird sometime, please pardon | 10:49 |
| mkrai | sambetts: I am able to powe on/off the node, does it mean the communication from conductor to node is successful? | 10:54 |
| mkrai | How do I make sure for the part from node to ironic-api? | 10:54 |
| *** udesale__ has joined #openstack-ironic | 10:54 | |
| *** udesale has quit IRC | 10:57 | |
| amarao | https://bugs.launchpad.net/ironic/+bug/1657083 | 11:01 |
| openstack | Launchpad bug 1657083 in Ironic "instance rebuild cause creation of two ports in two networks simultaneously" [Undecided,New] | 11:01 |
| sambetts | mkrai: that means that your conductor can talk on your servers management network, but the conductor needs to be connected to your nodes data network in order to be able to transfer the images to the node | 11:03 |
| *** amotoki has joined #openstack-ironic | 11:03 | |
| *** g3ek has quit IRC | 11:03 | |
| mkrai | sambetts: how do I confirm that? | 11:03 |
| sambetts | mkrai: that depends on your equipment, but easiest way is to connect to the nodes KVM/visual console and confirm that the node is able to PXE boot | 11:05 |
| mkrai | the node's screen shows "PXE-E61 Media test failure, check cable" | 11:07 |
| *** amotoki has quit IRC | 11:09 | |
| vdrok | amarao: rebuild is supported, yes. during rebuild, ironic still creates port in provisioning network, then it is deleted when you switch to tenant network | 11:10 |
| sambetts | mkrai: that means that your server is unable to PXE boot because its network cable is broken or possibly disabled at the switch | 11:10 |
| mkrai | Thanks sambetts | 11:12 |
| *** g3ek has joined #openstack-ironic | 11:12 | |
| mkrai | I will try correcting the issue and then get back | 11:13 |
| mkrai | Thank you all for quick response :) | 11:13 |
| amarao | vdrok, ironic should remove old port before creating new. Otherwise it asks ML2 driver to put switch iface in untagged mode in two different vlans, which is impossible. | 11:14 |
| *** tiendc_ has joined #openstack-ironic | 11:15 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Pass os_identity_api_version into functional tests https://review.openstack.org/414957 | 11:16 |
| *** udesale__ has quit IRC | 11:17 | |
| *** tiendc has quit IRC | 11:18 | |
| *** cdearborn has joined #openstack-ironic | 11:22 | |
| jroll | morning y'all | 11:33 |
| *** dtantsur|bbl is now known as dtantsur | 11:35 | |
| dtantsur | morning jroll | 11:36 |
| lucasagomes | morning jroll | 11:36 |
| *** amotoki has joined #openstack-ironic | 11:37 | |
| *** jkilpatr has quit IRC | 11:44 | |
| *** moshele has quit IRC | 11:47 | |
| *** geertsky has joined #openstack-ironic | 11:47 | |
| geertsky | Newbie here... | 11:48 |
| geertsky | Bare metal provisioning WHOHOOOO!!! | 11:48 |
| jroll | geertsky: welcome :) | 11:48 |
| geertsky | But how is the software eventually installed? | 11:48 |
| geertsky | kickstart? image? | 11:49 |
| lucasagomes | geertsky, image | 11:49 |
| jroll | yeah, image | 11:49 |
| geertsky | aaah good! so how is this realized? during initrd? | 11:49 |
| lucasagomes | geertsky, the idea is make baremetal deployment looks&feel like deploying VMs in openstack (in the user POV) | 11:49 |
| jroll | geertsky: this may be helpful http://docs.openstack.org/developer/ironic/deploy/user-guide.html | 11:49 |
| jroll | tl;dr we boot a ramdisk running some python code, ironic-conductor process interacts with it to do the things | 11:50 |
| jroll | and then we reboot | 11:50 |
| jroll | there's also out of band control like ipmi of course | 11:50 |
| jroll | these flowcharts are super helpful: http://docs.openstack.org/developer/ironic/deploy/user-guide.html#example-1-pxe-boot-and-iscsi-deploy-process | 11:51 |
| geertsky | nice!! Sounds good man! actually flushes around 2 years of my work down the drain but hey! :D | 11:51 |
| jroll | heh | 11:53 |
| geertsky | I've been working for around 2 years on an implementation for xCAT to partition and deploy nodes using rsync. | 11:58 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-python-agent: Use sh instead of bash when installing grub https://review.openstack.org/421102 | 11:58 |
| *** tuanluong has quit IRC | 11:59 | |
| *** tiendc_ has quit IRC | 11:59 | |
| geertsky | It has been used on a 750 node cluster and all works fine. | 12:00 |
| geertsky | But due to bad advertising or something no further interest... | 12:01 |
| geertsky | Thanks allot for the pointers! I'll look into it! | 12:03 |
| geertsky | Cya! | 12:03 |
| *** geertsky has quit IRC | 12:03 | |
| mkrai | jroll: what's that python code running inside ramdisk? | 12:03 |
| mkrai | Is it some agent? | 12:03 |
| sambetts | mkrai: https://github.com/openstack/ironic-python-agent | 12:03 |
| *** lucasagomes is now known as lucas-hungry | 12:11 | |
| *** links has quit IRC | 12:12 | |
| *** rbudden has joined #openstack-ironic | 12:14 | |
| cdaniel-afk | morning all | 12:14 |
| *** cdaniel-afk is now known as cdaniel | 12:15 | |
| jroll | who was it that said the other day we don't pass network metadata into the configdrive yet? because we totally do :P | 12:17 |
| *** nicodemos has joined #openstack-ironic | 12:18 | |
| jroll | $ cat network_data.json | 12:18 |
| jroll | {"services": [], "networks": [{"network_id": "3140867c-7ad5-42e2-a7e3-f0114cdcf92e", | 12:18 |
| patchbot | Error: Missing "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. | 12:18 |
| jroll | etc | 12:18 |
| *** jkilpatr has joined #openstack-ironic | 12:18 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add os_project_name to functional tests configuration script https://review.openstack.org/408235 | 12:18 |
| openstackgerrit | Aparna proposed openstack/ironic: Doc for sanitize erase support in iLO drivers https://review.openstack.org/408441 | 12:19 |
| *** aparnav has joined #openstack-ironic | 12:21 | |
| sambetts | jroll: we pass it into the generator, but nova didn't used to include the file in the config driver, that must have fixed it at some point :/ | 12:21 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Change tenant to project https://review.openstack.org/420055 | 12:21 |
| jroll | sambetts: it's included it for quite a while now, afaik | 12:22 |
| *** tesseract has quit IRC | 12:23 | |
| jroll | sambetts: https://review.openstack.org/#/c/153097/ | 12:24 |
| patchbot | patch 153097 - nova - Expose Neutron network data in metadata service (MERGED) | 12:24 |
| jroll | july 2015 | 12:24 |
| jroll | :P | 12:24 |
| sambetts | jroll: hmmm... /me had to include this https://review.openstack.org/#/c/289412/1/nova/virt/ironic/driver.py@720 when I orginally wrote the code otherwise it wouldn't include the file | 12:24 |
| jroll | liberty, I guess | 12:24 |
| patchbot | patch 289412 - nova - [WIP] Allow mapping multiple Neutron ports to Iron... (ABANDONED) | 12:24 |
| nicodemos | good morning! | 12:25 |
| jroll | huh | 12:25 |
| dtantsur | am I doing something wrong, or is IPA CI really unstable? | 12:26 |
| amarao | jroll, there is a bug with metadata. | 12:27 |
| jroll | dtantsur: I haven't seen that but haven't looked at IPA much | 12:27 |
| amarao | jroll: https://bugs.launchpad.net/nova/+bug/1656854 | 12:27 |
| openstack | Launchpad bug 1656854 in OpenStack Compute (nova) "Incorrect metada in ConfigDrive when using barematal ports under neutron" [Undecided,In progress] - Assigned to Vasyl Saienko (vsaienko) | 12:27 |
| jroll | amarao: yes, I recall this from yesterday | 12:27 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add DB layer support for node tags filter https://review.openstack.org/272624 | 12:27 |
| *** skazi has quit IRC | 12:27 | |
| sambetts | jroll: I think that must have been against Kilo it seems :/ man have I really been working on this feature that long XD | 12:27 |
| amarao | I've actually run patched version already. I knew that this is crude fix, but it works. | 12:28 |
| jroll | sambetts: lol! | 12:28 |
| *** SerenaFeng has quit IRC | 12:30 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Pin basepython for docs tox target to python2 https://review.openstack.org/421244 | 12:30 |
| openstackgerrit | Merged openstack/ironic-ui: Refactor the edit-node/nodeUpdatePatch class https://review.openstack.org/417988 | 12:30 |
| jroll | I wonder if that network metadata bug is just nova caching network info too hard | 12:32 |
