*** alexpilotti has joined #openstack-ironic | 00:11 | |
*** alexpilotti has quit IRC | 00:12 | |
*** amotoki has joined #openstack-ironic | 00:17 | |
*** Sukhdev has joined #openstack-ironic | 00:19 | |
*** chlong has joined #openstack-ironic | 00:19 | |
*** moshele has quit IRC | 00:21 | |
*** hoangcx has joined #openstack-ironic | 00:47 | |
*** amotoki has quit IRC | 00:59 | |
*** amotoki has joined #openstack-ironic | 01:21 | |
*** zhenguo_ has joined #openstack-ironic | 01:31 | |
*** lintan_ is now known as lintan | 01:36 | |
*** amotoki has quit IRC | 01:47 | |
*** chlong has quit IRC | 01:56 | |
*** jaybeale has quit IRC | 01:59 | |
*** jaybeale has joined #openstack-ironic | 02:01 | |
*** szhsong has joined #openstack-ironic | 02:03 | |
*** mbound has joined #openstack-ironic | 02:09 | |
*** amotoki has joined #openstack-ironic | 02:10 | |
*** mbound has quit IRC | 02:14 | |
*** yuanying_ is now known as yuanying | 02:17 | |
*** jaybeale has quit IRC | 02:22 | |
*** jaybeale has joined #openstack-ironic | 02:24 | |
*** sivaramakrishna has joined #openstack-ironic | 02:40 | |
openstackgerrit | Merged openstack/ironic: Fix syntax error in devstack create-node script https://review.openstack.org/316421 | 02:42 |
---|---|---|
*** jaybeale has quit IRC | 02:48 | |
*** jaybeale has joined #openstack-ironic | 02:48 | |
*** jaybeale has quit IRC | 02:51 | |
openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/315057 | 02:53 |
openstackgerrit | Nisha Agarwal proposed openstack/proliantutils: [WIP] Test patch for gpu device count https://review.openstack.org/315402 | 03:02 |
*** harlowja_at_home has joined #openstack-ironic | 03:16 | |
openstackgerrit | Tan Lin proposed openstack/ironic: Make use of oslo-config-generator https://review.openstack.org/247331 | 03:18 |
openstackgerrit | Tan Lin proposed openstack/ironic: Add require_exclusive_lock decorators to conductor methods https://review.openstack.org/295734 | 03:19 |
*** baoli has joined #openstack-ironic | 03:26 | |
*** baoli has quit IRC | 03:30 | |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 03:49 |
*** links has joined #openstack-ironic | 03:49 | |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add network drivers https://review.openstack.org/285852 | 03:49 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Update the deploy drivers with network flipping logic https://review.openstack.org/213262 | 03:49 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Added operator documentation for ironic portgroups https://review.openstack.org/228496 | 03:50 |
*** appu103_ has joined #openstack-ironic | 03:56 | |
*** appu103_ is now known as aparnav | 03:56 | |
*** amotoki has quit IRC | 04:15 | |
*** amotoki has joined #openstack-ironic | 04:16 | |
*** jcoufal has joined #openstack-ironic | 04:46 | |
*** harlowja_at_home has quit IRC | 04:50 | |
*** mtanino has quit IRC | 04:58 | |
*** yuriyz has quit IRC | 05:03 | |
*** yuriyz has joined #openstack-ironic | 05:04 | |
*** rcernin has joined #openstack-ironic | 05:16 | |
*** rcernin has quit IRC | 05:21 | |
*** rcernin has joined #openstack-ironic | 05:38 | |
*** moshele has joined #openstack-ironic | 05:51 | |
*** amotoki_ has joined #openstack-ironic | 05:53 | |
*** amotoki has quit IRC | 05:54 | |
*** vmud213 has joined #openstack-ironic | 06:04 | |
openstackgerrit | vinay kumar muddu proposed openstack/ironic-python-agent: Wait for all interfaces to be up before node lookup https://review.openstack.org/303503 | 06:05 |
pas-ha | morning Ironicers | 06:28 |
*** yuanying has quit IRC | 06:30 | |
*** vsaienko has quit IRC | 06:39 | |
*** amotoki_ is now known as amotoki | 06:58 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 06:58 |
*** sacharya has quit IRC | 07:00 | |
*** Sukhdev has quit IRC | 07:00 | |
*** vsaienko has joined #openstack-ironic | 07:01 | |
*** milan has joined #openstack-ironic | 07:03 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for dib based agent ramdisk in lib/ironic https://review.openstack.org/264579 | 07:08 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 07:13 |
*** vmud213 has quit IRC | 07:15 | |
*** jtomasek has joined #openstack-ironic | 07:26 | |
dtantsur | Morning Ironic | 07:34 |
*** yuanying has joined #openstack-ironic | 07:34 | |
*** stendulker has joined #openstack-ironic | 07:36 | |
*** vmud213 has joined #openstack-ironic | 07:39 | |
vdrok | morning ironic, pas-ha dtantsur | 07:41 |
*** yuanying has quit IRC | 07:42 | |
*** yuanying has joined #openstack-ironic | 07:42 | |
*** itamarl has joined #openstack-ironic | 07:43 | |
*** aarefiev has quit IRC | 07:43 | |
*** vsaienko has quit IRC | 07:43 | |
*** yuriyz has quit IRC | 07:43 | |
*** yuanying has quit IRC | 07:43 | |
*** yuriyz has joined #openstack-ironic | 07:44 | |
*** vsaienko has joined #openstack-ironic | 07:45 | |
*** vsaienko1 has joined #openstack-ironic | 07:48 | |
*** vsaienko has quit IRC | 07:48 | |
*** ohamada has joined #openstack-ironic | 07:50 | |
*** vmud213 has quit IRC | 07:52 | |
*** irf has joined #openstack-ironic | 07:58 | |
irf | morning Ironicers!! | 07:58 |
*** lennyb has joined #openstack-ironic | 07:58 | |
*** zzzeek has quit IRC | 08:00 | |
*** zzzeek has joined #openstack-ironic | 08:00 | |
*** sacharya has joined #openstack-ironic | 08:00 | |
*** amotoki has quit IRC | 08:03 | |
*** sacharya has quit IRC | 08:06 | |
*** yuanying has joined #openstack-ironic | 08:08 | |
*** mbound has joined #openstack-ironic | 08:09 | |
*** amotoki has joined #openstack-ironic | 08:11 | |
*** mbound has quit IRC | 08:14 | |
lucasagomes | morning all | 08:18 |
lucasagomes | sambetts|afk, awesome, +1 | 08:19 |
irf | today suddenly , i faced error in the nova conductor logs. Port 00:22:4d:af:79:f4 is still in use.\n' | 08:23 |
irf | and due to that , now i am not able to create instances | 08:23 |
irf | :( | 08:23 |
openstackgerrit | Atsushi SAKAI proposed openstack/ironic: [api-ref]Add/Update driver method samples https://review.openstack.org/316651 | 08:25 |
*** e0ne has joined #openstack-ironic | 08:25 | |
irf | google search did not help me lot ... | 08:27 |
*** amotoki has quit IRC | 08:27 | |
irf | last friday, i spawned the baremetal node using this port ... and after that i deleted that node ... | 08:30 |
*** jistr has joined #openstack-ironic | 08:30 | |
irf | now when i am trying to spawn the node using this port ,,, but not able to cretae that | 08:31 |
irf | i deleted the port as well and recreated that | 08:31 |
irf | but still no luck ... | 08:31 |
dtantsur | harlowja, https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L1140-L1177 | 08:31 |
irf | pleaese suggets ... | 08:31 |
*** dmk0202 has joined #openstack-ironic | 08:32 | |
dtantsur | irf, how did you delete the instance? if you used 'nova delete', it should have removed the neutron port | 08:33 |
irf | no i did not use nova delete.. i deleted form horizon dashboard | 08:34 |
*** kromanenko_ has joined #openstack-ironic | 08:37 | |
*** kromanenko__ has joined #openstack-ironic | 08:37 | |
irf | if i run the command : neutron port-list , it gives me bunch of output ... but if i run the command : ironic port-list then it shows me port no... | 08:39 |
irf | port no is nothing but MAC_ADDRS | 08:39 |
irf | now i am not getting what comamnd i need to execute to get rid of these issue | 08:40 |
dtantsur | irf, so, in horizon, did you delete the instance or the ironic node? I didn't quite get it.. | 08:40 |
irf | i deleted the instance ... | 08:42 |
irf | which was of flavor type baremetal | 08:43 |
dtantsur | it's weird that the port was not deleted.. | 08:44 |
dtantsur | irf, did you try to delete the neutron port and retry? | 08:44 |
irf | i tried ironic port-delete but does not happen anything | 08:45 |
*** derekh has joined #openstack-ironic | 08:45 | |
*** amotoki has joined #openstack-ironic | 08:46 | |
irf | shud i run the neutron port-list | 08:46 |
irf | when i run the neutron port-list , it is not showing my port no ... | 08:47 |
irf | i ran the comamnd : neutron port-list | grep 00:22:4d:af:79:f4 | 08:49 |
irf | and it showed me one output | 08:49 |
irf | will delete this port and see .... | 08:49 |
dtantsur | neutron ports and ironic ports are independent | 08:53 |
dtantsur | ironic ports are created by you, neutron ones - by nova | 08:53 |
irf | ok | 08:55 |
*** zzzeek has quit IRC | 08:56 | |
lucasagomes | irf, check neutron yeah | 08:56 |
lucasagomes | you may need to do a port delete on the port with that mac address | 08:57 |
lucasagomes | irf, when you deleted the instance form the node, did you do it via nova? (nova delete ...) | 08:57 |
irf | nop | 08:57 |
irf | i just deleetd from horizon ... | 08:57 |
*** zzzeek has joined #openstack-ironic | 08:58 | |
*** electrofelix has joined #openstack-ironic | 08:59 | |
lucasagomes | right, well that should do it via nova internally | 09:00 |
lucasagomes | strange the port is still taken | 09:01 |
irf | ok .. i have deleted the port and now try to create the instance manually... | 09:03 |
irf | it is getting created and now it is ins spawning state | 09:04 |
sambetts|afk | Morning all | 09:07 |
*** sambetts|afk is now known as sambetts | 09:07 | |
*** vmud213 has joined #openstack-ironic | 09:08 | |
*** amotoki has quit IRC | 09:10 | |
*** yuanying has quit IRC | 09:14 | |
*** vmud213 has quit IRC | 09:18 | |
*** vmud213 has joined #openstack-ironic | 09:18 | |
irf | its resolved now | 09:23 |
irf | i deleted the port and now instance is getting spawned. instance means not the baremetal but normal VM ... | 09:23 |
*** vmud213 has quit IRC | 09:28 | |
*** vmud213 has joined #openstack-ironic | 09:28 | |
lucasagomes | irf, cool | 09:30 |
irf | :) | 09:30 |
irf | thanks much :) | 09:30 |
lucasagomes | yw | 09:31 |
dtantsur | lucasagomes, well, maybe we should not | 09:34 |
lucasagomes | dtantsur, hah my bad on the openstack-qa | 09:34 |
dtantsur | lucasagomes, inspector operations are pluggable, while nova's are not | 09:34 |
lucasagomes | didn't notice I was on that channel | 09:34 |
lucasagomes | dtantsur, yeah, I personally like the inspector ones more (it's similar to what I suggested) | 09:35 |
*** vmud213 has quit IRC | 09:35 | |
lucasagomes | but yeah I do feel like the nova ones are more consistent overall | 09:35 |
lucasagomes | are you OK with root device hints mimic'ing the oslo ones then? | 09:35 |
lucasagomes | perhaps using the oslo spec matcher that jroll proposed | 09:35 |
dtantsur | lucasagomes, yeah, I think it's fine | 09:35 |
lucasagomes | ack, thanks | 09:35 |
*** moshele has quit IRC | 09:37 | |
*** moshele has joined #openstack-ironic | 09:37 | |
*** vmud213 has joined #openstack-ironic | 09:47 | |
*** dtantsur is now known as dtantsur|brb | 09:53 | |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: refactor ironic enroll-node code https://review.openstack.org/256364 | 09:55 |
*** vmud213 has quit IRC | 09:55 | |
*** vmud213 has joined #openstack-ironic | 09:56 | |
*** vmud213 has quit IRC | 09:58 | |
*** vmud213 has joined #openstack-ironic | 09:58 | |
lucasagomes | strange error in one agent_ssh job http://logs.openstack.org/93/315593/1/gate/gate-tempest-dsvm-ironic-agent_ssh/13730e8/logs/screen-ir-cond.txt.gz#_2016-05-16_09_35_44_263 | 10:00 |
* lucasagomes digs more, but looks like some race condition in the tests somewhere | 10:00 | |