| jroll | amarao: any chance you could get nova-compute logs around where it's generating the configdrive? not sure if useful but it may be | 12:33 |
| sambetts | jroll: the bug is because at the point that nova generates the config-drive the ports are unbound | 12:33 |
| jroll | mmm | 12:33 |
| sambetts | jroll: so it calls the link type unbound which is an unsupported vif type | 12:33 |
| jroll | oh right, we bind them at the end of the deploy | 12:33 |
| jroll | urgh | 12:33 |
| sambetts | jroll: in our case all vifs are always bound, because they are physically | 12:33 |
| sambetts | existing | 12:33 |
| sambetts | s/vifs/nics | 12:33 |
| jroll | s/our/flat/ ? | 12:34 |
| sambetts | both, nics don't go away because of driver type :-P | 12:34 |
| jroll | uh | 12:34 |
| jroll | it's the multitenant stuff that breaks this, because it doesn't bind tenant ports until the deploy is done | 12:34 |
| jroll | whether they are physical or not | 12:35 |
| sambetts | right :/ but IMO all Ironic ports should be included in network_data.json in links because they physically exist, then the 'networks' section should be dedicated to VIF specific stuff | 12:36 |
| jroll | well, this is the 'links' section that's broken | 12:36 |
| sambetts | the reason I guess that nova puts them in there as unbound is because, in the VM case, when a NIC is unbound it won't physically exist on the VM and therefore won't get processed by cloud-init | 12:36 |
| *** vikrant has quit IRC | 12:37 | |
| jroll | I assume the neutron port returns unbound somewhere | 12:37 |
| jroll | right | 12:37 |
| jroll | or maybe nova converts status:down | 12:37 |
| jroll | but I don't see unbound anywhere relevant in nova | 12:37 |
| jroll | so no clue where that's coming from | 12:37 |
| jroll | neutron/extensions/portbindings.py | 12:38 |
| jroll | 81:VIF_TYPE_UNBOUND = 'unbound' | 12:38 |
| jroll | whomp | 12:38 |
| * jroll wonders if an ml2 plugin can fix that or if we have to hack around it | 12:38 | |
| *** hw_wutianwei has quit IRC | 12:39 | |
| sambetts | jroll: the ports will still be unbound in m-t case | 12:39 |
| * sambetts really wishes we could move config driver generation later | 12:39 | |
| jroll | well, yes | 12:39 |
| jroll | we need to hack around it somewhere | 12:40 |
| jroll | just trying to decide if that should be in the nova driver or ml2 plugin or cloud-init | 12:40 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Update troubleshooting docs on no valid host found error https://review.openstack.org/419595 | 12:40 |
| openstackgerrit | Merged openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/420206 | 12:40 |
| jroll | and we bind them after we write the configdrive, so moving the generation doesn't help :P | 12:40 |
| sambetts | jroll: oh yeah :/ fuuu... | 12:40 |
| sambetts | jroll: I think I have an idea for a fix, give me a minute :) | 12:41 |
| jroll | OTOH it's justification to just hack it | 12:41 |
| * jroll curious | 12:41 | |
| *** athomas has quit IRC | 12:42 | |
| *** dprince has joined #openstack-ironic | 12:43 | |
| amarao | jroll, do you need logs with my ugly patch or without it? I can gather them a bit later after next deployment attempt (chef cookbooks debug, etc). AFAIK there were no any useful info around this place. | 12:45 |
| jroll | dtantsur: what were you seeing around IPA CI? | 12:45 |
| jroll | amarao: don't worry about it, I was going down the wrong path | 12:45 |
| dtantsur | jroll, didn't check the exact error, but I see non-inspector jobs failing on a purely inspector change, so it may be not my fault | 12:47 |
| jroll | dtantsur: link? | 12:47 |
| jroll | :) | 12:47 |
| dtantsur | https://review.openstack.org/417871 | 12:47 |
| patchbot | patch 417871 - ironic-python-agent - Remove compatibility with old bash-based introspec... | 12:47 |
| jroll | thanks | 12:47 |
| jroll | I think julia mentioned lots of failures in ovh around nested virt crap | 12:47 |
| * jroll tries to remember how that presents | 12:48 | |
| *** skazi has joined #openstack-ironic | 12:49 | |
| jroll | ah, I think http://logs.openstack.org/71/417871/1/check/gate-tempest-dsvm-ironic-ipa-partition-pxe_ipmitool-coreos-src-ubuntu-xenial/70aaffc/logs/libvirt/qemu/node-0.txt.gz | 12:49 |
| jroll | maybe we should indeed turn that off | 12:50 |
| jroll | it makes things so fast when it works though :( | 12:50 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Do not use loopback BMC addresses for lookup https://review.openstack.org/419526 | 12:51 |
| dtantsur | jroll, sigh :( | 12:51 |
| jroll | indeed :/ | 12:51 |
| vdrok | morning jroll and cdaniel | 12:52 |
| jroll | ohai vdrok :) | 12:52 |
| *** athomas has joined #openstack-ironic | 12:52 | |
| *** hoangcx has joined #openstack-ironic | 12:54 | |
| * jroll breakfasts | 12:55 | |
| vdrok | jroll: that bug is fixed now in the port attach patch | 12:55 |
| vdrok | jroll: https://review.openstack.org/#/c/364413/22/nova/virt/ironic/driver.py@706 | 12:56 |
| patchbot | patch 364413 - nova - Support Ironic interface attach/detach in nova virt | 12:56 |
| jroll | vdrok: ah, just hardcoded? | 12:56 |
| cdaniel | morning vdrok :) | 12:56 |
| vdrok | jroll: yup. we do the same with ports inside portgroup in subsequent patch | 12:56 |
| *** hoangcx has quit IRC | 12:56 | |
| jroll | yeah, cool | 12:56 |
| openstackgerrit | Aparna proposed openstack/ironic: Doc for sanitize erase support in iLO drivers https://review.openstack.org/408441 | 12:56 |
| sambetts | vdrok, jroll: https://review.openstack.org/#/c/421254/ | 12:57 |
| patchbot | patch 421254 - nova (stable/newton) - Ensure we mark baremetal links as phy links | 12:57 |
| jroll | sambetts: agree that would fix it... now we'll need to get it in, in a backportable way | 12:58 |
| jroll | underneath the attach/detach stuff | 12:58 |
| sambetts | yeh... | 12:58 |
| vdrok | jroll: sambetts tho the thing is, It does not seem to actually break anything | 12:58 |
| *** gcb has quit IRC | 12:58 | |
| *** links has joined #openstack-ironic | 12:58 | |
| vdrok | I mean, the info is wrong, cloud-init ignores it, but interfaces still come uo | 12:58 |
| jroll | dtantsur: re different hardware types on each conductor, that throws quite the wedge in my RPC patch... I'll need to rpc to every conductor :/ | 12:58 |
| vdrok | up | 12:58 |
| vdrok | in case of usual ports | 12:59 |
| jroll | vdrok: even if you're in an environment without dhcp? :) | 12:59 |
| jroll | I assume they just happen to work because the system will configure them to dhcp and everybody is happy | 12:59 |
| jroll | but cloud-init needs to configure them if we're using a static IP and such | 12:59 |
| vdrok | jroll: yup, hm, neutron allows to have no dhcp? no expert in neutron sorry :) | 13:00 |
| jroll | vdrok: yes, e.g. rackspace cloud doesn't have dhcp at all | 13:00 |
| dtantsur | jroll, yeah.. we have an option to query the database directly, though it poses a risk if we ever upgrade it | 13:00 |
| vdrok | jroll: yeah. we can strip that bit from attach stuff for easy backporting | 13:00 |
| *** hoangcx has joined #openstack-ironic | 13:00 | |
| sambetts | it'll be ok as long as cloud-init treats unbound as phy but I'm not sure that it dpes | 13:00 |
| *** links has quit IRC | 13:01 | |
| vdrok | sambetts: no, it fails in case of unknown type, ValueError or somethingh | 13:01 |
| jroll | dtantsur: not really, I need to load the hardware types to get supported interfaces, and this also means API configuration needs to include all possible hardware types and defaults and such | 13:01 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Deprecate setting IPMI credentials https://review.openstack.org/417041 | 13:01 |
| dtantsur | jroll, right, yeah.. | 13:01 |
| dtantsur | jroll, wait, don't we put only supported types per hardware type to this database? | 13:02 |
| jroll | oh, right | 13:02 |
| jroll | so the config is the only issue (for defaults) | 13:02 |
| dtantsur | I wonder if we can just store the default | 13:03 |
| jroll | we could add a default column... yeah | 13:03 |
| jroll | FUN | 13:03 |
| jroll | I guess that's the right thing to do | 13:05 |
| sambetts | vdrok: a more backportable version https://review.openstack.org/#/c/421261/ | 13:05 |
| patchbot | patch 421261 - nova (stable/newton) - Ensure we mark baremetal links as phy links | 13:05 |
| jroll | dtantsur: I'll go ahead with that, then, and deal with all the comments in the stack today | 13:05 |
| dtantsur | cool! | 13:06 |
| jroll | I can do the spec update too (ugh) | 13:06 |
| dtantsur | if you have time, sure | 13:06 |
| * dtantsur is dragged into all directions currently | 13:06 | |
| vdrok | sambetts: I can strip the bits from the attach patch in nova and place it in the very beginning of the chain | 13:06 |
| vdrok | and then backport it everywhere properly | 13:06 |
| jroll | dtantsur: I'm surprisingly not too busy this week, though I have a string of meetings this afternoon, so we'll see if I get it done today | 13:06 |
| jroll | tomorrow at the latest | 13:07 |
| vdrok | sambetts: as this is the preferred way I think, instead of proposing code straight to stable :) | 13:07 |
| * jroll bbiab | 13:07 | |
| sambetts | vdrok: I've created it as a backport patch now, the change-id is all updated etc | 13:07 |
| sambetts | and I cherrypick the code on the code from the other patches and made it nicer for covering the case of stable/newton | 13:08 |
| vdrok | sambetts: ah. You need to have cherry-picked from too in the commit message I guess, along with unittest | 13:08 |
| *** amoralej is now known as amoralej|lunch | 13:09 | |
| *** jheroux has joined #openstack-ironic | 13:10 | |
| vdrok | sambetts: also, do you think there is a need to do list ports, instead of just setting all links as phy? | 13:10 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Do not fail in Inspector.__init__ if [inspector]enabled is False https://review.openstack.org/416232 | 13:11 |
| sambetts | vdrok: we could but I think it safer if we ensure that we only touch the link we know are our Ironic ports | 13:12 |
| *** bfournie has quit IRC | 13:12 | |
| *** dtantsur is now known as dtantsur|brb | 13:16 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Change tenant to project https://review.openstack.org/420055 | 13:17 |
| *** lucas-hungry is now known as lucasagomes | 13:21 | |
| *** rloo has joined #openstack-ironic | 13:22 | |
| *** rloo has quit IRC | 13:22 | |
| *** aparnav has quit IRC | 13:24 | |
| *** milan has joined #openstack-ironic | 13:29 | |
| milan | morning Ironic! :) | 13:31 |
| *** ralonsoh_ has joined #openstack-ironic | 13:33 | |
| *** ralonsoh has quit IRC | 13:33 | |
| lucasagomes | milan, morning | 13:34 |
| milan | morning lucasagomes! :) | 13:34 |
| *** vmud213 has quit IRC | 13:36 | |
| vdrok | morning milan | 13:40 |
| milan | morning vdrok! :) | 13:40 |
| *** ccamacho is now known as ccamacho|lunch | 13:40 | |
| *** bfournie has joined #openstack-ironic | 13:46 | |
| openstackgerrit | Noam Angel proposed openstack/ironic: make sure OVS_PHYSICAL_BRIDGE is up before bring up vlan interface https://review.openstack.org/421279 | 13:48 |
| openstackgerrit | Noam Angel proposed openstack/ironic: make sure OVS_PHYSICAL_BRIDGE is up before bring up vlan interface https://review.openstack.org/421279 | 13:53 |
| openstackgerrit | Merged openstack/bifrost: Allow user to specify private SSH key https://review.openstack.org/420210 | 13:59 |
| *** baoli has joined #openstack-ironic | 13:59 | |
| *** ccamacho|lunch is now known as ccamacho | 13:59 | |
| *** krtaylor has joined #openstack-ironic | 14:00 | |
| *** baoli has quit IRC | 14:00 | |
| *** baoli has joined #openstack-ironic | 14:01 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add usage documentation for Baremetal OSC Plugin https://review.openstack.org/415261 | 14:06 |
| *** Goneri has joined #openstack-ironic | 14:08 | |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Find a node by multiple attributes https://review.openstack.org/421285 | 14:09 |
| vdrok | sambetts: there is one problem with approach you're suggesting | 14:09 |
| vdrok | wrt portgroups metadata | 14:09 |
| *** aparnav has joined #openstack-ironic | 14:09 | |
| vdrok | we can have portgroup.address to be None | 14:10 |
| *** dtantsur|brb is now known as dtantsur | 14:10 | |
| vdrok | so we can not always map vif to ironic portgroup by address | 14:10 |
| sambetts | vdrok: when you vif_attach, the portgroup address is (should be) set to the Neutron ports address | 14:11 |
| vdrok | sambetts: if it is present, yes | 14:11 |
| sambetts | then there is no way for the address to be none | 14:11 |
| sambetts | its either set by the admin, or set by vif_attach | 14:11 |
| vdrok | sambetts: no, I mean it is None on ironic side | 14:11 |
| sambetts | and I said when you vif_attach, the portgroup address is (should be) set to the Neutron ports address | 14:12 |
| *** amoralej|lunch is now known as amoralej | 14:13 | |
| sambetts | I believe thats what we decided was going to be what happened, or perhaps I misunderstood :/ | 14:13 |
| *** jerrygb has joined #openstack-ironic | 14:14 | |
| vdrok | sambetts: yeah, it was I guess. lemme check this | 14:14 |
| *** aparnav has quit IRC | 14:14 | |
| milan | aarefiev, dtantsur, sambetts: multi-matching in inspector: https://review.openstack.org/#/c/421285/1 ; please, let me know wdyt | 14:15 |
| patchbot | patch 421285 - ironic-inspector - Find a node by multiple attributes | 14:15 |
| dtantsur | cool | 14:15 |
| *** [1]cdearborn has joined #openstack-ironic | 14:15 | |
| sambetts | milan: cool I'll add it to my list :) | 14:16 |
| milan | dtantsur, sambetts txh ;) | 14:16 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Put logic in place for using new Ironic network simulation https://review.openstack.org/421288 | 14:17 |
| milan | dtantsur, I think it might solve the port issue too by not failing because of multiple inserts of the same address (should anyone use a non-local IP for the virtualbmc) | 14:17 |
| dtantsur | milan, yep, I guess so. note, however (and make sure to reflect in the release note) that it will prevent us from detecting duplicate IPMI addresses and things like that at early stages | 14:18 |
| milan | speaking about release notes :D I'm missing one | 14:19 |
| milan | sambetts, your w/r your patch, could you please fwd a link to the ironic update? | 14:19 |
| * milan 'd like to check | 14:19 | |
| sambetts | milan: I've dropped it in a comment on the patch :) | 14:20 |
| milan | sambetts, cool thanks! :) | 14:21 |
| vdrok | sambetts: so, in fact I have not added it to that attach to portgroup patch. vsaienk0 has a concern with it | 14:22 |
| vdrok | sambetts: if we get it from neutron mac_address and set in ironic | 14:22 |
| vdrok | then delete neutron port, address still set on ironic port | 14:22 |
| vdrok | then that address can be used for another neutron port | 14:22 |
| sambetts | vif_detach should handle it some how, we might need to place a flag in the internal_info so we know to clean it up | 14:23 |
| vdrok | sambetts: how are we going to distinguish user-set address and neutron-set address then? | 14:23 |
| *** vsaienko has joined #openstack-ironic | 14:23 | |
| vdrok | ah, flag | 14:23 |
| sambetts | vif_attach sets neutron_mac=True and the vif_detach deletes it | 14:24 |
| vdrok | sambetts: that feels really complex now | 14:24 |
| *** rloo has joined #openstack-ironic | 14:25 | |
| openstackgerrit | Andrey Shestakov proposed openstack/bifrost: Fixes to allow install Ironic on remote node https://review.openstack.org/409213 | 14:25 |
| vdrok | jroll: wdyt on ^^? | 14:26 |
| TheJulia | Good morning | 14:26 |
| vdrok | morning TheJulia | 14:26 |
| sambetts | Morning TheJulia | 14:26 |
| rloo | good morning TheJulia, vdrok, sambetts, jroll | 14:27 |
| vdrok | morning rloo | 14:27 |
| milan | morning rloo! :) | 14:28 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Functional tests for port groups in ironicclient https://review.openstack.org/297727 | 14:28 |
| rloo | hi milan! | 14:28 |
| *** jcoufal has joined #openstack-ironic | 14:29 | |
| milan | and morning jroll, sambetts and TheJulia! :) I think I've not said hi today yet ;) | 14:29 |
| TheJulia | Good morning rloo, milan | 14:29 |
| * TheJulia tries to wake up with more coffee | 14:30 | |
| *** cdearborn has quit IRC | 14:30 | |
| milan | sambetts, qq shouldn't the inspector part Depend-on: the ironic patches? | 14:30 |
| rloo | vdrok: you're going to update this? https://review.openstack.org/#/c/388756/14/nova/virt/ironic/driver.py | 14:30 |