*** daemontool has joined #openstack-ironic | 10:01 | |
*** sacharya has joined #openstack-ironic | 10:02 | |
*** vmud213 has quit IRC | 10:02 | |
*** vmud213 has joined #openstack-ironic | 10:02 | |
openstackgerrit | Debayan Ray proposed openstack/proliantutils: Add supported boot modes to server capabilities https://review.openstack.org/275053 | 10:03 |
openstackgerrit | Aline Bousquet proposed openstack/ironic-lib: Add support for metrics https://review.openstack.org/301526 | 10:06 |
*** sacharya has quit IRC | 10:07 | |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add Link-Local-Connection info to ironic port https://review.openstack.org/256365 | 10:12 |
*** yuanying has joined #openstack-ironic | 10:13 | |
*** amotoki has joined #openstack-ironic | 10:14 | |
*** vmud213 has quit IRC | 10:19 | |
*** mgould has joined #openstack-ironic | 10:23 | |
mgould | morning Ironic! | 10:24 |
sambetts | o/ mgould | 10:24 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 10:25 |
lucasagomes | mgould, good morning | 10:25 |
mgould | sambetts, lucasagomes morning! | 10:25 |
vdrok | morning mgould sambetts and lucasagomes | 10:25 |
mgould | morning vdrok | 10:25 |
* mgould is currently suffering from being ill at the weekend and going climbing anyway :-( | 10:26 | |
*** caoshufeng has joined #openstack-ironic | 10:27 | |
irf | morning mgould | 10:27 |
sambetts | mgould: :( | 10:28 |
mgould | irf, morning! | 10:28 |
caoshufeng | Hi, I use agent_ssh driver deploy ironic. We I use "nova boot" , I get such error: | 10:28 |
caoshufeng | iPXE 1.0.0+ (dc795b9f) -- Open Source Network Boot Firmware -- http://ipxe.org Features: DNS HTTP iSCSI TFTP AoE ELF MBOOT PXE bzImage Menu PXEXT | 10:28 |
caoshufeng | net0: 52:54:00:04:09:9c using virtio-net on PCI00:04.0 (open) | 10:29 |
caoshufeng | [Link:up, TX:0 TXE:0 RX:0 RXE:0] | 10:29 |
caoshufeng | Configuring (net0 52:54:00:04:09:9c)...... ok | 10:29 |
caoshufeng | net0: 192.168.1.4/255.255.255.0 gw 192.168.1.1 | 10:29 |
caoshufeng | Next server: 10.167.133.33 | 10:29 |
caoshufeng | Filename: pxelinux.0 | 10:29 |
caoshufeng | tftp://10.167.133.33/pxelinux.0... Connection timed out (http://ipxe.org/4c12603 | 10:29 |
caoshufeng | No bootable device. | 10:29 |
*** hoangcx has quit IRC | 10:30 | |
caoshufeng | This is the console log of one node | 10:30 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Allow to use NETWORK_DRIVERS in devstack https://review.openstack.org/293520 | 10:30 |
*** lennyb has quit IRC | 10:32 | |
*** moshele has quit IRC | 10:34 | |
lucasagomes | caoshufeng, apparently the node can't talk to the TFTP server to chainload the iPXE ROM | 10:34 |
lucasagomes | caoshufeng, either the TFTP service is not setup/running or is inaccessible from the node | 10:34 |
lucasagomes | the PXE ROM (sorry, no iPXE there) | 10:35 |
*** vmud213 has joined #openstack-ironic | 10:35 | |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Update Ironic VM network connection https://review.openstack.org/256366 | 10:36 |
openstackgerrit | Merged openstack/ironic: Allow vendor drivers to acquire shared locks https://review.openstack.org/226058 | 10:37 |
openstackgerrit | Merged openstack/ironic: Only install DIB if going to use DIB https://review.openstack.org/315795 | 10:37 |
openstackgerrit | Merged openstack/ironic: Modify doc for RAID clean steps in manual cleaning https://review.openstack.org/315998 | 10:37 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add configure_provision_network function https://review.openstack.org/256367 | 10:38 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add Ironic/Neutron integration documentation https://review.openstack.org/258596 | 10:39 |
openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add test to verify ironic multitenancy https://review.openstack.org/269157 | 10:39 |
*** moshele has joined #openstack-ironic | 10:40 | |
*** sivaramakrishna has quit IRC | 10:44 | |
caoshufeng | lucasagomes, Hi, so what do I need to to? How can I make private net 192.168.1.4 access to the host IP 10.167.133.33? | 10:51 |
*** vmud213 has quit IRC | 10:51 | |
*** vmud213 has joined #openstack-ironic | 10:52 | |
lucasagomes | caoshufeng, are you using a flat network for the deployment? | 10:53 |
caoshufeng | vxlan | 10:53 |
lucasagomes | you need the nodes being deployed to be able to talk to the control panel | 10:53 |
caoshufeng | lucasagomes. I deploy my environment according to this manual: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack | 10:56 |
caoshufeng | I only changed this: IRONIC_DEPLOY_DRIVER=pxe_ssh -------> IRONIC_DEPLOY_DRIVER=agent_ssh | 10:56 |
lucasagomes | caoshufeng, strange then... can you check if the tftp service is up and running? | 11:00 |
lucasagomes | caoshufeng, xinetd | 11:00 |
lucasagomes | you can also try to run "tftp" and do a "get pxelinux.0" | 11:01 |
lucasagomes | see if that works for you | 11:01 |
*** yuanying has quit IRC | 11:01 | |
caoshufeng | lucasagomes, xinetd is active | 11:02 |
caoshufeng | lucasagomes, tftp is not a command in my environment. | 11:03 |
lucasagomes | caoshufeng, can you install it ? Just to check | 11:04 |
*** vmud213 has quit IRC | 11:05 | |
*** vmud213 has joined #openstack-ironic | 11:05 | |
*** yuanying has joined #openstack-ironic | 11:05 | |
caoshufeng | (to) get pxelinux.0 | 11:05 |
caoshufeng | Error: Name or service not known | 11:05 |
caoshufeng | get: unknown host | 11:06 |
caoshufeng | lucasagomes, seems something wrong here. | 11:06 |
lucasagomes | maybe syntax | 11:06 |
* lucasagomes checks | 11:07 | |
*** stendulker has quit IRC | 11:08 | |
caoshufeng | lucasagomes, tftp command works in my environment. | 11:08 |
caoshufeng | I forget to run "connect" before "get" | 11:09 |
lucasagomes | caoshufeng, right hmm that error (connection timed out) happens on every attempt? | 11:10 |
caoshufeng | Yes. | 11:10 |
* lucasagomes thinks | 11:10 | |
*** krtaylor has quit IRC | 11:10 | |
caoshufeng | 1. I deployed my devstack 2. I run "nova boot". Do I miss some steps? | 11:11 |
caoshufeng | lucasagomes, Do I need to deal with iptables or nova secgroup? | 11:12 |
lucasagomes | caoshufeng, you shouldn't no | 11:13 |
lucasagomes | caoshufeng, are you using ubuntu as a host? | 11:13 |
sambetts | caoshufeng: can you check your other OpenStack logs, especially the q-* ones, to make sure there are no errors from any of the other services | 11:13 |
*** _fortis has quit IRC | 11:13 | |
lucasagomes | caoshufeng, ^ yeah as well... | 11:13 |
lucasagomes | sambetts, tho by the error ir seems something tftp-related | 11:13 |
sambetts | lucasagomes: I have found that if you delete the neutron router/ or its not setup right, it will cause errors like this | 11:14 |
*** jtomasek_ has joined #openstack-ironic | 11:15 | |
caoshufeng | lucasagomes, I use Fedora 23. | 11:15 |
lucasagomes | caoshufeng, right, that's what I use as well... should be fine hmm | 11:15 |
lucasagomes | yeah check the logs as sambetts asked, you can do a "screen -x" to attach to the screen session | 11:16 |
lucasagomes | and you will see the service logs there | 11:16 |
caoshufeng | No error log found in q-*.log. | 11:19 |
lucasagomes | caoshufeng, started a devstack setup in a f23 here, will take a look | 11:20 |
caoshufeng | Oh, Thanks! | 11:21 |
* lucasagomes go grab some quick lunch in between | 11:21 | |
*** lucasagomes is now known as lucas-hungry | 11:21 | |
*** leakypipes is now known as jaypipes | 11:22 | |
*** krtaylor has joined #openstack-ironic | 11:24 | |
*** dtantsur|brb is now known as dtantsur | 11:28 | |
caoshufeng | lucasagomes, It's 7:30PM in my timezone now, I will go off work now. | 11:29 |
caoshufeng | Thanks again. | 11:30 |
sambetts | o/ | 11:30 |
caoshufeng | lucasagomes, I will read your answer tomorrow. | 11:30 |
caoshufeng | sambetts, thanks. | 11:30 |
*** vmud213 has quit IRC | 11:30 | |
*** vmud213 has joined #openstack-ironic | 11:31 | |
*** amotoki has quit IRC | 11:36 | |
*** jtomasek_ has quit IRC | 11:38 | |
*** cdearborn has joined #openstack-ironic | 11:40 | |
openstackgerrit | Sam Betts proposed openstack/ironic-specs: Add spec for VLAN Aware Baremetal Instances https://review.openstack.org/277853 | 11:46 |
openstackgerrit | Merged openstack/ironic: DevStack: Support to install virtualbmc from source https://review.openstack.org/315593 | 11:48 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Migrate to os-api-ref library https://review.openstack.org/316754 | 11:49 |
jroll | mornin y'all | 11:49 |
dtantsur | morning jroll | 11:49 |
jroll | :D | 11:49 |
*** yuanying has quit IRC | 11:49 | |
dtantsur | jroll, https://review.openstack.org/316655 | 11:50 |
jroll | dtantsur: wooo | 11:51 |
*** moshele has quit IRC | 11:52 | |
jroll | dtantsur: one thing but lgtm | 11:52 |
dtantsur | sambetts, I realized that we might need to implement https://bugs.launchpad.net/ironic-inspector/+bug/1580525 to be able to implement manage_boot | 11:52 |
openstack | Launchpad bug 1580525 in Ironic Inspector "[RFE] Cache MAC addresses from previous introspection runs to enhance our blacklist" [Wishlist,Confirmed] | 11:52 |
*** amotoki has joined #openstack-ironic | 11:52 | |
dtantsur | jroll, we still have kilo, and it's everywhere around, so dunno really | 11:53 |
*** moshele has joined #openstack-ironic | 11:53 | |
*** yuanying has joined #openstack-ironic | 11:53 | |
*** Goneri has quit IRC | 11:53 | |
*** vmud213 has quit IRC | 11:53 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove verbose option https://review.openstack.org/315990 | 11:55 |
jroll | ^ that should be landable now | 11:55 |
dtantsur | cool | 11:55 |
*** mtanino has joined #openstack-ironic | 11:57 | |
sambetts | Morning jroll | 11:57 |
jroll | dtantsur: omg, why can't I find this email | 11:57 |
jroll | aha | 11:58 |
jroll | dtantsur: http://lists.openstack.org/pipermail/openstack/2016-May/016142.html | 11:58 |
jroll | * Please note, that this is the final release of Kilo and no further | 11:58 |
jroll | updates will be provided * | 11:58 |
dtantsur | I know that | 11:58 |
dtantsur | still, we have a stable/kilo branch, and project-config is not updated yet | 11:58 |
*** dprince has joined #openstack-ironic | 11:59 | |
jroll | sure, they're working on the branches right now | 11:59 |
jroll | just don't think we should add yet another piece of to-be-cruft | 11:59 |
sambetts | do the branches just get archieved and nuked? | 12:00 |
jroll | yeah, they create a kilo-eol branch and remove kilo | 12:00 |
sambetts | ah makes sense | 12:00 |
mat128 | good monday morning ironic o/ | 12:01 |
sambetts | morning mat128 | 12:01 |
dtantsur | jroll, updated; lets see what infra folks say | 12:02 |
dtantsur | morning mat128 | 12:02 |
jroll | dtantsur: thanks | 12:02 |
jroll | +1'd | 12:02 |
*** cdearborn has quit IRC | 12:03 | |
*** aparnav has quit IRC | 12:05 | |
sambetts | dtantsur: I'm not sure if we need that for unmanaged boot, (I guess it depends on the setting they have in their configuration about which macs to store in Ironic) | 12:06 |
dtantsur | sambetts, I thought we might need a table to store MAC's.. but on a second glance we might do with only the current database | 12:07 |
sambetts | dtantsur: also I've just thought, what happens if they set the config to store pxe port only, and then inspect something using virt media? | 12:07 |