| patchbot | patch 388756 - nova - Generate necessary network metadata for ironic por... | 14:30 |
| milan | maybe just to track that | 14:30 |
| dtantsur | morning rloo, TheJulia | 14:30 |
| milan | more coffee \o/ | 14:31 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add RPC and object version pinning https://review.openstack.org/407491 | 14:31 |
| rloo | hi dtantsur | 14:31 |
| sambetts | milan: Unfortunatly because inspector gates on my ironic changes, inspector needs to be able to deal with my ironic changes before they merge | 14:31 |
| milan | sambetts, I'm reading that sentence 3rd time :D | 14:32 |
| milan | OK | 14:32 |
| rloo | vdrok: cuz wondering if the dependent (attach/detach) patch also will be updated, it uses .internal_info too | 14:32 |
| vdrok | rloo: working on that. yet another issue in ironic code as it seems | 14:32 |
| vdrok | rloo: yup, in case of ports it can be mitigated | 14:32 |
| rloo | vdrok: ouch. better now than later though. | 14:32 |
| vdrok | in case of portgroups, not | 14:32 |
| vdrok | sambetts: wdyt about adding api to get port/portgroup by vif? | 14:33 |
| TheJulia | good morning dtantsur | 14:33 |
| vdrok | we'll have to do multiple requests tho while metadata generation :( | 14:34 |
| vdrok | anyway, the flag seems to be the only quick way forward | 14:35 |
| *** aparnav has joined #openstack-ironic | 14:36 | |
| *** gcb has joined #openstack-ironic | 14:36 | |
| rloo | vdrok: I'm fine with a #TODO | 14:37 |
| sambetts | vdrok: I avoided that because in the case of trunked vlans there would be multiple vifs per port/portgroup. /me wonders out loud if we should be using the neutron port binding_profile to store which Ironic port the neutron port is assigned to instead of storing it in Ironic :/ sigh... openstack networking is a complex thing... | 14:38 |
| *** sasha2 has joined #openstack-ironic | 14:39 | |
| *** sasha21 has quit IRC | 14:40 | |
| vsaienk0 | sambetts: neutron has trunk port, in case of attaching trunk to ironic node we still have only one VIF - vif of trunked port | 14:40 |
| sambetts | vsaienk0: yeah... thats a good point, I keep forgetting the data strcutures | 14:41 |
| vsaienk0 | all network abstraction (trunk details) should be stored on newtron side | 14:41 |
| *** bfournie1 has joined #openstack-ironic | 14:41 | |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add version column https://review.openstack.org/412397 | 14:41 |
| *** bfournie has quit IRC | 14:42 | |
| sambetts | maybe vif_list should return vif: port(group)_id :/ | 14:42 |
| vsaienk0 | sambetts: so for now if we know after vif_attach to which excatly ironic port we mapped a vif, might be add a new API to allow get ironic port|portgroup uuid where specific VIF is attached? | 14:42 |
| *** aparnav has quit IRC | 14:43 | |
| vsaienk0 | sambetts: yeah, that is what I'm thinking about | 14:43 |
| *** hoangcx has quit IRC | 14:44 | |
| *** ralonsoh_ is now known as ralonsoh | 14:45 | |
| vdrok | sambetts: vsaienk0. ok, let's wait what jroll thinks about that. We'll need another api micorversion and client release if we want that way. and flag if we're short on time | 14:45 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add version column https://review.openstack.org/412397 | 14:47 |
| sambetts | +1 | 14:47 |
| *** hoangcx has joined #openstack-ironic | 14:47 | |
| rloo | vdrok, sambetts: can we code it in nova with what we have now? I'm hesitant to rush through a new REST API/client unless it is urgent. | 14:49 |
| mariojv | thanks xek, i'll re-review today | 14:49 |
| sambetts | rloo: I think its possibe, we'd need to make an update to the vif_attach and vif_detach logic on the Ironic side, (not API just how it processes the request) and then all ironic mac address should (if attached to a vif) match the mac addresses in Neutron | 14:50 |
| TheJulia | lucasagomes: w/r/t cinder, do you think we should just completely remove the option and leave a note indicating an open bug? | 14:50 |
| sambetts | rloo: it'll just require more requests to Ironic than just one request to get the vif_list | 14:50 |
| lucasagomes | TheJulia, hi there... I really don't mind, I know it's not pratical because it's not really a "option" if it just accept one value | 14:51 |
| lucasagomes | TheJulia, but, mostly imporatnt I think we should report it in cinder | 14:51 |
| sambetts | milan: working on it ;) | 14:51 |
| lucasagomes | and leave a FIXME note in the code maybe | 14:51 |
| rloo | sambetts: oh, we don't want more requests to ironic. or neutron. | 14:51 |
| vdrok | sambetts: rloo the easiest way tho is to leave this stuff with checking specific fields for now, and update later | 14:51 |
| TheJulia | lucasagomes: writing a bug now, no idea if they will take any action on it | 14:51 |
| vdrok | but yeah, if only ironic side, we can do that pretty quick | 14:52 |
| rloo | vdrok: yes, that's what i mean. if the code as-is works, leave it for now with a TODO | 14:52 |
| lucasagomes | TheJulia, yeah, we at least did our part | 14:52 |
| lucasagomes | TheJulia, it's strange for the API to accept noauth but the client not. Def a bug IMHO | 14:52 |
| TheJulia | lucasagomes: I tried bringing it up with a few cinder folks recently and they didn't really seem interested. Looking at their testing, they just use noauth for api tests. | 14:52 |
| TheJulia | Where as the client has a bunch of logic predicated on keystone :( | 14:53 |
| lucasagomes | :-( | 14:53 |
| lucasagomes | TheJulia, hope someone new might get interested on that problem and fix it | 14:53 |
| sambetts | rloo: oh right, yeah it should work as is, just it means that nova is aware of internals :'( | 14:53 |
| lucasagomes | if the current devs are not interested | 14:53 |
| TheJulia | lucasagomes: Yeah | 14:54 |
| rloo | sambetts: right. so i'm ok with it as long as nova are too. if we put a note/todo, that should be ok? | 14:55 |
| sambetts | I don't have a problem with it :) | 14:55 |
| rloo | see what jroll thinks ^^ | 14:55 |
| rloo | sambetts: what if we don't get the nova patches landed? no support for portgroups :-( ok, i answered my question. | 14:57 |
| *** dtantsur is now known as dtantsur|mtg | 14:57 | |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add new dbsync command with first online data migration https://review.openstack.org/408556 | 14:58 |
| *** udesale has joined #openstack-ironic | 14:59 | |
| *** hoangcx has quit IRC | 15:03 | |
| *** rajinir has joined #openstack-ironic | 15:03 | |
| *** rama_y has joined #openstack-ironic | 15:04 | |
| *** trown is now known as trown|brb | 15:04 | |
| jroll | rloo: vdrok: sambetts: can someone summarize? | 15:04 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add new dbsync command with first online data migration https://review.openstack.org/408556 | 15:05 |
| *** jaosorior has joined #openstack-ironic | 15:06 | |
| vdrok | jroll: portgroup address is not mandatory. to do the thing sambetts suggested in the portgroup metadata change, we need to be able to map neutron vifs to ironic objects by address. to do that, we need to set portgroup address from neutro port during attach call, and remember that it needs to be removed during detach, by adding a flag to internal_info. Or add | 15:06 |
| vdrok | a new api that will output vif_id and port(group)_uuid. or leave as is, with checking the internal_info/extra and change later, when we figure out the best way | 15:06 |
| jroll | vdrok: does the "leave as is" option mean portgroups without addresses cannot be used? | 15:07 |
| vdrok | jroll: nope, they will be used. it just means that nova is still aware about fields vif is stored in | 15:08 |
| jroll | vdrok: oh, I see | 15:08 |
| jroll | +1 for leaving as is, and fixing next cycle (preferably with vif-list just returning everything we need for network metadata) | 15:09 |
| jroll | which I commented on the nova patch, btw :) | 15:09 |
| sambetts | +1 | 15:09 |
| vdrok | ++, that is what I'd love to have in our API too | 15:09 |
| rloo | +1 (phew) | 15:09 |
| jroll | I'd like GET /nodes/foo/vifs/ to return [{'id': 'xxx', 'port': 'port-uuid', 'properties': {'bond_mode': 'blah', ...}, ...}, ...] | 15:10 |
| vdrok | okie, I'll change the other things pointed out and create an rfe to add the api | 15:10 |
| jroll | or something | 15:10 |
| jroll | then it's one api call to get everything we need and we're all happy | 15:10 |