*** amotoki has quit IRC | 12:08 | |
sambetts | jroll: https://bugs.launchpad.net/ironic/+bug/1582188 | 12:08 |
openstack | Launchpad bug 1582188 in Ironic "[RFE] Add interface attach API" [Undecided,New] - Assigned to Sam Betts (sambetts) | 12:08 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Migrate to os-api-ref library https://review.openstack.org/316754 | 12:08 |
dtantsur | sambetts, correct, we should document that it won't work | 12:09 |
*** jtomasek_ has joined #openstack-ironic | 12:10 | |
*** amotoki has joined #openstack-ironic | 12:10 | |
sambetts | dtantsur: The current DB model handles multiple mac addresses, we may just need a different attribute name for non-ironic mac addresses | 12:12 |
jroll | sambetts: woot | 12:12 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Migrate to os-api-ref library https://review.openstack.org/316754 | 12:12 |
sambetts | jroll: I was going through the VLAN aware BM spec this morning answering the questions, and adding that API and hiding the proposed vif_port_ids/trunk_port_id would be sooooo much nicer | 12:13 |
jroll | sambetts: yeah, totally agree | 12:15 |
jroll | sambetts: I can't wait to see the discussion about what the path should be :) | 12:16 |
*** yuanying has quit IRC | 12:17 | |
*** amotoki has quit IRC | 12:17 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Eliminate warnings about rm in api-ref build https://review.openstack.org/316793 | 12:22 |
*** ohamada has quit IRC | 12:22 | |
*** ohamada has joined #openstack-ironic | 12:22 | |
*** lucas-hungry is now known as lucasagomes | 12:26 | |
lucasagomes | jroll, morning | 12:26 |
jroll | heya lucasagomes | 12:29 |
*** ohamada_ has joined #openstack-ironic | 12:29 | |
*** ohamada has quit IRC | 12:29 | |
*** ohamada__ has joined #openstack-ironic | 12:29 | |
*** ohamada_ has quit IRC | 12:29 | |
*** trown|brb is now known as trown | 12:31 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Doc: Replace nova image-list https://review.openstack.org/316796 | 12:36 |
*** xavierr has joined #openstack-ironic | 12:39 | |
xavierr | Good morning Ironic :) | 12:40 |
lucasagomes | xavierr, morning | 12:41 |
xavierr | morning lucasagomes o/ | 12:42 |
mgould | xavierr, jroll morning | 12:42 |
*** ppiela has joined #openstack-ironic | 12:42 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: [WIP] Support manage_boot parameter to introspection API https://review.openstack.org/316801 | 12:45 |
xavierr | guys, I have a question. Why don't we have 'last_error' and 'maintenance_reason' when filtering for a node? | 12:45 |
xavierr | here: https://github.com/openstack/ironic/blob/master/ironic/db/sqlalchemy/api.py#L186 | 12:45 |
lucasagomes | caoshufeng, tftp://192.168.122.158/pxelinux.0... ok hmm seems fine here :-/ | 12:46 |
lucasagomes | caoshufeng, Linux localhost.localdomain 4.2.3-300.fc23.x86_64 #1 SMP Mon Oct 5 15:42:54 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux | 12:46 |
lucasagomes | will check if the deployment completes fine | 12:46 |
dtantsur | xavierr, cause someone complained that DB2 (maybe other databases) can't use TEXT fields in such queries | 12:46 |
*** irf has quit IRC | 12:47 | |
*** ohamada__ has quit IRC | 12:48 | |
mgould | we support DB2? wow | 12:48 |
*** ohamada__ has joined #openstack-ironic | 12:48 | |
dtantsur | well, these people apparently used it | 12:48 |
dtantsur | we don't need fancy database features, so I pretty much expect us to support everything sane | 12:49 |
xavierr | dtantsur, I didn't know that | 12:49 |
mgould | dtantsur, ISTR being bitten by differences in DB semantics in some surprisingly non-fancy features | 12:50 |
openstackgerrit | Atsushi SAKAI proposed openstack/ironic: api-ref:Add REST API for / and /v1 https://review.openstack.org/316806 | 12:51 |
dtantsur | mgould, well, we were also bit by relying on sqlite in gate | 12:51 |
dtantsur | I forgot it does not even enforce foreign keys | 12:51 |
mgould | huh, I didn't know that | 12:52 |
mgould | on second thoughts, I think it may have been sqlite/postgres differences that bit me | 12:53 |
dtantsur | yeah, and it a has weak type system (in a sense of weak vs strong, not in a sense of "bad") | 12:53 |
mgould | so if sqlite isn't included in {sane databases}, I withdraw that objection :-) | 12:53 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Doc: Replace nova image-list https://review.openstack.org/316796 | 12:53 |
mgould | OTOH, inability to filter by a TEXT field is pretty unimpressive IMHO | 12:55 |
* lucasagomes sometimes think that we should use postgres in gate because it's more strict overall | 12:55 | |
mgould | are we currently using sqlite in gate? | 12:55 |
lucasagomes | nop, mysql | 12:55 |
*** links has quit IRC | 12:55 | |
mgould | oh, ofc | 12:56 |
* mgould votes for "test with what we recommend people deploy with" :-) | 12:56 | |
e0ne | hi. can anybody help me with https://review.openstack.org/#/c/287938/, please? I don't see any errors, only NOT_REGISTERED jobs | 12:56 |
*** rbudden has joined #openstack-ironic | 12:56 | |
dtantsur | e0ne, NOT_REGISTERED is a temporary condition, recheck it | 12:56 |
e0ne | dtantsur: we tried it few times. should I recheck once more? | 12:57 |
lucasagomes | e0ne, ask the guys in #openstack-qa if you already retried it | 12:57 |
lucasagomes | sorry #openstack-infra | 12:57 |
dtantsur | e0ne, ping #openstack-infra for advice then. but it might take time, so I'd just wait | 12:58 |
e0ne | oops. I've asked it in a wrong channel:( | 12:58 |
e0ne | I'm sorry | 12:58 |
lucasagomes | all good :-) | 12:58 |
dtantsur | :) | 12:58 |
e0ne | thanks for the help, team:) | 12:58 |
dtantsur | e0ne, it usually means that the jobs were being updated when your gate run happened. it might take time. | 12:58 |
*** athomas has quit IRC | 13:01 | |
*** alexpilotti has joined #openstack-ironic | 13:02 | |
xavierr | dtantsur, Any idea what could we do for work around this ('last_error' and 'maintenance_reason' queries)? | 13:02 |
*** Goneri has joined #openstack-ironic | 13:03 | |
dtantsur | xavierr, what's your use case? | 13:03 |
*** rloo has joined #openstack-ironic | 13:04 | |
xavierr | dtantsur, my case is, when something goes wrong in the node caused by any OneView problem, I'm updating the field last_error with the massage that describes the error | 13:05 |
xavierr | dtantsur, after that I want to repair that nodes that have last_error | 13:06 |
xavierr | so, I want to filter them, but actually I can't :( | 13:07 |
dtantsur | xavierr, then anyway a direct == comparison is suboptimal for you. it makes sense to introduce something with LIKE operator or regexps | 13:07 |
*** mtanino has quit IRC | 13:08 | |
* lucasagomes remembers we approved a similar spec for querying node names with LIKE | 13:08 | |
lucasagomes | perhaps we could re-use some of that logic | 13:09 |
xavierr | dtantsur, do you mean bring all nodes and make a loop comparing the 'last_error'? | 13:09 |
lucasagomes | xavierr, http://specs.openstack.org/openstack/ironic-specs/specs/approved/node-name-regexp-api.html#proposed-change | 13:09 |
mgould | yay, O(n) network traffic... | 13:09 |
*** baoli has joined #openstack-ironic | 13:09 | |
dtantsur | xavierr, I mean maybe a new API parameter last_error_like="blah-blah %" | 13:09 |
mgould | xavierr, how many nodes do you have? | 13:09 |
dtantsur | xavierr, similar to what lucasagomes posted | 13:10 |
* mgould supposes it doesn't really matter what n is here, we want to work at roflscale | 13:11 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Doc: Replace nova image-list https://review.openstack.org/316796 | 13:11 |
mgould | lucasagomes, lgtm | 13:11 |
lucasagomes | :-) | 13:11 |
lucasagomes | ty | 13:12 |
mgould | np | 13:12 |
*** jaybeale has joined #openstack-ironic | 13:14 | |
xavierr | dtantsur, lucasagomes, mgould, I'm trying to filter nodes inside a driver, using 'base_manager.iter_nodes' in a periodic task. Do you think would be ok using the a new parameter API in a periodic task? | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated driver_periodic_task https://review.openstack.org/316820 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "discoverd" configuration group https://review.openstack.org/316821 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [agent]/manage_tftp option https://review.openstack.org/316822 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated [pxe]/http_* options https://review.openstack.org/316823 | 13:21 |
jroll | spaaaaaam | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "periodic_interval" config option https://review.openstack.org/316824 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "message" attribute support from IronicException https://review.openstack.org/316825 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [conductor]/clean_nodes config option https://review.openstack.org/316826 | 13:21 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove backward compat for Liberty cleaning https://review.openstack.org/316827 | 13:21 |
*** vmud213 has joined #openstack-ironic | 13:21 | |
dtantsur | wow | 13:21 |
sambetts | hahaha | 13:21 |
dtantsur | much clean so delete deprecation wow | 13:21 |
mgould | :-) | 13:22 |
dtantsur | xavierr, I didn't quite get the question. what do you mean by "using the a new parameter API in a periodic task"? | 13:22 |
dtantsur | jroll, now I suspect you have to update the sample for all these ;) | 13:22 |
*** causten_ has joined #openstack-ironic | 13:22 | |
jroll | gabriel-bezerra: hey, I'm cleaning up deprecations, there was one in the oneview driver, I'm wondering if your team could clean that up? | 13:22 |
jroll | dtantsur: heh, yep, doing it now | 13:23 |
mgould | dtantsur, xavierr: more basic question: do drivers make API calls? I thought the API called drivers, not the other way round... | 13:23 |
dtantsur | mgould, you're right, hence my question | 13:23 |
lucasagomes | I think he means API because we call the module api.py (db/sqlalchemy/api.py) | 13:24 |
mgould | oh, OK | 13:24 |
lucasagomes | xavierr, if you refer to the DB code, yes I think it would be an optimization | 13:24 |
mgould | I guess there's an API in the broader sense | 13:24 |
lucasagomes | yeah | 13:24 |
lucasagomes | xavierr, not a requirement for what you are doing, because you still can loop on _all_ nodes (sure it will suffer some speed penalty doing that) | 13:25 |
xavierr | dtantsur, mgould, lucasagomes oh, understood | 13:26 |
*** ChrisAusten has quit IRC | 13:26 | |
*** causten_ has quit IRC | 13:27 | |
mgould | lucasagomes, dtantsur: but xavierr could add an add_node_filter_by_last_error method to sqlalchemy.api, right? | 13:27 |
lucasagomes | yup | 13:27 |
lucasagomes | even in the API, I think it iwll look similar to the node names one | 13:27 |
lucasagomes | ReST API* | 13:28 |
* mgould nods | 13:28 | |
mgould | so I guess there are three tasks: | 13:28 |
mgould | 1) write backend filter-by-last-error code | 13:28 |
mgould | 2) provide an HTTP endpoint for it | 13:28 |
mgould | 3) write driver code, possibly using 1) | 13:29 |
*** alexpilo_ has joined #openstack-ironic | 13:29 | |
mgould | 3) is made faster by 1), but doesn't require it | 13:29 |