| rloo | jroll: my only concern is if the nova folks balk at the use of .internal_info[] | 15:10 |
| vdrok | rloo: it's not really internal :) just read-only | 15:11 |
| jroll | rloo: should be fine, if it wasn't meant to be read it wouldn't be in the api | 15:11 |
| jroll | rloo: also, if we have a solid plan to fix later I don't think they will mind | 15:11 |
| vdrok | it's a bad naming that we have | 15:11 |
| rloo | well, it isn't bad naming. it was meant for internal stuff and i know one person that didn't want that exposed. | 15:11 |
| rloo | vdrok: we should add a good NOTE or TODO though :) | 15:12 |
| jroll | ++ | 15:12 |
| vdrok | rloo: sure. | 15:12 |
| jroll | I think that better list vifs api will be super useful when we're doing trunking, btw | 15:13 |
| jroll | is there anything that needs reviewing quickly before I jump into rewriting half of this driver composition code? :P | 15:13 |
| rloo | jroll: the response is diff for the better vif-list. are we ok with that via a version bump? or should we scramble to fix it now or maybe it is too late. | 15:14 |
| vdrok | rloo: so you think we should not output port(group).internal_info and node.driver_internal_info? | 15:14 |
| vdrok | rloo: we can do it as separate endpoint even | 15:14 |
| rloo | vdrok: well, it is internal stuff. should only be useful for debug. | 15:14 |
| jroll | rloo: let's go next cycle with a version bump, I think it will take some time to design it well | 15:14 |
| rloo | jroll: ok with me. | 15:14 |
| jroll | rloo: I thought it might be :) | 15:14 |
| rloo | jroll: and yeah, get that driver composition done, will you? | 15:15 |
| jroll | dtantsur|mtg: oh, I wanted to chat with you about final release timeline when you're free | 15:15 |
| jroll | rloo: on it! | 15:15 |
| rloo | jroll: ^^ was about to ask you if you had chatted with him! | 15:15 |
| *** pcaruana has quit IRC | 15:15 | |
| jroll | :D | 15:15 |
| dtantsur|mtg | jroll, I'll be free in roughly 45 minutes, if it works for you | 15:16 |
| jroll | dtantsur|mtg: yep, that's fine | 15:16 |
| jroll | thanks | 15:16 |
| *** zackf has quit IRC | 15:19 | |
| openstackgerrit | Sam Betts proposed openstack/ironic: Rework Ironic devstack baremetal network simulation https://review.openstack.org/392959 | 15:19 |
| *** trown|brb is now known as trown | 15:20 | |
| *** udesale has quit IRC | 15:21 | |
| *** vsaienko has quit IRC | 15:21 | |
| *** gcb has quit IRC | 15:26 | |
| *** xuao has joined #openstack-ironic | 15:29 | |
| *** xuao has quit IRC | 15:37 | |
| *** david-lyle has joined #openstack-ironic | 15:42 | |
| *** bfournie1 is now known as bfournie | 15:43 | |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Rolling upgrades related dev documentation https://review.openstack.org/419439 | 15:49 |
| *** david-lyle has quit IRC | 15:55 | |
| *** zackf has joined #openstack-ironic | 15:57 | |
| *** ohamada has quit IRC | 16:01 | |
| openstackgerrit | Merged openstack/bifrost: Add keystone to git contents download https://review.openstack.org/407215 | 16:04 |
| *** david-lyle has joined #openstack-ironic | 16:07 | |
| *** absubram has joined #openstack-ironic | 16:09 | |
| *** dtantsur|mtg is now known as dtantsur | 16:10 | |
| dtantsur | jroll, back | 16:10 |
| *** rpioso has joined #openstack-ironic | 16:10 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 16:12 | |
| jroll | dtantsur: hey, so did you see doug's email yesterday? tl;dr they're branching g-r just after cycle-with-milestones are released: http://lists.openstack.org/pipermail/openstack-dev/2017-January/110218.html | 16:13 |
| jroll | which means we'll be running with pike requirements after that | 16:13 |
| jroll | so wondering if we should also do final release and branch then, or risk stable/ocata (or master?) CI being broken after we branch | 16:13 |
| dtantsur | yeah, I've seen it | 16:14 |
| dtantsur | jroll, Feb 2 sounds like a good date to wrap up the feature work to me | 16:14 |
| * jroll looks at calendar | 16:15 | |
| jroll | dtantsur: do we want to branch at that time or give some time to handle bugs? | 16:15 |
| dtantsur | jroll, we can make a release on the same time (give or take) as folks' RC1 and call it stable/ocata | 16:15 |
| *** livelace has quit IRC | 16:15 | |
| jroll | yeah | 16:15 |
| dtantsur | jroll, then we need to reintroduce feature freezes :) | 16:15 |
| jroll | so that was kind of my thought, just means final touches on features have to be done | 16:15 |
| jroll | yeah | 16:15 |
| jroll | :( | 16:15 |
| dtantsur | backporting bug fixes is not too complex fwiw | 16:16 |
| jroll | this is true | 16:16 |
| jroll | it just means no features at all past feb 2 | 16:16 |
| jroll | as opposed to keeping features minimal | 16:16 |
| dtantsur | well.. yeah, so kind of a feature freeze in the end.. | 16:17 |
| jroll | which is fine | 16:17 |
| jroll | this is just not what I expected, I guess | 16:17 |
| jroll | and only 2 weeks notice :| | 16:17 |
| jroll | (3?) | 16:17 |
| dtantsur | as to branching later. I'm not particularly afraid of broken CI, but from packager point of view I don't quite like requirements going back in time when we branch | 16:17 |
| jroll | well, we'd freeze requirements time | 16:18 |
| jroll | er, s/time/bot/ | 16:18 |
| dtantsur | hmm | 16:18 |
| NobodyCam | Good Morning Ironic'ers :) | 16:18 |
| dtantsur | morning NobodyCam | 16:19 |
| jroll | dtantsur: which I guess means CI wouldn't break | 16:19 |
| NobodyCam | hey hey dtantsur morning :) | 16:19 |
| jroll | er, yes it would | 16:19 |
| jroll | constraints | 16:19 |
| dtantsur | jroll, we can't freeze u-c | 16:19 |
| dtantsur | right | 16:19 |
| NobodyCam | morning jroll :) | 16:19 |
| jroll | dtantsur: we could actually point our tox to the branched version (we do that after branching anyway) | 16:19 |
| jroll | hey NobodyCam :) | 16:19 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Rolling upgrades related dev documentation https://review.openstack.org/419439 | 16:19 |
| dtantsur | jroll, aha, and then revert it on master | 16:20 |
| vdrok | morning NobodyCam | 16:20 |
| jroll | yeah | 16:20 |
| NobodyCam | morning vdrok :) | 16:21 |
| dtantsur | jroll, when would you branch then? R-1 or R-2 (I guess R-0 is too much)? | 16:22 |
| *** Nisha_Agarwal has quit IRC | 16:22 | |
| jroll | dtantsur: r-1 would be the deadline | 16:22 |
| jroll | we could do r-2 | 16:22 |
| *** Nisha_Agarwal has joined #openstack-ironic | 16:22 | |
| jroll | depends how code is looking I guess | 16:22 |
| jroll | I'd still want to soft FF aroudn feb 2 | 16:23 |
| dtantsur | jroll, I'd have r-1 as branching point and r-3 as a soft freeze (only priorities, bug fixes and small improvements) | 16:23 |
| dtantsur | hah | 16:23 |
| * dtantsur types too slowly | 16:23 | |
| jroll | rloo: ^ fyi, seems we can mitigate CI problems | 16:23 |
| jroll | :P | 16:23 |
| *** mjura has quit IRC | 16:23 | |
| rloo | jroll: in meeting, will look later but summary sounds good! :) | 16:24 |
| jroll | rloo: cool :) | 16:24 |
| jroll | dtantsur: I'd call that a plan then | 16:24 |
| jroll | thanks | 16:24 |
| *** swatson has left #openstack-ironic | 16:24 | |
| * jroll goes to standup then continues addressing comments on driver composition series | 16:24 | |
| dtantsur | jroll, I suggest to run this idea through openstack-release to avoid misunderstanding | 16:24 |
| jroll | dtantsur: sure, I can do that | 16:25 |
| dtantsur | cool | 16:25 |
| jlvillal | Good morning Ironic | 16:26 |
| *** jaosorior has quit IRC | 16:26 | |
| lucasagomes | jlvillal, morning | 16:26 |
| NobodyCam | good morning jlvillal | 16:26 |
| NobodyCam | morning lucasagomes | 16:26 |
| lucasagomes | NobodyCam, o/ morning | 16:27 |
| NobodyCam | :) | 16:27 |
| jlvillal | :) | 16:27 |
| milan | morning jlvillal, NobodyCam ! :) | 16:27 |
| NobodyCam | hey there Morning milan :) | 16:28 |
| milan | :) | 16:28 |
| *** aparnav has joined #openstack-ironic | 16:28 | |
| rama_y | Hi dtantsur, milan, this is regarding https://review.openstack.org/#/c/406790/ | 16:29 |
| patchbot | patch 406790 - ironic-inspector-specs - Collect Bios, Memory and Processor Info from Dmide... | 16:29 |
| rama_y | Would you be able to give review comments on this? | 16:29 |
| milan | hi rama_y! :) | 16:29 |
| rama_y | hi milan! | 16:29 |
| milan | rama_y, yeah, in my todo, I guess tomorrow | 16:30 |
| rama_y | Thanks, milan! | 16:30 |
| milan | rama_y, no prob :) | 16:30 |
| *** baoli has quit IRC | 16:30 | |
| dtantsur | rama_y, at want point in time, yes. sorry, we're very close to the release and too many things to finish still left | 16:30 |
| *** aparna has joined #openstack-ironic | 16:30 | |
| rama_y | dtantsur, thanks. | 16:32 |
| *** appu103 has joined #openstack-ironic | 16:33 | |
| *** aparnav has quit IRC | 16:34 | |
| *** aparna has quit IRC | 16:36 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic-specs: Intel Node Manager sensor collection support https://review.openstack.org/413249 | 16:36 |
| *** e0ne has quit IRC | 16:37 | |
| *** ijw_ has joined #openstack-ironic | 16:37 | |
| *** Syed__ has joined #openstack-ironic | 16:39 | |
| *** ijw has quit IRC | 16:40 | |
| *** moshele has joined #openstack-ironic | 16:41 | |
| jroll | dtantsur: one thing we missed: "your master branch will still be tested against master of the other branches during that period, which may introduce issues. that may end up causing issues with the constraints, for example" | 16:47 |
| jroll | tested against master services, to be clear | 16:48 |
| jroll | also "I think you understand what issues might come up by delaying, so I'd say go ahead with your plan but keep an eye on things. If it becomes too bad, you can always branch at that point in time, I suppose." | 16:48 |
| jroll | so I guess let's evaluate weekly (daily?) as we get closer | 16:48 |
| dtantsur | jroll, oh right.. yes, I still believe the probability of breakage is not too high. but it's true that we no longer test ocata with ocata, so it's kind of an argument for making the window shorter | 16:50 |
| jroll | dtantsur: yeah, I agree | 16:50 |
| jroll | fwiw I would prefer a sooner branch, but I'd also prefer we finish everything, so :) | 16:50 |
| dtantsur | heh, let's concentrate and try to finish $everything in 2 weeks \o/ | 16:51 |
| jroll | +1 :) | 16:51 |
| * jroll gets back to work | 16:51 | |
| * dtantsur dives into reviews for the night is dark and full of terrors.. | 16:53 | |
| jroll | what is this black magic https://github.com/openstack/oslo.db/blob/c61d2d47bd39c31f1cd06faf497ffa85ae66cdd4/oslo_db/sqlalchemy/exc_filters.py#L96 | 16:57 |
| * jroll sighs | 16:57 | |
| sambetts | lol | 16:57 |
| openstackgerrit | Merged openstack/ironic-python-agent: Removes unnecessary utf-8 encoding https://review.openstack.org/413866 | 17:03 |
| dtantsur | btw a handy tip for gertty users: you can prepare your own priority dashboard by something like | 17:05 |
| dtantsur | - name: "Priority" | 17:05 |
| dtantsur | query: "status:open AND (is:starred OR topic:bug/1559691 OR topic:bug/1526283 OR topic:bug/1526226)" | 17:05 |
| dtantsur | key: "f3" | 17:05 |
| dtantsur | put in ~/.gertty.yaml | 17:05 |
| dtantsur | (this is not all priorities, so copy-paste with care :) | 17:06 |
| *** jaosorior has joined #openstack-ironic | 17:07 | |
| *** jrist has quit IRC | 17:08 | |
| jroll | heh, nice, thanks | 17:13 |
| *** jrist has joined #openstack-ironic | 17:13 | |
| openstackgerrit | Sam Betts proposed openstack/ironic: Rework Ironic devstack baremetal network simulation https://review.openstack.org/392959 | 17:14 |
| *** amarao has quit IRC | 17:15 | |
| * sambetts probably needs to start using gertty ... it just takes sooooo long to sync the first time lol | 17:15 | |
| *** amarao has joined #openstack-ironic | 17:15 | |
| *** trown is now known as trown|lunch | 17:15 | |
| TheJulia | sambetts: The last couple times I've fired it up, it went fairly quick, at least for what I cared about :) | 17:17 |
| sambetts | it might have imporved quite a bit since I last used it, I think I tried to set it up for the vancover summit (I think) | 17:17 |
| TheJulia | I think it did get some improvements after that because I seem to remember setting it up about three months after that and it was faster | 17:19 |
| *** ijw has joined #openstack-ironic | 17:19 | |
| dtantsur | yep, it's definitely improved since vancouver. it was really horribly slow back then | 17:19 |
| dtantsur | it's sometimes horribly slow nowadays, but usually it's quite fine | 17:20 |
| sambetts | awesome :) | 17:20 |
| *** srobert has joined #openstack-ironic | 17:22 | |
| *** rama_y has quit IRC | 17:22 | |
| *** ijw_ has quit IRC | 17:22 | |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: [WIP] Put logic in place for using new Ironic network simulation https://review.openstack.org/421288 | 17:24 |
| *** vinsh has joined #openstack-ironic | 17:30 | |
| *** sacharya has quit IRC | 17:33 | |
| *** sacharya has joined #openstack-ironic | 17:33 | |
| *** athomas has quit IRC | 17:35 | |
| *** moshele has quit IRC | 17:36 | |
| *** causten has joined #openstack-ironic | 17:36 | |
| *** baoli has joined #openstack-ironic | 17:37 | |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add RPC and object version pinning https://review.openstack.org/407491 | 17:40 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add version column https://review.openstack.org/412397 | 17:42 |
| rloo | jroll, dtantsur: I read ^^ about releasing and am not sure I understand. Are we talking about (soft) feature freeze week of Jan 29 and branching stable/ocata week of feb 13? | 17:43 |
| dtantsur | rloo, yep, seems like that | 17:44 |
| jroll | rloo: yes, preferably branching before, if we can get things done earlier | 17:44 |
| rloo | i think we should branch at the latest the week before. but just my thinking. i don't like waiting til the 'last' minute. | 17:45 |
| rloo | i mean branch by week of Feb 6. | 17:45 |
| TheJulia | rloo: ++ | 17:46 |
| jroll | don't disagree, but we've been doing it at the deadline for the last few cycles | 17:46 |
| rloo | jroll: oh, we have? i thought we didn't wait til the last week but i don't keep track of that. if so, then let's continue with it. | 17:47 |
| *** livelace has joined #openstack-ironic | 17:48 | |
| jroll | idk, I seem to remember we did, I could be wrong | 17:48 |
| rloo | jroll: i think it is worth sending email about the soft feature freeze date. we talked a bit about it in Monday's meeting but we didn't mention the feature freeze date. it is in 2ish weeks. | 17:48 |
| *** derekh has quit IRC | 17:49 | |
| jroll | rloo: yes, I agree | 17:50 |
| jroll | I kind of want to play things by ear | 17:50 |
| jroll | but I will send one to that effect | 17:50 |
| rloo | jroll: thx. i don't want people coming to us later and saying they didn't know, blah blah. | 17:51 |
| jroll | meh, if they didn't know how we work they would assume we follow standard feature freeze, so I tend not to care | 17:51 |
| jroll | s/would/should | 17:52 |
| *** lucasagomes is now known as lucas-afk | 17:53 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Update troubleshooting docs on no valid host found error https://review.openstack.org/419595 | 17:53 |
| *** rama_y has joined #openstack-ironic | 17:55 | |
| *** absubram has quit IRC | 17:55 | |
| dtantsur | see you tomorrow | 17:56 |
| *** dtantsur is now known as dtantsur|afk | 17:56 | |
| NobodyCam | night dtantsur|afk | 17:56 |
| jroll | later dtantsur|afk | 17:56 |
| sambetts | o/ dtantsur|afk | 17:56 |
| *** mgoddard has quit IRC | 18:02 | |
| *** ralonsoh has quit IRC | 18:02 | |
| *** absubram has joined #openstack-ironic | 18:07 | |
| *** causten has quit IRC | 18:07 | |
| *** livelace has quit IRC | 18:09 | |
| *** causten has joined #openstack-ironic | 18:10 | |
| *** livelace has joined #openstack-ironic | 18:10 | |
| *** Sukhdev has joined #openstack-ironic | 18:10 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Update troubleshooting docs on no valid host found error https://review.openstack.org/419595 | 18:18 |
| NobodyCam | lol dtantsur|afk so much for afk :p | 18:19 |
| *** jcoufal has quit IRC | 18:23 | |
| *** dprince has quit IRC | 18:28 | |
| *** ijw has quit IRC | 18:28 | |
| *** jcoufal has joined #openstack-ironic | 18:32 | |
| *** milan has quit IRC | 18:34 | |
| *** krtaylor has quit IRC | 18:36 | |
| jlvillal | JayF: Easy review for you :) https://review.openstack.org/#/c/420789/ | 18:38 |
| patchbot | patch 420789 - ironic (stable/newton) - Trivial: fix warning message formatting | 18:38 |