mgould | 3) and 2) are basically orthogonal | 13:29 |
mgould | yes? | 13:29 |
sambetts | mgould: https://review.openstack.org/#/c/306092/6/specs/approved/node-search-api.rst | 13:30 |
sambetts | could roll all that into ^ | 13:30 |
mgould | sambetts, excellent | 13:30 |
*** alexpil__ has joined #openstack-ironic | 13:30 | |
xavierr | roger that, thank you guys | 13:31 |
xavierr | thank you mgould :) | 13:31 |
*** alexpilotti has quit IRC | 13:31 | |
jroll | oh boo, git blame shows the date the change was first committed, not when it merged :| | 13:32 |
jroll | https://review.openstack.org/#/c/264720/ merged feb 24 | 13:32 |
jroll | which means we aren't at the 3-month hit yet | 13:33 |
* jroll will re-order | 13:33 | |
*** sabeen has joined #openstack-ironic | 13:33 | |
sambetts | :( | 13:33 |
*** alexpilo_ has quit IRC | 13:34 | |
*** rloo has quit IRC | 13:35 | |
* mgould tries to work out how to get "date this line was merged" automatically | 13:35 | |
mgould | not easy, AFAICT | 13:35 |
*** rloo has joined #openstack-ironic | 13:36 | |
mgould | git log --merges $hash, and then take the last line | 13:36 |
sambetts | mgould: git blame | grep <line> | awk to get the git sha | xargs git show | grep date: ?? | 13:37 |
mgould | sambetts, that'll give you the commit date too | 13:37 |
sambetts | ah, i forgot those are often different :/ | 13:38 |
* mgould will spend five more minutes on this rabbithole | 13:38 | |
jroll | lol | 13:39 |
* sambetts is trying to work out why his CI has blown up :'( | 13:39 | |
jroll | git blame | grep <line> | awk to get the git sha | xargs git show | grep Change-Id | more awk | query gerrit api with change id | 13:39 |
mgould | heh | 13:40 |
mgould | git: beautiful internal model, really horrible UI | 13:40 |
mgould | almost as if it were thrown together one command at a time by C programmers... | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove backward compat for Liberty cleaning https://review.openstack.org/316827 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [conductor]/clean_nodes config option https://review.openstack.org/316826 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "message" attribute support from IronicException https://review.openstack.org/316825 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "periodic_interval" config option https://review.openstack.org/316824 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated [pxe]/http_* options https://review.openstack.org/316823 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [agent]/manage_tftp option https://review.openstack.org/316822 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "discoverd" configuration group https://review.openstack.org/316821 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated driver_periodic_task https://review.openstack.org/316820 | 13:41 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Regenerate sample config https://review.openstack.org/316849 | 13:41 |
rloo | hi and a sunny good morning to everyone. mgould, sambetts, xavierr, lucasagomes, dtantsur, jroll the spammer | 13:41 |
jroll | rebase -i ftw | 13:42 |
jroll | morning rloo :) | 13:42 |
krotscheck | ppiela: Comment on https://review.openstack.org/#/c/316406/3 -> There's a thing you're doing which _might_ be intentional, but might not? | 13:42 |
dtantsur | morning rloo | 13:42 |
sambetts | rloo: Hey! | 13:42 |
NobodyCam | good morning Ironic'ers | 13:42 |
krotscheck | It's a rloo! | 13:42 |
krotscheck | Hi rloo! | 13:42 |
jroll | dtantsur: those should be good to go, I wip'd the futurist-related things until we're at three months | 13:42 |
sambetts | o/ NobodyCam | 13:42 |
jroll | \o NobodyCam | 13:42 |
rloo | morning NobodyCam, krotscheck! | 13:42 |
krotscheck | This channel is so much more fun now that I can put faces to names. | 13:42 |
NobodyCam | morning sambetts, rloo, jroll, mgould | 13:42 |
NobodyCam | hehe :) morning krotscheck | 13:43 |
rloo | folks, lucasagomes mentioned that with ironic-lib, we don't need to worry about changes that might break any out-of-tree use of it. Want to make sure everyone agrees. wrt https://review.openstack.org/#/c/311613/4/ironic_lib/disk_partitioner.py | 13:43 |
* jroll brb, breakfast | 13:43 | |
mgould | morning NobodyCam | 13:44 |
mgould | morning rloo | 13:44 |
jroll | rloo: before I go, that's an interesting question, it is a "public" method | 13:44 |
rloo | jroll: exactly. which is why i assumed it couldn't be broken. should i bring it up in our meeting instead? | 13:45 |
lucasagomes | I still think it's fine | 13:45 |
jroll | rloo: given it isn't hard to support both, I say support both, mark bootable deprecated, remove it later in a 'major' release | 13:45 |
lucasagomes | as long as it doesn't break ironic-projects | 13:45 |
NobodyCam | morning lucasagomes :) | 13:45 |
lucasagomes | NobodyCam, morning | 13:46 |
jroll | the readme says "A common library to be used by various projects in the Ironic ecosystem." interpret as you wish :) | 13:46 |
rloo | jroll: i agree it isn't hard to support both. what i want is our 'policy' on this going forward. | 13:46 |
*** joprovost has joined #openstack-ironic | 13:46 | |
jroll | rloo: yeah, I get that, personally I'd send an email | 13:46 |
dtantsur | jroll, all looks good, I'll go check if tripleo uses something deprecated :D | 13:46 |
jroll | I don't have a large problem with either direction | 13:46 |
jroll | dtantsur: heh | 13:46 |
rloo | ok, i'll send email then. (just wanted a decision fast but...). | 13:46 |
lucasagomes | jroll, right, maybe it should say ironic umbrella? I don't think we ever said ironic-lib can be used for projects outside it | 13:46 |
jroll | dtantsur: my next task is grep our devstack logs for deprecated things | 13:47 |
rloo | lucasagomes: you good if I bring it up in email? | 13:47 |
lucasagomes | rloo, sure | 13:47 |
jroll | lucasagomes: rloo: maybe the answer is just do it, and if anyone complains we tell them sorry? | 13:47 |
jroll | regardless, we should properly version that release | 13:47 |
sambetts | lucasagomes: surly it can be used by third party out of tree hardware managers and drivers right/ | 13:47 |
sambetts | ? | 13:47 |
rloo | fwiw, it is much easier to do as you suggest lucasagomes, so i'd prefer that. i just don't know if it is 'right'. | 13:47 |
jroll | great point | 13:48 |
rloo | sambetts: that's my question | 13:48 |
* jroll really brb now | 13:48 | |
lucasagomes | sambetts, not sure... honestly the way it was idealized IMHO was to share code between the official ironic projects | 13:49 |
lucasagomes | it's not oslo, or any public library out there | 13:49 |
sambetts | I would prefer we make it a public lib, and actually make it work similar to what neutron is doing with neutron-lib right now, their plan is that no external drivers inherit anything from neutron directly, anything they want to make accessible goes into neutron-lib, like the driver base classes etc | 13:50 |
lucasagomes | https://review.openstack.org/#/c/157636/6 | 13:50 |
*** jaybeale has quit IRC | 13:50 | |
*** wajdi has joined #openstack-ironic | 13:50 | |
*** joprovost has quit IRC | 13:50 | |
lucasagomes | sambetts, right, I'm not against doing it. Just pointing out that I believe that wasn't part of the conception of the library | 13:51 |
lucasagomes | and AFAIK we never made any statement that the lib is a public lib | 13:51 |
rloo | lucasagomes: just because we say it will be used by ironic & ipa, we didn't say it wouldn't be supported for use by other. | 13:51 |
lucasagomes | we don't even document the method there | 13:51 |
lucasagomes | methods* | 13:51 |
sambetts | yeah :/ not sure about that | 13:51 |
rloo | my impression, when we created that library, was that it was a library and hence could be used by whatever. | 13:51 |
dtantsur | "Option "http_url" from group "pxe" is deprecated. Use option "http_url" from group "deploy"." someone is gonna be broken :( | 13:51 |
lucasagomes | rloo, right, yeah maybe we interpreted things differently | 13:52 |
jlvillal | jroll, When you get a chance. A fix for the stable/mitaka release notes issue: https://review.openstack.org/#/c/316288/ | 13:52 |
lucasagomes | cause I never thought it as a oslo-ish lib | 13:52 |
lucasagomes | of it* | 13:52 |
mgould | jroll, sambetts: git log --merges --ancestry-path {sha1 of offending commit}.. | grep Date | tail -n 1 | 13:54 |
mgould | whew | 13:54 |
*** mtanino has joined #openstack-ironic | 13:54 | |
sambetts | mgould: haha how was that rabbithole? | 13:55 |
*** ChrisAusten has joined #openstack-ironic | 13:55 | |
xavierr | morning rloo :) | 13:55 |
mgould | sambetts, the sun is but a distant memory | 13:56 |
jlvillal | mgould, sambetts Do you know about vim-fugitive? Slightly related | 13:56 |
*** zzzeek has quit IRC | 13:56 | |
jlvillal | Not sure if you use vim or not | 13:56 |
* jlvillal is pretty sure he told lucasagomes about vim-fugitive and Gblame | 13:56 | |
mgould | jlvillal, I have fugitive installed, but don't use it all that often | 13:56 |
mgould | Gblame is one of the things I use it for, though | 13:57 |
* jlvillal loves Gblame :) | 13:57 | |
lucasagomes | jlvillal, yup, I remember that | 13:57 |
lucasagomes | quite useful | 13:57 |
*** ametts has joined #openstack-ironic | 13:57 | |
sambetts | mgould, jlvillal: Not used it personally | 13:57 |
jlvillal | With '~' or '^' to step back in time. | 13:57 |
rloo | lucasagomes, sambetts, jroll: email sent out. Let the discussion begin :) | 13:57 |
lucasagomes | rloo, thanks | 13:57 |
jlvillal | and then 'o' to open the offending commit | 13:57 |
*** zzzeek has joined #openstack-ironic | 13:58 | |
*** sacharya has joined #openstack-ironic | 14:03 | |
*** jaybeale has joined #openstack-ironic | 14:05 | |
dtantsur | jroll, JayF, one more funny issue >_< our lsblk command in IPA hardware.py returns nothing during introspection, but properly returns a RAID logical volume if we ssh into the ramdisk and run it there. have you seen anything like that? | 14:07 |
jroll | jlvillal: this started with "Gblame only tells me when the commit was made, not when it merges" :) | 14:07 |
dtantsur | quick googling gives surprisingly nothing.... | 14:08 |
jroll | dtantsur: haven't seen that, maybe racing with the thing that starts up lvm or mdraid? | 14:08 |
*** sacharya has quit IRC | 14:08 | |
dtantsur | I think that's hardware RAID.. investigating moar | 14:08 |
jlvillal | jroll, Ah. Yeah that can be annoying on how git works with the date of when the patch was made, not when it merges :( | 14:09 |
dtantsur | just don't tell me it's one more DIB fanciness | 14:09 |
sambetts | dtantsur: heh, I wonder XD | 14:10 |
NobodyCam | morning dtantsur :) | 14:11 |
dtantsur | morning NobodyCam, jlvillal | 14:13 |
dtantsur | aha, wonderful. the RAID driver gets loaded after IPA runs >_< | 14:13 |
dtantsur | don't tell me we need one more wait loop......... | 14:13 |
*** alexpil__ has quit IRC | 14:13 | |
sambetts | dtantsur: isn't there a way to make literally the last thing that runs as part of boot up? | 14:14 |
dtantsur | I wonder | 14:15 |
dtantsur | damned DIB y u so complex | 14:15 |
*** links has joined #openstack-ironic | 14:15 | |
sambetts | dtantsur: I'd expect that to be the same issue with a non-DIB ramdisk right? | 14:15 |
openstackgerrit | Merged openstack/ironic: Doc: Replace nova image-list https://review.openstack.org/316796 | 14:15 |
*** jaoh has joined #openstack-ironic | 14:16 | |
dtantsur | sambetts, dunno.... | 14:16 |
sambetts | dtantsur: hmm, so in the coreos build we inject all the services as part of the cloud-config.yaml :/ | 14:19 |
*** Mr_T has quit IRC | 14:23 | |
mgould | jroll, did you see the git command I posted earlier? not pretty, but should do what you want | 14:23 |
jroll | mgould: I did, thank you | 14:24 |
*** electrofelix has quit IRC | 14:24 | |
* dtantsur thinks of inserting 'sleep 120' in the IPA systemd unit in the DIB build | 14:26 | |
* dtantsur is nearly serious about that | 14:26 | |
lucasagomes | dtantsur, lol gosh heh | 14:26 |
sambetts | XD | 14:27 |
dtantsur | lemme put it straight: in the DIB build IPA starts when half of the system is not initialized yet | 14:27 |
lucasagomes | dtantsur, https://www.freedesktop.org/software/systemd/man/modules-load.d.html ? | 14:27 |
dtantsur | lucasagomes, the module IS loaded, it's just loaded half way through inspection >_< | 14:27 |
* sambetts is pretty sure thats the reason your having issues with the sysfs networking stuff too | 14:28 | |
* dtantsur is pretty sure too | 14:28 | |
sambetts | dtantsur: did you see my latest commit to update the service files? | 14:29 |
dtantsur | nope, not yet | 14:29 |
lucasagomes | dtantsur, right, so maybe we should add an "After: systemd-modules-load.service" ? | 14:29 |
*** Mr_T has joined #openstack-ironic | 14:29 | |
jroll | After: EVERYTHING ELSE SRSLY | 14:29 |
dtantsur | jroll++ I would love it | 14:29 |
*** vmud213 has quit IRC | 14:31 | |
dtantsur | does not look possible.. https://www.reddit.com/r/linuxquestions/comments/1lkf21/start_systemd_service_after_everything_else/ | 14:32 |
lucasagomes | yeah, wonder if there's a way to guarantee that | 14:32 |
lucasagomes | cause systemd only dispatch stuff | 14:32 |
lucasagomes | so you can say "after you've dispatched everything" but I don't think it does guarantee that things are loaded/running already | 14:33 |
* lucasagomes think we should experiment the After/Requires systemd-modules... | 14:33 | |
dtantsur | yeah, I think we're asking too much of an init system ;) | 14:35 |
*** electrofelix has joined #openstack-ironic | 14:35 | |
dtantsur | like if we would require an init system to have a full-featured logging daemon | 14:35 |
* dtantsur is too sarcastic today probably | 14:35 | |
jroll | lol | 14:35 |
*** rajinir has joined #openstack-ironic | 14:37 | |
* lucasagomes remembers people raging against systemd in the memo-list | 14:40 | |
* lucasagomes memo-list is an internal Red Hat ML | 14:40 | |
dtantsur | lucasagomes, systemd-modules-load.service might not be what we're looking for. it will fire module loading for sure, but I don't think it's synchronous | 14:40 |
dtantsur | lucasagomes, don't get me wrong, I like systemd more than all the other alternatives :) | 14:41 |
lucasagomes | me too | 14:41 |
lucasagomes | and I like journald as well, I think it's great | 14:41 |
lucasagomes | dtantsur, yeah I think systemd will only guarantee that it was dispatched :-/ | 14:41 |
lucasagomes | dtantsur, alternatives to a sleep would be perhaps checking if the module is loaded in the code? | 14:42 |
*** dprince has quit IRC | 14:42 | |
dtantsur | lucasagomes, the module is loaded, I'm afraid. it's just not fully initialized. I'm not sure how it's represented in lsmod | 14:42 |
*** cdearborn has joined #openstack-ironic | 14:43 | |
* lucasagomes don't know either | 14:43 | |
lucasagomes | don't think there's an interface for it in user-space ;-/ | 14:44 |
*** [1]cdearborn has joined #openstack-ironic | 14:45 | |
dtantsur | so, loop'n'wait again? :( I wonder if we need to create a loop for memory_mb.. y'know, just in case | 14:46 |
lucasagomes | :-/ yeah | 14:47 |
jroll | dtantsur: wait for what, exactly, what if only some disks are in the raid | 14:48 |
jroll | :) | 14:48 |
dtantsur | jroll, wait for at least one disk? yeah, does not sound robust, but I have no clues what else I can do... | 14:48 |
dtantsur | maybe my idea with sleep 120 is not so bad..... | 14:49 |
jroll | dtantsur: idk, I'd honestly start with emailing the systemd list and ask if there's a thing, and if not, if people care to add a thing | 14:49 |
*** amotoki has joined #openstack-ironic | 14:49 | |
dtantsur | yeah, maybe... | 14:50 |
*** xhku has joined #openstack-ironic | 14:50 | |
*** ChrisAusten has quit IRC | 14:50 | |
dtantsur | though I've already found network-online target in their docs.. and then found that it does not work for u | 14:51 |
dtantsur | * for us | 14:51 |
jroll | :| | 14:51 |
sambetts | dtantsur: maybe we need to make it so its After:multi-user.target? | 14:54 |
dtantsur | sounds interesting, maybe | 14:54 |
*** superdan is now known as dansmith | 14:56 | |
dtantsur | looks promising: http://unix.stackexchange.com/a/61417 | 14:58 |
dtantsur | still, it does not guarantee that the module is *actually* initialized | 14:58 |
*** kcalman has joined #openstack-ironic | 14:59 | |
*** kcalman has quit IRC | 14:59 | |
sambetts | no, but unless we have checks in IPA start up that check and then exit 1 if we can't gurantee that | 15:00 |
*** itamarl has quit IRC | 15:01 | |
*** cdearborn has quit IRC | 15:02 | |
*** ChrisAusten has joined #openstack-ironic | 15:03 | |
*** anush has quit IRC | 15:04 | |
*** lennyb has joined #openstack-ironic | 15:04 | |
*** yarkot has quit IRC | 15:05 | |
* mgould wonders why this is so hard | 15:06 | |
mgould | is it because systemd only tracks "this service has been started" and not "this service has finished initializing, and now accepts commands"? | 15:06 |
dtantsur | mgould, s/service/kernel module/ now I think you see the problem | 15:07 |
mgould | 'cos "only start X once Y is ready" seems like core functionality | 15:07 |
*** yarkot has joined #openstack-ironic | 15:07 | |
dtantsur | also, reading systemd documentation, they prefer folks to listen to events instead of polling | 15:07 |
dtantsur | which makes sense, except for during inspection we don't know anything about devices and how much we're supposed to wait for notifications | 15:07 |
dtantsur | (and which notifications) | 15:07 |
jlvillal | Good morning dtantsur mgould sambetts jroll lucasagomes NobodyCam rloo TheJulia and everyone else I missed seeing in the backscroll | 15:08 |
dtantsur | morning jlvillal | 15:08 |
mgould | dtantsur, not sure I follow: I guess I don't know enough about how kernel modules are initialized | 15:08 |
mgould | morning jlvillal | 15:08 |
NobodyCam | good morning jlvillal :) | 15:08 |
sambetts | dtantsur: I think After=multi-user.target will get us as late in the boot process as possible right? | 15:08 |
dtantsur | sambetts, yes, but not necessary late enough | 15:08 |
rloo | morning jlvillal | 15:09 |
sambetts | :/ the system should be pretty stable before multi-user.target is marked as complete right? | 15:09 |
dtantsur | the bug is public btw, you can take a look https://bugzilla.redhat.com/show_bug.cgi?id=1328585#c21 | 15:10 |
openstack | bugzilla.redhat.com bug 1328585 in openstack-ironic-inspector "Introspection errors with 'The following required parameters are missing: ['local_gb']'. Passes on readding nodes and rerunning introspection." [Unspecified,New] - Assigned to dtantsur | 15:10 |
*** joprovost has joined #openstack-ironic | 15:11 | |
*** keedya has joined #openstack-ironic | 15:12 | |
*** keedya has quit IRC | 15:13 | |
jlvillal | mgould: Have time to chat about grenade? | 15:14 |
*** keedya has joined #openstack-ironic | 15:15 | |
dtantsur | sambetts, yeah, worth experimenting, I guess | 15:15 |
lucasagomes | jlvillal, morning :-) | 15:16 |
mgould | jlvillal, sure | 15:18 |
*** vmud213 has joined #openstack-ironic | 15:19 | |
jlvillal | mgould: So jroll figured out that the issue is that a new network is created (not private) during the resource phase create. | 15:19 |
jlvillal | And thus the conductor can't communicate with the node's IPA instance. | 15:20 |
jlvillal | An 'early_create' phase has now been added to Grenade. And the Neutron stuff has been moved there. | 15:20 |
jlvillal | mgould: Then I created: https://review.openstack.org/#/c/316234/ to leverage the 'early_create' phase stuff | 15:20 |
*** sacharya has joined #openstack-ironic | 15:21 | |
jlvillal | But I'm still failing... | 15:21 |
mgould | jlvillal, cool! | 15:21 |
mgould | about the progress, not the "still failing", obviously | 15:21 |
jlvillal | So I have the VM up and I am trying to run things manually and try to figure out why the network isn't seen. | 15:21 |
mgould | gotcha | 15:21 |
jroll | jlvillal: where/how are you failing there? | 15:21 |
jlvillal | Let me copy some log files to a public location. | 15:22 |
*** sdake has joined #openstack-ironic | 15:22 | |
sdake | devananda ping | 15:22 |
sdake | re kolla + bifrost | 15:22 |
jroll | oh, I'd hoped you were running in the gate | 15:22 |
jlvillal | It seems similar as before. Though I am seeing an iscsi error, which I don't remember before in the conductor log | 15:22 |
jroll | jlvillal: orly, so the node boots the ramdisk but then we have connection issues with iscsi? | 15:23 |
jlvillal | Seems like it. Let me find my log file | 15:23 |
sdake | Is Julia the indvidiuals name that leads bifrost? | 15:23 |
mgould | sdake, yes | 15:23 |
sdake | if so, what is her irc nick? anyone know? | 15:24 |
mgould | she's TheJulia here | 15:24 |
sdake | cool thanks | 15:24 |
rloo | sdake: she's away for another week i think | 15:24 |
openstackgerrit | Merged openstack/ironic: Regenerate sample config https://review.openstack.org/316849 | 15:24 |
sdake | rloo oh bummer ok thanks :) | 15:24 |
cinerama | hi sdake | 15:24 |
sdake | hey cinerama | 15:25 |
sdake | did you see sean's email? | 15:25 |
sambetts | jlvillal: link to the logs?? I'm seeing a iscsi error in my third party CI too | 15:25 |
sdake | Looks like he has a rough prototype - wondering what should be next steps on ironic side of things | 15:25 |
sambetts | jlvillal: wonder if its the same | 15:25 |
cinerama | sdake: subject? i'll have a look soon | 15:25 |
sdake | cinerama [openstack-dev] [kolla] [bifrost] bifrost container poc update | 15:26 |
sdake | tia ;) | 15:26 |
sdake | cinerama if you could provide sean some guidance, i'd super appreciate it | 15:26 |
sdake | I had alook ofer his patch, and it doesn't look too invasive ot me | 15:26 |
cinerama | cool | 15:26 |
sdake | although it definately needs restructring | 15:26 |
jroll | sdake: so the goal with this is to use bifrost to kick a bunch of baremetal, to deploy openstack on? | 15:26 |
jlvillal | jroll: sambetts: http://www.jirantis.com/grenade-logs/2016-05-16/ | 15:27 |
jlvillal | mgould: ^^ | 15:27 |
jroll | jlvillal: 403 on http://www.jirantis.com/grenade-logs/2016-05-16/grenade.sh.txt | 15:27 |
mgould | jlvillal, thanks | 15:27 |
jlvillal | :9 | 15:27 |
jlvillal | :( | 15:27 |
jroll | jlvillal: console logs might be useful too | 15:27 |
jlvillal | One moment | 15:27 |
mgould | jroll, sdake, that sounds awfully like tripleo... | 15:27 |
jroll | mgould: :) | 15:28 |
sdake | jroll right | 15:28 |
*** amotoki has quit IRC | 15:28 | |
sambetts | mgould: JAOI (just another openstack installer) heh | 15:28 |