| vdrok | good morning jlvillal and good night dtantsur|afk ! | 18:43 |
| jlvillal | :) | 18:43 |
| *** trown|lunch is now known as trown | 18:43 | |
| *** mgoddard has joined #openstack-ironic | 18:46 | |
| *** david-lyle has quit IRC | 18:47 | |
| *** ijw has joined #openstack-ironic | 18:50 | |
| *** nicodemos has quit IRC | 18:54 | |
| *** aNuposic has joined #openstack-ironic | 18:54 | |
| *** ijw has quit IRC | 18:54 | |
| *** nicodemos has joined #openstack-ironic | 18:55 | |
| *** dprince has joined #openstack-ironic | 18:56 | |
| *** amoralej is now known as amoralej|off | 19:01 | |
| aNuposic | Hi Folks, I am trying to use already defined auth conf options like 'auth_url' and 'username' in one of the unit test, are there any pointers on how to use that? As right now I see those auth options are not registered and so unit test failing with "no such option in group". | 19:02 |
| aNuposic | JayF: mariojv: jlvillal: rloo: ^ | 19:03 |
| JayF | CONF.set_override iirc | 19:03 |
| JayF | is how you set those to temporary values for tests | 19:03 |
| *** jkilpatr has quit IRC | 19:04 | |
| openstackgerrit | Noam Angel proposed openstack/ironic: make sure OVS_PHYSICAL_BRIDGE is up before bring up vlan interface https://review.openstack.org/421279 | 19:05 |
| jroll | dtantsur|afk: I feel like you and I need this shirt: https://teespring.com/shop/maksesoftwarebetter2#pid=6&cid=647&sid=front | 19:07 |
| jroll | JayF: aNuposic: self.config(auth_url='foo') | 19:08 |
| jroll | and you can add a group='keystone' or whatever to that | 19:08 |
| jroll | with that error you may just have the wrong group | 19:08 |
| *** rbartal has quit IRC | 19:12 | |
| aNuposic | jroll: JayF: I am using set_override right now and group seems to be correct, I am using 'swift' group. Let me try self.config | 19:12 |
| *** jkilpatr has joined #openstack-ironic | 19:18 | |
| TheJulia | aNuposic: You may want to double check that the parameter is actually used. Some are imported via templates and may not actually be used in certain groups. | 19:20 |
| *** vsaienko has joined #openstack-ironic | 19:24 | |
| aNuposic | TheJulia: So you think it should be defined again? I tried to define them in ironic/conf/swift.py but it throws error "DuplicateOptError" | 19:24 |
| * rloo wonders what kind of yaks jroll and dtantsur|afk were shaving... | 19:24 | |
| aNuposic | TheJulia: JayF: jroll: this is what I tried and getting the error http://paste.openstack.org/show/595263/ | 19:24 |
| TheJulia | aNuposic: That would cause it I believe. That imports ?auth? I think. | 19:25 |
| jroll | hrm, no clue | 19:27 |
| jroll | rloo: oh, you'll see in a moment | 19:27 |
| *** vsaienko has quit IRC | 19:28 | |
| aNuposic | TheJulia: so do you think I shall use ConfigOpts.import_opt() method? I am not clear on what should I import to use options imported via templates if auth is getting imported like that | 19:28 |
| TheJulia | aNuposic: Give me a minute, looking | 19:29 |
| * jroll is getting lost in driver composition | 19:31 | |
| aNuposic | TheJulia: Sure. | 19:31 |
| *** Nisha_Agarwal has quit IRC | 19:32 | |
| *** mgoddard has quit IRC | 19:35 | |
| TheJulia | aNuposic: https://github.com/openstack/ironic/blob/5603a215b726ca465eed2867d830821bb24b1a72/releasenotes/notes/keystone-auth-3155762c524e44df.yaml | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: WIP: API changes for driver-composition https://review.openstack.org/417970 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Enable manual-management hardware type in devstack https://review.openstack.org/417920 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Expose default interface calculation from driver_factory https://review.openstack.org/418573 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add hardware types to the hash ring https://review.openstack.org/419065 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Register/unregister hardware interfaces for conductors https://review.openstack.org/412631 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add default column to ConductorHardwareInterfaces https://review.openstack.org/421464 | 19:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Refactor active conductor check https://review.openstack.org/421465 | 19:36 |
| jroll | jeeeeeeeeebus. | 19:36 |
| jroll | I think that only involved 10 rebase -i commands | 19:37 |
| openstackgerrit | Merged openstack/ironic: Use TENANT_VIF_KEY constant everywhere https://review.openstack.org/421121 | 19:37 |
| *** ChubYann has joined #openstack-ironic | 19:37 | |
| *** appu103 has quit IRC | 19:39 | |
| openstackgerrit | Merged openstack/ironic: Update troubleshooting docs on no valid host found error https://review.openstack.org/419595 | 19:39 |
| *** vsaienko has joined #openstack-ironic | 19:40 | |
| TheJulia | aNuposic: checkout https://github.com/openstack/ironic/blob/5603a215b726ca465eed2867d830821bb24b1a72/ironic/tests/unit/common/test_keystone.py#L29 | 19:41 |
| aNuposic | TheJulia: Okay. I checked this file, but then do I need similar setup() for swift? | 19:44 |
| TheJulia | For your test since, at least as I understand it, the options should be loaded out from keystoneauth, which might be why your getting the error trying to assert auth_url in the swift group now. | 19:45 |
| aNuposic | TheJulia: well it seems it loads 'password' plugin, do I need to load password plugin in my test file? | 19:45 |
| aNuposic | TheJulia: do you think importing this setup() from keystoneauth in my test file will help? | 19:49 |
| *** xavierr has joined #openstack-ironic | 19:50 | |
| TheJulia | aNuposic: I wouldn't be able to tell you for sure if it would help to load the plugin without having an understanding of why your directly attempting to leverage or test for the auth url in a test when the client should be using a keystone session | 19:50 |
| *** david-lyle has joined #openstack-ironic | 19:50 | |
| *** rpioso has quit IRC | 19:53 | |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Agent driver rescue implementation https://review.openstack.org/400437 | 19:54 |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add API and conductor methods for [un]rescue https://review.openstack.org/350831 | 19:54 |
| aNuposic | TheJulia: Working on patch for RFE https://bugs.launchpad.net/ironic/+bug/1642719 where I am using Ceph RADOSGW authentication instead of keystone and hence need auth_url,username and password | 19:55 |
| openstack | Launchpad bug 1642719 in Ironic "[RFE] RADOSGW authentication mechanism for storing config drive in Ceph" [Wishlist,In progress] - Assigned to Anup (anup-d-navare) | 19:55 |
| aNuposic | TheJulia: Here is the patch https://review.openstack.org/#/c/407639/ | 19:56 |
| patchbot | patch 407639 - ironic - Config drive support for Ceph | 19:56 |
| TheJulia | aNuposic: I _think_ you'll need to create a new configuration section explicitly for collecting those parameters since keystoneauth wouldn't be helping manage those parameters | 19:57 |
| TheJulia | aNuposic: Yeah, you'll need to create new config group for talking to ceph directly | 19:58 |
| aNuposic | TheJulia: Ah, so the new config group would be a separate group in ironic.conf? And thus will have a new .py file where I will define these options, is that what you mean? | 20:00 |
| *** vsaienko has quit IRC | 20:03 | |
| TheJulia | aNuposic: Correct, from there you'd likely want to also put radosgw related actions/params in it's own common module and just use the config parameter in swift.py to determine which as to where it should be routed. But I also don't know the radosgw mechanics | 20:03 |
| *** sambetts is now known as sambetts|afk | 20:03 | |
| vdrok | good night everyone! | 20:03 |
| *** rama_y has quit IRC | 20:03 | |
| *** Sukhdev_ has joined #openstack-ironic | 20:04 | |
| NobodyCam | night vdrok | 20:06 |
| aNuposic | TheJulia: I see. I think so I will create a separate group and read values from there. But I wonder, is there no way to use these defined auth values somehow because if we need auth_url, username, pass for glance later for something then again there has to be a different section | 20:07 |
| TheJulia | Not really, the clients call auth_plugin.get_url() and the session handling logic (if memory serves) attempts to validate keystone tokens, so they are extremely keystone centric. | 20:09 |