mgould | how many are we up to now? | 15:28 |
jroll | sdake: right on, thanks | 15:28 |
sdake | kolla does more then innstall | 15:28 |
sdake | kolla also responsible for past day-0 deployment mangement ;) | 15:28 |
jlvillal | jroll: fixed | 15:29 |
jlvillal | sambetts: mgould fixed | 15:29 |
jroll | sdake: as in, just upgrades, or operations as well | 15:29 |
jroll | jlvillal: thanks | 15:29 |
cinerama | sdake: oh right, i looked at this patch the other day. so this patch splits the install and config stuff out into includes. i was wondering if we should instead split the install process into multiple roles along those lines instead | 15:29 |
jlvillal | jroll: console logs are in old/ironic-bm-logs/ | 15:29 |
sdake | jroll atm upgrade and reconfigure a runnign environment from single source of truth | 15:29 |
jroll | jlvillal: oh duh | 15:29 |
jroll | sdake: so as deployment tools should :) | 15:29 |
sdake | cinerama i think separate roles makes sense to me but might be more invasive too | 15:29 |
jlvillal | jroll: Well when I first did there were NO directories displayed, due to permission issues :( | 15:30 |
sdake | jroll we are adding more to come ;) | 15:30 |
jroll | jlvillal: aha | 15:30 |
jlvillal | So my mistake, not yours | 15:30 |
cinerama | sdake: main disadvantage to doing it as roles is that we'd have some amount of duplication for setting defaults e.g. the OS-specific things | 15:30 |
*** keedya_ has joined #openstack-ironic | 15:30 | |
jlvillal | mgould: Any updates from your end? | 15:30 |
sdake | cinerama ya that is annoying aspect of ansible roles... | 15:30 |
mgould | jlvillal, nope, sorry | 15:30 |
cinerama | sdake: well, either way it's going to be a big change. | 15:31 |
jlvillal | mgould: Okay. Thanks | 15:31 |
* mgould is not pulling his weight here :-( | 15:31 | |
sdake | cinerama to Sean's quesiton though, does Ironic require specs for this sort of work? | 15:31 |
mgould | sdake, tripleo does some post-deployment stuff: http://docs.openstack.org/developer/tripleo-docs/post_deployment/post_deployment.html | 15:31 |
* jlvillal is reminded his weight is 100 kilos ;) | 15:31 | |
cinerama | sdake: so bifrost do not currently do specs, what we do do though is have launchpad blueprints. additionally this is the first time we've really had a big cross-project initiative like this | 15:32 |
sambetts | jlvillal, jroll: the BM log 0 seems fine, I see a bunch of weird errors in the 1 and 2 logs thoughs | 15:32 |
sdake | cinerama ack on the first time thing | 15:32 |
mgould | jlvillal, :-) | 15:32 |
cinerama | sdake: which is awesome :) but will just need to figure stuff out | 15:32 |
sdake | mgould ya - i suspect most deployment management tools do post deployment operations | 15:33 |
jroll | jlvillal: aha, so it can't reach the agent over http to tell the agent to start up the iscsi thing | 15:33 |
jroll | sambetts: I suspect the first one was the pre-resource-create thing | 15:33 |
jroll | jlvillal: I suspect you need the rule for port 9999 here https://review.openstack.org/#/c/315268/1/projects/60_nova/resources.sh | 15:33 |
sdake | cinerama cool, well I don't want Sean to be blocked | 15:33 |
*** keedya has quit IRC | 15:33 | |
sdake | cinerama any chance you can respond on Sean's thread and let him know next steps for Ironic? | 15:34 |
jlvillal | jroll: Let me look | 15:34 |
*** harlowja_at_home has joined #openstack-ironic | 15:34 | |
*** keedya has joined #openstack-ironic | 15:35 | |
jlvillal | jroll: I think I am doing that: https://github.com/JohnVillalovos/devstack-gate-test/blob/master/ansible/roles/devstack-gate/files/update-projects.sh#L24 | 15:35 |
jlvillal | I stole your patch... | 15:35 |
johnthetubaguy | jroll: bonded nics, do we have specs up for that with ironic+nova? | 15:35 |
jroll | jlvillal: well then I guess it isn't working, like vsaienko1 mentioned :| | 15:36 |
johnthetubaguy | jroll: my reason for asking is this spec, for the same thing with virt: https://review.openstack.org/#/c/182242/14/specs/newton/approved/user-controlled-sriov-ports-allocation.rst | 15:36 |
cinerama | sdake: so what specifically does sean want from our end? just process? | 15:36 |
jroll | whyyyyyyyyyyy | 15:36 |
*** rama_y has joined #openstack-ironic | 15:36 | |
jroll | johnthetubaguy: sambetts has https://review.openstack.org/#/c/277853/ | 15:36 |
johnthetubaguy | jroll: ah, got it, for the ironic side at least | 15:37 |
jroll | johnthetubaguy: yeah, nothing in nova (yet) | 15:37 |
sambetts | johnthetubaguy, jroll: https://review.openstack.org/#/c/237067 | 15:37 |
johnthetubaguy | yeah, not sure that includes the bonding stuff though | 15:37 |
jroll | oh, right | 15:37 |
cinerama | sdake: from where i sit i think it's fine to just propose the change as a review in gerrit | 15:37 |
jroll | johnthetubaguy: so that does bonding in the sense of LAG, I think you want trunking though | 15:37 |
jroll | johnthetubaguy: and/or bond config being done in the instance | 15:38 |
*** keedya_ has quit IRC | 15:38 | |
johnthetubaguy | jroll: possibly... anyways, thanks sambetts + jroll, that saved my an age of searching :) | 15:38 |
johnthetubaguy | jroll: yeah, bond inside the instance, is what I was thinking about | 15:38 |
jroll | johnthetubaguy: cool, no problem | 15:39 |
jroll | jlvillal: this may be relevant to your thing https://review.openstack.org/#/c/316703/4/projects/50_neutron/resources.sh | 15:39 |
*** keedya has quit IRC | 15:40 | |
sambetts | johnthetubaguy, jroll https://specs.openstack.org/openstack/nova-specs/specs/liberty/implemented/metadata-service-network-info.html | 15:40 |
jlvillal | jroll: Thanks. I'll investigate that. | 15:40 |
*** keedya has joined #openstack-ironic | 15:40 | |
sambetts | jroll: sigh I'd hoped we'd work this out at the summit but if virt are doign it we need to push neutron to model bonding ... | 15:41 |
sdake | cinerama cool thanks - should sean file la blueprint first then? | 15:43 |
jroll | sambetts: yeah, seems like it | 15:43 |
sambetts | jroll: if thats the case, we shouldn't merge the portgroups API because that will make it a nightmare for us right? | 15:43 |
*** dmk0202 has quit IRC | 15:44 | |
jlvillal | vsaienko1: ping? | 15:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove backward compat for Liberty cleaning https://review.openstack.org/316827 | 15:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [conductor]/clean_nodes config option https://review.openstack.org/316826 | 15:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "message" attribute support from IronicException https://review.openstack.org/316825 | 15:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "periodic_interval" config option https://review.openstack.org/316824 | 15:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated driver_periodic_task https://review.openstack.org/316820 | 15:45 |
vsaienko1 | jlvillal o/ | 15:45 |
jroll | sambetts: will it? we still need to represent them somehow, right? (for physical infra) | 15:45 |
jlvillal | vsaienko1: Looking at: https://review.openstack.org/#/c/316703/ | 15:45 |
jlvillal | vsaienko1: What do you think about adding that to: https://review.openstack.org/#/c/316234/ | 15:45 |
jlvillal | vsaienko1: Putting it into the Ironic code base, and not Grenade directly? | 15:46 |
sambetts | jroll: if they are modeled in neutron, then we'll just create 2 neutron ports and bond them together right? we wouldn't need to model them in Ironic any more | 15:46 |
jlvillal | vsaienko1: I can do it, but wanted to get your input. | 15:46 |
jlvillal | vsaienko1: Or maybe not possible... | 15:47 |
jroll | sambetts: oh, right, because neutron has a bonding model, hmm | 15:47 |
*** Sukhdev has joined #openstack-ironic | 15:48 | |
sambetts | yeah, it would be just down to the mech driver to support setting it up, when we port-update | 15:48 |
jroll | right | 15:49 |
jlvillal | vsaienko1: ??? | 15:51 |
vsaienko1 | lvillal: I'm checking it at openstack-qa channel | 15:51 |
jlvillal | vsaienko1: Ah | 15:51 |
*** hshiina has joined #openstack-ironic | 15:54 | |
*** MattMan has quit IRC | 15:55 | |
*** MattMan has joined #openstack-ironic | 15:55 | |
*** keedya_ has joined #openstack-ironic | 15:55 | |
openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: [WIP] Tempest: add basic test https://review.openstack.org/293926 | 15:56 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/316936 | 15:56 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-lib: Updated from global requirements https://review.openstack.org/316937 | 15:56 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/316938 | 15:56 |
*** thiagop has joined #openstack-ironic | 15:56 | |
*** vishwanathj has joined #openstack-ironic | 15:56 | |
jlvillal | jroll: Not sure what you think of: https://review.openstack.org/#/c/316234/ | 15:57 |
jlvillal | I know it is similar to your patch. | 15:57 |
jlvillal | I was just trying to minimize code duplication with mine. | 15:57 |
thiagop | Good morning, Ironicers | 15:57 |
mgould | morning thiagop | 15:57 |
NobodyCam | hey hey thiagop good morning... | 15:57 |
jroll | jlvillal: yeah, totally fine, I never mind people picking up my work :) | 15:58 |
jlvillal | :) | 15:58 |
*** keedya has quit IRC | 15:58 | |
jlvillal | vsaienko1: Let me know if I can help at all | 15:58 |
jlvillal | vsaienko1: I would be very happy to test it!! | 15:58 |
rloo | jroll: easy +2 for you I think: https://review.openstack.org/#/c/312608 | 16:00 |
rloo | jroll, who are the release folks that can +2, you, dtantsur, devananda? Is there a link that shows that info? | 16:00 |
* jlvillal feels like this Monday morning is kind of hectic... | 16:00 | |
*** keedya__ has joined #openstack-ironic | 16:00 | |
rloo | jlvillal: more so than normal? | 16:01 |
dtantsur | rloo, https://review.openstack.org/#/admin/groups/950,members | 16:01 |
jlvillal | rloo: Feels like I have been in multiple IRC conversations and trying to code at the same time.. Trying to keep up :) | 16:02 |
rloo | thx dtantsur. Am going to update our wiki with that. | 16:02 |
rloo | jlvillal: OH. 'multitasking' :( | 16:02 |
rloo | jlvillal: just means that you're in demand ;) | 16:02 |
*** moshele has quit IRC | 16:02 | |
jroll | rloo: to find that link, click the gear next to the project name in gerrit, then 'access' at the top | 16:03 |
*** keedya_ has quit IRC | 16:03 | |
rloo | dtantsur: the stable maint folks are for all of the projects under ironic umbrella right? | 16:03 |
rloo | jroll: thx, that is good to know. | 16:04 |
dtantsur | rloo, not necessary, I think subprojects have different procedures | 16:04 |
dtantsur | but definitely true for ironic, ironicclient, ironic-lib and IPA | 16:04 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 16:05 |
*** keedya has joined #openstack-ironic | 16:05 | |
jlvillal | rloo: Luckily I'm not in demand that much. | 16:06 |
*** earlephilhower has quit IRC | 16:06 | |
*** fragatina has quit IRC | 16:07 | |
*** fragatina has joined #openstack-ironic | 16:08 | |
*** keedya__ has quit IRC | 16:08 | |
jlvillal | sambetts: Hopefully Cisco CI out of control on May 10th was fixed :) https://review.openstack.org/#/c/311109/ | 16:09 |
jlvillal | Holy cow! | 16:09 |
jlvillal | ;) | 16:09 |
rloo | jlvillal: let us know if you would like us to change it so that being in demand is the norm ;) | 16:10 |
sambetts | jlvillal: oooooh god?! I have no idea whats happened there!!! | 16:10 |
jlvillal | rloo: -1 :P | 16:10 |
jlvillal | sambetts: Yeah, I kept hitting page down and going wow :) | 16:10 |
sambetts | jlvillal: yeah I don't know why that happened... | 16:10 |
sambetts | sigh... | 16:10 |
rloo | jlvillal, sambetts: yeah, I saw that last week. sorry, maybe i shouldn't have ignored it but figured you knew. | 16:11 |
*** keedya_ has joined #openstack-ironic | 16:11 | |
lucasagomes | jlvillal, that's a lot of depends-on :-/ | 16:13 |
*** keedya has quit IRC | 16:15 | |
jlvillal | vsaienko1: Do you think it is NOT needed to call 'create_ovs_taps'? https://review.openstack.org/#/c/316234/ | 16:22 |
jlvillal | vsaienko1: Sorry I meant: https://review.openstack.org/#/c/311109/15/devstack/upgrade/resources.sh | 16:23 |
*** e0ne has quit IRC | 16:25 | |
*** keedya__ has joined #openstack-ironic | 16:26 | |
vsaienko1 | jlvillal looks like needed | 16:28 |
*** derekh has quit IRC | 16:28 | |
mgould | jlvillal, what do we need to update in https://etherpad.openstack.org/p/ironic-newton-grenade-whiteboard ? | 16:29 |
*** keedya_ has quit IRC | 16:30 | |
openstackgerrit | Merged openstack/ironic-python-agent: Make TinyCore mirror configurable https://review.openstack.org/315649 | 16:30 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 16:31 |
vsaienko1 | jlvillal, yes taps are needed, without them VMs will not have an access to new network | 16:31 |
*** keedya has joined #openstack-ironic | 16:32 | |
openstackgerrit | Hugo Nicodemos proposed openstack/python-oneviewclient: Adding dynamic allocation operations https://review.openstack.org/258699 | 16:33 |
rloo | lucasagomes: when you have a chance, you had some comments, so want to make sure you are OK with this: https://review.openstack.org/#/c/302403/ | 16:33 |
*** keedya__ has quit IRC | 16:34 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/316938 | 16:36 |
*** jaoh has quit IRC | 16:37 | |
* mgould -> relocate | 16:37 | |
*** mgould has quit IRC | 16:37 | |
*** vmud213 has quit IRC | 16:37 | |
*** vmud213 has joined #openstack-ironic | 16:38 | |
lucasagomes | rloo, hi there, will take a look | 16:40 |
*** keedya_ has joined #openstack-ironic | 16:42 | |
lucasagomes | rloo, +a | 16:44 |
rloo | thx lucasagomes! | 16:44 |
*** trown is now known as trown|lunch | 16:45 | |
*** keedya has quit IRC | 16:45 | |
*** kromanenko_ has quit IRC | 16:46 | |
*** kromanenko__ has quit IRC | 16:47 | |
*** keedya has joined #openstack-ironic | 16:47 | |
*** links has quit IRC | 16:47 | |
*** lekha is now known as lekha-away | 16:48 | |
*** keedya_ has quit IRC | 16:49 | |
*** ChrisAusten has quit IRC | 16:49 | |
jlvillal | vsaienko1: On the patch. Why do we need a new network? https://review.openstack.org/#/c/311109/16/devstack/upgrade/resources.sh | 16:50 |
jlvillal | vsaienko1: Just wondering if we can use the already existing network and just update it? | 16:50 |
vsaienko1 | jlvillal: do you mean to use existing network that is created by neutron resources? or the 'private' network that was created during stack.sh? | 16:51 |
jlvillal | vsaienko1: The neutron resources one | 16:51 |
jlvillal | Since that is what nova will be using. I think. | 16:52 |
vsaienko1 | this network has the same CIDR as network that was created during devstack installation. I doubt that it is possible to setup routing to overlapping networks | 16:52 |
jlvillal | vsaienko1: Hmm. Okay. I will need to investigate more and learn more then. | 16:53 |
devananda | moring, all | 16:53 |
vsaienko1 | during upgrade instances that were started in old network should be still available right? | 16:53 |
*** harlowja_at_home has quit IRC | 16:53 | |
*** blakec has joined #openstack-ironic | 16:54 | |
vsaienko1 | Morning devananda | 16:54 |
jlvillal | vsaienko1: I believe so. | 16:54 |
jlvillal | devananda: Good morning | 16:54 |
jlvillal | vsaienko1: Interested to see how your patch does in the gate :) | 16:54 |
vsaienko1 | jlvillal so we need to create resources in new network | 16:54 |
vsaienko1 | me too :) | 16:55 |
* jlvillal remembers he needs to update the whiteboard | 16:55 | |
vsaienko1 | jlvillal, btv I have a problem. I can't find a way how to pass OS_TEST_TIMEOUT to grenade job | 16:55 |
jroll | vsaienko1: tempest.conf | 16:56 |
* jlvillal was about to say that jroll probably knows... | 16:56 | |
*** keedya has quit IRC | 16:56 | |
vsaienko1 | https://review.openstack.org/#/c/315422/8/.testr.conf | 16:56 |
jroll | er, no I lied | 16:56 |
vsaienko1 | jroll the variable should be exported before running tox ... | 16:56 |
jroll | vsaienko1: right, yep | 16:56 |
*** keedya has joined #openstack-ironic | 16:57 | |
jroll | that doesn't work | 16:57 |
jroll | ? | 16:57 |
jroll | I've had success with: OS_TEST_TIMEOUT=foo tox -e... | 16:57 |
vsaienko1 | that works, but it is hack | 16:57 |
jroll | well, that's how it's meant to be run afaik | 16:57 |
rloo | 3 minute warning | 16:57 |
*** mgould has joined #openstack-ironic | 16:57 | |
jroll | exporting it in devstack-gate or project-config should work no? | 16:57 |
vsaienko1 | I have tried https://review.openstack.org/#/c/316662 and https://review.openstack.org/#/c/316055/ but no luck | 16:58 |
jroll | vsaienko1: oh, sudo doesn't pass through environment variables | 16:58 |
jroll | see sudo -E | 16:59 |
vsaienko1 | but the tests are launched from stack https://github.com/openstack-dev/grenade/blob/master/grenade.sh#L248 | 16:59 |
*** sivaramakrishna has joined #openstack-ironic | 17:00 | |
jroll | meeting time | 17:00 |
*** clif has joined #openstack-ironic | 17:00 | |
*** clif is now known as clif_h | 17:00 | |
*** keedya has quit IRC | 17:01 | |
*** keedya has joined #openstack-ironic | 17:01 | |
*** fragatina has quit IRC | 17:02 | |
*** jistr has quit IRC | 17:02 | |
jroll | Sukhdev: can you drop the links you mentioned here, and what you want me to check out? running the ironic meeting right now but don't want to forget | 17:03 |
jroll | or hshiina ^ | 17:04 |
Sukhdev | jroll : here you go - https://review.openstack.org/297895 and https://review.openstack.org/206163 | 17:04 |
lucasagomes | jlvillal no virtual nor physical ? | 17:04 |
lucasagomes | (re midcycle) | 17:04 |
jroll | Sukhdev: what do I need to "check out" here? | 17:05 |
jlvillal | lucasagomes: I wouldn't be able to attend the physical midcycle as I would be on vacation for when it was planned. | 17:05 |
lucasagomes | jlvillal, ah, fair enough. Hope the virtual will work for ya | 17:05 |
*** ohamada_ has joined #openstack-ironic | 17:05 | |
*** ohamada__ has quit IRC | 17:05 | |
Sukhdev | jroll : hshiina wants to confirm that those are the only two nova patches that need to merge with the other 4 critical ironic patches | 17:05 |
jroll | Sukhdev: yes | 17:06 |
Sukhdev | hshiina : I hope you got your answer | 17:06 |
*** keedya_ has joined #openstack-ironic | 17:06 | |
hshiina | Sukhdev, jroll, thanks. | 17:09 |
*** keedya has quit IRC | 17:09 | |
jlvillal | lucasagomes: The initial plan was physical at same time and location as Nova midcycle. I am on vacation that week. | 17:11 |
*** ohamada_ has quit IRC | 17:11 | |
*** Sukhdev has quit IRC | 17:11 | |
*** ohamada_ has joined #openstack-ironic | 17:11 | |
jlvillal | lucasagomes: So hopefully virtual will be when I am not on vacation :) | 17:11 |
lucasagomes | ++ totally | 17:12 |
jlvillal | lucasagomes: I would actually be up for a virtual sooner rather than later. | 17:12 |
lucasagomes | jlvillal, yeah not sure what the dates jroll has in mind, I believe we will talk about it soon-ish | 17:12 |
*** david-lyle_ has quit IRC | 17:12 | |
lucasagomes | the sooner we decide the better | 17:13 |
*** david-lyle has joined #openstack-ironic | 17:13 | |
*** ohamada_ has quit IRC | 17:21 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for dib based agent ramdisk in lib/ironic https://review.openstack.org/264579 | 17:21 |
openstackgerrit | Merged openstack/ironic: Extend IRONIC_RAMDISK_TYPE to support 'dib' https://review.openstack.org/302403 | 17:22 |
*** sivaramakrishna has quit IRC | 17:23 | |
*** trusted has joined #openstack-ironic | 17:24 | |
*** dims_ has joined #openstack-ironic | 17:25 | |
*** dims has quit IRC | 17:26 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 17:27 |
*** sinval has joined #openstack-ironic | 17:28 | |
sinval | morning folks | 17:28 |
*** piet has joined #openstack-ironic | 17:29 | |
*** jcoufal has quit IRC | 17:29 | |
*** joprovost has quit IRC | 17:30 | |
*** joprovost has joined #openstack-ironic | 17:30 | |
*** joprovost has quit IRC | 17:32 | |
*** joprovost has joined #openstack-ironic | 17:32 | |
*** keedya_ has quit IRC | 17:32 | |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Update devstack plugin with new auth options https://review.openstack.org/297312 | 17:33 |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Migrate to using keystoneauth Sessions https://review.openstack.org/236982 | 17:33 |
*** joprovost has quit IRC | 17:36 | |
*** joprovost has joined #openstack-ironic | 17:36 | |
*** vmud213 has quit IRC | 17:37 | |
*** hshiina has quit IRC | 17:37 | |
*** electrofelix has quit IRC | 17:38 | |
*** trusted has quit IRC | 17:38 | |
*** keedya has joined #openstack-ironic | 17:39 | |
*** trusted has joined #openstack-ironic | 17:39 | |
*** keedya has quit IRC | 17:41 | |
*** keedya has joined #openstack-ironic | 17:41 | |
dtantsur | devananda, how would you create a nova instance for an adopted node (asking here not to side-track the discussion) | 17:46 |
*** trown|lunch is now known as trown | 17:46 | |
*** rcernin has quit IRC | 17:47 | |
dtantsur | mat128, you're asking the same question, I guess ^^? | 17:48 |
*** e0ne has joined #openstack-ironic | 17:52 | |
*** baoli has quit IRC | 17:54 | |
*** baoli has joined #openstack-ironic | 17:55 | |
mgould | I've just read the microversions spec, and I don't see why my suggestion is "totally against the idea behind microversions" | 17:56 |
mgould | see use cases 5 and 6 in https://specs.openstack.org/openstack/ironic-specs/specs/kilo-implemented/api-microversions.html | 17:57 |
mgould | if the client can't handle the truth, it should fail; but a new enough client should handle all possible microversions that the server might support | 17:57 |
mgould | which is what I suggested, I thought... | 17:57 |
mgould | sorry if I didn't express myself clearly enough | 17:58 |
mat128 | dtantsur: yup | 17:58 |
dtantsur | mgould, I guess you're talking about version negotiation, which has been our hot topic for quite some time :) | 17:58 |
mgould | dtantsur, yep | 17:59 |
mgould | but if we can't ever say "go away, you're too old" then we'll keep having the "what lie should we tell?" discussion | 17:59 |
*** baoli has quit IRC | 18:00 | |
mgould | so I'm relieved to see that "go away, you're too old" is allowed by the spec | 18:00 |
lucasagomes | mgould, I think it was because when the microversion was introduced it was done to solve an old client breakage problem | 18:01 |
jroll | rloo: right, the priority of search API changes | 18:01 |
dtantsur | mgould, it is, but in this case we'll just break all old clients (e.g. instantly break the whole tripleo and nova) | 18:01 |