| TheJulia | Unless, radosgw speaks Keystone that is... | 20:09 |
| TheJulia | s/speaks/groks/ | 20:09 |
| *** e0ne has joined #openstack-ironic | 20:10 | |
| *** rpioso has joined #openstack-ironic | 20:11 | |
| aNuposic | TheJulia: ah yeah. That's true every service is mostly authenticated using keystone. Cool. Thanks a ton! | 20:15 |
| TheJulia | aNuposic: No problem, sorry for the earlier confusion :) | 20:16 |
| aNuposic | TheJulia: Not at all :) | 20:16 |
| *** Sukhdev has quit IRC | 20:21 | |
| *** absubram has quit IRC | 20:23 | |
| *** krtaylor has joined #openstack-ironic | 20:26 | |
| *** formica has joined #openstack-ironic | 20:27 | |
| *** mtanino has joined #openstack-ironic | 20:28 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-ui: Updated from global requirements https://review.openstack.org/420396 | 20:29 |
| *** vsaienko has joined #openstack-ironic | 20:30 | |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Find a node by multiple attributes https://review.openstack.org/421285 | 20:33 |
| openstackgerrit | Merged openstack/ironic: Do not fail in Inspector.__init__ if [inspector]enabled is False https://review.openstack.org/416232 | 20:44 |
| *** vsaienko has quit IRC | 20:50 | |
| *** absubram has joined #openstack-ironic | 20:54 | |
| *** vsaienko has joined #openstack-ironic | 20:54 | |
| *** sacharya_ has joined #openstack-ironic | 20:57 | |
| *** athomas has joined #openstack-ironic | 20:57 | |
| *** sacharya has quit IRC | 20:59 | |
| *** ijw_ has joined #openstack-ironic | 21:07 | |
| *** absubram has quit IRC | 21:11 | |
| *** ijw_ has quit IRC | 21:12 | |
| *** nicodemos has quit IRC | 21:16 | |
| *** ijw has joined #openstack-ironic | 21:17 | |
| *** nicodemos has joined #openstack-ironic | 21:17 | |
| *** formica1 has joined #openstack-ironic | 21:23 | |
| *** dyasny has quit IRC | 21:31 | |
| *** vsaienko has quit IRC | 21:34 | |
| *** swatson has joined #openstack-ironic | 21:34 | |
| *** vsaienko has joined #openstack-ironic | 21:38 | |
| *** trown is now known as trown|outtypewww | 21:41 | |
| *** jkilpatr has quit IRC | 21:43 | |
| *** vsaienko has quit IRC | 21:43 | |
| *** vsaienko has joined #openstack-ironic | 21:43 | |
| *** dyasny has joined #openstack-ironic | 21:46 | |
| *** e0ne has quit IRC | 21:49 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: Allow to specify VIF fields and interface_type https://review.openstack.org/421544 | 21:50 |
| *** srobert_ has joined #openstack-ironic | 21:51 | |
| *** bfournie has quit IRC | 21:54 | |
| *** srobert has quit IRC | 21:55 | |
| swatson | Is anyone in the channel familiar with the migration tests that ironic runs on the gate? I'm trying to find what step of the process creates the openstack_citest user, if any | 21:55 |
| *** srobert_ has quit IRC | 21:56 | |
| *** d0ugal has quit IRC | 21:59 | |
| *** d0ugal has joined #openstack-ironic | 21:59 | |
| *** d0ugal has quit IRC | 21:59 | |
| *** d0ugal has joined #openstack-ironic | 21:59 | |
| *** krtaylor has quit IRC | 21:59 | |
| *** harlowja has quit IRC | 22:01 | |
| *** sacharya_ has quit IRC | 22:06 | |
| *** sacharya has joined #openstack-ironic | 22:06 | |
| *** jkilpatr has joined #openstack-ironic | 22:07 | |
| *** rama_y has joined #openstack-ironic | 22:07 | |
| *** rama_y_ has joined #openstack-ironic | 22:07 | |
| *** jheroux has quit IRC | 22:10 | |
| *** priteau has quit IRC | 22:13 | |
| *** Goneri has quit IRC | 22:16 | |
| *** vsaienko has quit IRC | 22:18 | |
| *** formica1 has quit IRC | 22:22 | |
| *** wrouesnel has quit IRC | 22:27 | |
| TheJulia | swatson: I just replied in infra with more info | 22:29 |
| swatson | TheJulia: I'm checking it out, thanks | 22:29 |
| *** wrouesnel has joined #openstack-ironic | 22:31 | |
| *** bfournie has joined #openstack-ironic | 22:32 | |
| *** baoli has quit IRC | 22:32 | |
| *** jcoufal has quit IRC | 22:35 | |
| JayF | KVM: entry failed, hardware error 0x0 | 22:42 |
| JayF | were caused by enabling nested kvm, right? | 22:42 |
| JayF | If so, should we consider disabling it again? I see a lot of patches with at least one failure on those as I go through | 22:42 |
| JayF | and it's affecting job stability considerably | 22:42 |
| mariojv | rloo: vdrok: hey, i made a suggestion on the notification payload test cases: https://review.openstack.org/#/c/418148/ | 22:45 |
| patchbot | patch 418148 - ironic - Add tests for Payloads with SCHEMAs | 22:45 |
| mariojv | i think it's probably fine assuming that *Payload classes are going to be notification payloads | 22:46 |
| rloo | mariojv: just looking at it now. am thinking... :) | 22:46 |
| mariojv | but there's a way to avoid just looking at the class name, in case another revision happens | 22:46 |
| TheJulia | JayF: I'd +2 reverting nested kvm | 22:47 |
| *** wrouesnel has quit IRC | 22:47 | |
| rloo | TheJulia, JayF, I'm for it too if it is happening frequently. I see someone added more failure cases yest, to our etherpad | 22:49 |
| JayF | I'll poke at logstash some and make a case along with a nested kvm disable patch | 22:50 |
| JayF | for us to talk about in | 22:50 |
| TheJulia | Yeah, I spotted a few more recently | 22:50 |
| TheJulia | JayF: afaik, the logs it is reported in don't get indexed in ES | 22:50 |
| JayF | oh, that's really unfortunate | 22:50 |
| TheJulia | I've been kind of talking to some folks about maybe making something a little smarter than ES that can hunt for oddities/differences between success/failed jobs, but it is just conceptual vaporware right now | 22:51 |
| JayF | I can at least take a job or two | 22:51 |
| JayF | and try to index it with failures over time | 22:51 |
| *** david-lyle has quit IRC | 22:51 | |
| JayF | to see if I can show the corrolation | 22:52 |
| TheJulia | connect to the AFS cluster and look for node-0.log that is greater than thirty six bytes or something like that | 22:53 |
| TheJulia | ? | 22:53 |
| TheJulia | actually that is not a bad idea... | 22:55 |
| JayF | I nkow what those words mean | 22:56 |
| JayF | but had no idea we could do bulk querying of logs via afs | 22:56 |
| rloo | exit | 22:56 |
| NobodyCam | rloo: lol | 22:56 |
| rloo | :) | 22:56 |
| rloo | ciao | 22:56 |
| *** rloo has quit IRC | 22:57 | |
| *** dsneddon is now known as dsneddon_afk | 22:59 | |
| TheJulia | JayF: well... find since normally the log is nearly empty if not zero bytes, when the jobs fail they the files can be several kilobytes | 22:59 |
| JayF | TheJulia: you mean the qemu log, not the node-0 logs? | 23:00 |
| * TheJulia switches brain gears | 23:00 | |
| TheJulia | JayF: http://logs.openstack.org/35/420835/4/check/gate-tempest-dsvm-ironic-ipa-partition-agent_ipmitool-tinyipa-ubuntu-xenial/7fadba8/logs/libvirt/qemu/node-0.txt.gz | 23:01 |
| JayF | yeah, the qemu logs | 23:01 |
| JayF | I literally had one of those up | 23:01 |
| JayF | ironic-bm-logs/node-0 = blank on fail | 23:01 |
| JayF | libvirt/qemu/node-0 = smaller on fail | 23:02 |
| JayF | right? | 23:02 |
| TheJulia | so I think a normal run's qemu log is smaller than when it is failing right now | 23:03 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Add tests for Payloads with SCHEMAs https://review.openstack.org/418148 | 23:04 |
| TheJulia | the bare metal log, I think I've seen with a few bytes, but essentially blank | 23:05 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Add tests for Payloads with SCHEMAs https://review.openstack.org/418148 | 23:05 |
| *** wrouesnel has joined #openstack-ironic | 23:17 | |
| * swatson is away: I'm away | 23:18 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Default ironic to not use nested KVM https://review.openstack.org/421581 | 23:25 |
| JayF | TheJulia: https://review.openstack.org/#/c/421581/1 if you find any interesting evidence, it can go there | 23:27 |
| patchbot | patch 421581 - ironic - Default ironic to not use nested KVM | 23:27 |
| JayF | TheJulia: I'm going to hit the mailing list about this too, I think | 23:27 |
| TheJulia | JayF: Thank you | 23:28 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Default ironic to not use nested KVM https://review.openstack.org/421581 | 23:35 |
| *** rpioso has quit IRC | 23:55 | |
| *** rajinir has quit IRC | 23:56 | |
| *** [1]cdearborn has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!