mgould | anyway, I get that this is a long-running argument, but I was surprised how vehement the reaction was | 18:01 |
lucasagomes | so when upgrading Ironic it wouldn't break nova (because nova only understood old states) | 18:01 |
lucasagomes | that was the conception of it | 18:01 |
vdrok | jroll: I have to bug you a bit with launchpad stuff again :( could you close these bps? https://blueprints.launchpad.net/python-ironicclient/+spec/ironicclient-functional-tests https://blueprints.launchpad.net/python-ironicclient/+spec/ironicclient-negative-functional-tests | 18:01 |
vdrok | https://blueprints.launchpad.net/python-ironicclient/+spec/ironic-client-advanced-functional-tests https://blueprints.launchpad.net/python-ironicclient/+spec/py33-support | 18:01 |
NobodyCam | thank you all great meeting | 18:02 |
jroll | mgould: well, your suggestion sounded like "every time there's an incompatibility, raise the minimum version and let everyone else break" | 18:02 |
jroll | mgould: which, at that point, there's no back compat anyway so why version | 18:02 |
vdrok | jroll: I'll talk to Sergii regrding functional tests, to move it to usual bugs | 18:02 |
mgould | ah, OK | 18:02 |
devananda | jroll: after reading our state machine again, I am wondering if hte nova driver is wrong, and so my assessment of the compatibility implications for that patch may have originally been correct | 18:02 |
jroll | devananda: eh? | 18:02 |
mgould | I meant something more like "require operators to update clients before updating ironic-api" | 18:03 |
*** jvgrant__ has quit IRC | 18:03 | |
*** trusted has quit IRC | 18:03 | |
jroll | mgould: yeah, that makes operations super hard imo | 18:03 |
mgould | jroll, and you'd need versioning precisely to detect "my client is now too old" | 18:03 |
jroll | mgould: because that's "find everyone in your company that talks to ironic and make sure they update all of their tooling" | 18:04 |
*** baoli has joined #openstack-ironic | 18:05 | |
mgould | jroll: ... and we can't rely on that to happen automatically because there are clients out there that don't respect the -Min-Version header? | 18:05 |
mgould | I think I see the problem now | 18:05 |
jroll | hehe | 18:05 |
*** fragatina has joined #openstack-ironic | 18:06 | |
* mgould was assuming an incompatible upgrade would be a big deal involving notifying users *anyway* | 18:06 | |
lucasagomes | folks, calling it a day. Have a great evening all | 18:06 |
mgould | "we're updating OpenStack, please upgrade your client software" | 18:06 |
sambetts | o/ lucasagomes | 18:06 |
lucasagomes | see you all tomorrow | 18:06 |
vdrok | night lucasagomes | 18:06 |
NobodyCam | have a good night lucasagomes | 18:06 |
dtantsur | me too, see you | 18:06 |
jroll | mgould: well, that's the thing, it isn't incompatible if the old tooling supports it (and the new tooling uses the older version) | 18:06 |
NobodyCam | night dtantsur | 18:06 |
*** dtantsur is now known as dtantsur|afk | 18:06 | |
*** lucasagomes is now known as lucas-dinner | 18:06 | |
jroll | later lucasagomes, dtantsur|afk \o | 18:07 |
vdrok | and good night dtantsur|afk :) | 18:07 |
*** keedya has quit IRC | 18:07 | |
mgould | jroll, right, sure, but I don't think it's sustainable to support clients with outdated semantics indefinitely | 18:08 |
devananda | jroll: https://github.com/openstack/nova/blob/master/nova/virt/ironic/driver.py#L74 vs. https://github.com/openstack/ironic/blob/master/ironic/common/states.py | 18:08 |
jroll | devananda: did you want to elaborate on what you said? | 18:08 |
jroll | aha | 18:08 |
devananda | jroll: 'delete' is not an allowed target for the DEPLOYING state | 18:09 |
devananda | even though it is listed in Nova's _UNPROVISION_STATES | 18:09 |
jroll | mgould: yeah, we may find ourselves taking a stand on that | 18:09 |
mgould | jroll, cool | 18:09 |
devananda | so I think that's a bug in Nova | 18:09 |
*** jaybeale has quit IRC | 18:09 | |
jroll | devananda: https://github.com/openstack/nova/commit/bfe52542f5449391b9dc152a90fd79afebcb3ff2 | 18:09 |
mgould | supporting outdated *syntax* is pretty low-impact, but outdated semantics are a bigger deal | 18:09 |
jroll | devananda: totally intentional because retries | 18:09 |
vdrok | jroll: oh, this one too - https://blueprints.launchpad.net/python-ironicclient/+spec/improve-baremetal-unittest, when you have time, it was already moved to bug | 18:09 |
devananda | jroll: ok. I thought that might be the case | 18:09 |
jroll | devananda: solves a major bug with orphaning things | 18:10 |
vdrok | good night everyone! | 18:10 |
devananda | jroll: totally makes sense then | 18:10 |
jroll | vdrok: done, good night! | 18:10 |
NobodyCam | night vdrok | 18:10 |
devananda | jroll: however, it also means that the API semantics for making ADOPTING appear like DEPLOYING would be consistent | 18:10 |
* mgould -> dinner; good night everyone! | 18:10 | |
vdrok | jroll: thanks! | 18:10 |
NobodyCam | night mgould | 18:10 |
sambetts | night mgould | 18:10 |
jlvillal | mgould: Ciao! | 18:10 |
*** mgould has quit IRC | 18:10 | |
devananda | jroll: as long as the same error code is returned fromthe API when a(n older) client requests to delete a node in ADOPTING state (it errors, then retries when the node is in ACTIVE state) | 18:11 |
*** jvgrant has joined #openstack-ironic | 18:11 | |
jroll | devananda: sure, I still disagree that we should lie about the state (what other actions/assumptions might someone make when they see a node in DEPLOYING?) | 18:11 |
devananda | jroll: so, I agree with the general principle in the discussion, but in this case, it looks like backwards compat could be maintained | 18:11 |
devananda | because the intent of the ADOPTING state is to import active instances from external systems | 18:12 |
devananda | that said - we can't know how someone else might use the state | 18:12 |
jroll | right, with my ops hat on, a false view of the system is much more dangerous than a poorly written client that might break down | 18:13 |
* sambetts is heading off | 18:14 | |
*** sambetts is now known as sambetts|afk | 18:15 | |
NobodyCam | night sambetts|afk | 18:15 |
devananda | jroll: do you have a bit / want to chat about api versions in general? | 18:15 |
jroll | devananda: sure | 18:18 |
*** jtomasek has quit IRC | 18:19 | |
devananda | jroll: two things I'd like to chat about: 1) raising the requested version in the client 2) as we EOL support for old server versions, do we begin to raise the server's min API version based on our supported server lifespan? | 18:23 |
devananda | I don't expect we'll resolve either right now , but I'd like to understand if you have objections to (a) -- eg, see https://review.openstack.org/#/c/305540/ | 18:23 |
devananda | and what your initial thoughts on (b) are | 18:23 |
devananda | gah. brains. s/a/1/ && s/b/2/ | 18:24 |
devananda | wow, the "music"in this cafe is bad. relocating and back in a few minutes | 18:27 |
*** e0ne has quit IRC | 18:28 | |
*** fragatin_ has joined #openstack-ironic | 18:28 | |
jroll | devananda: on (1), you're going to break people, (2) idk, that's a good question | 18:29 |
jroll | I don't currently have any thoughts on (2), I guess | 18:30 |
*** e0ne has joined #openstack-ironic | 18:31 | |
*** fragatina has quit IRC | 18:32 | |
*** Sukhdev has joined #openstack-ironic | 18:33 | |
*** aNupoisc has joined #openstack-ironic | 18:34 | |
*** aNupoisc has quit IRC | 18:34 | |
*** aNupoisc has joined #openstack-ironic | 18:35 | |
*** e0ne has quit IRC | 18:35 | |
*** penick has joined #openstack-ironic | 18:39 | |
*** keedya has joined #openstack-ironic | 18:41 | |
* devananda is back | 18:42 | |
*** keedya has quit IRC | 18:43 | |
*** keedya has joined #openstack-ironic | 18:43 | |
devananda | jroll: re (1) at what point do you think we should begin exposing any functionality added since 1.6 to users by default? | 18:43 |
*** krtaylor has quit IRC | 18:44 | |
[1]cdearborn | Hey folks, | 18:45 |
devananda | raising the default requested version doesn't prevent users from requesting older versions, if their application depends on a specific version of the server | 18:45 |
jroll | devananda: right, I guess the question is how many things are out there not pinning versions | 18:45 |
jroll | (using the python sdk) | 18:45 |
[1]cdearborn | was looking at the marketplace: https://www.openstack.org/marketplace/drivers/, and noticed that it doesn't list the drac driver. Anyone know how we could get it listed there? | 18:45 |
devananda | jroll: I don't think we can know that | 18:45 |
jroll | right | 18:46 |
jroll | [1]cdearborn: this is on my list to look at + fix this week | 18:46 |
jroll | [1]cdearborn: it's super super out of date | 18:46 |
devananda | jroll: conversely, how many new clients are not passing a specific version header and therefor not seeing any of the new features? | 18:46 |
[1]cdearborn | jroll: thx much! | 18:46 |
jroll | devananda: and how many of them care :) | 18:46 |
jroll | (for the former, how many would actually break) | 18:46 |
devananda | jroll: anyone who wants to use inspection, for example | 18:46 |
jroll | are they not using inspection because they don't see it in the python sdk? | 18:47 |
jroll | fwiw, I really like nova's model of python sdk is at minimum, cli is at maximum | 18:47 |
devananda | or raid config | 18:47 |
devananda | that's interesting | 18:47 |
*** anush_ has joined #openstack-ironic | 18:47 | |
devananda | and yea, that would be fine with me, too | 18:48 |
devananda | (at first thought) | 18:48 |
jroll | the idea being sdk is for computers, and version should always be set, cli is for humans and they'll figure it out | 18:48 |
*** keedya has quit IRC | 18:48 | |
*** keedya has joined #openstack-ironic | 18:48 | |
*** blakec has quit IRC | 18:50 | |
* jroll steps away for a while | 18:56 | |
*** zhihaisong has joined #openstack-ironic | 18:56 | |
* NobodyCam thinks jroll iss getting food of some kind :p | 18:57 | |
*** szhsong has quit IRC | 18:59 | |
*** baoli has quit IRC | 18:59 | |
*** keedya has quit IRC | 19:04 | |
*** keedya has joined #openstack-ironic | 19:04 | |
*** ChrisAusten has joined #openstack-ironic | 19:09 | |
*** keedya_ has joined #openstack-ironic | 19:10 | |
*** ChrisAusten has quit IRC | 19:13 | |
*** keedya has quit IRC | 19:13 | |
*** keedya_ has quit IRC | 19:15 | |
*** ChrisAusten has joined #openstack-ironic | 19:18 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/316936 | 19:19 |
openstackgerrit | Ricardo Araújo Santos proposed openstack/ironic-python-agent: Splits sfdisk output and removes trailing space when partition id is less than 10. https://review.openstack.org/317058 | 19:19 |
openstackgerrit | Sam Betts proposed openstack/ironic-python-agent: TinyIPA: Ensure that kernel modules are loaded at boot https://review.openstack.org/317059 | 19:20 |
*** fragatin_ has quit IRC | 19:23 | |
*** fragatina has joined #openstack-ironic | 19:23 | |
*** jaybeale has joined #openstack-ironic | 19:32 | |
*** lucas-dinner has quit IRC | 19:38 | |
*** e0ne has joined #openstack-ironic | 19:38 | |
*** e0ne has quit IRC | 19:40 | |
*** rcernin has joined #openstack-ironic | 19:42 | |
*** lucasagomes has joined #openstack-ironic | 19:43 | |
*** baoli has joined #openstack-ironic | 19:45 | |
*** baoli has quit IRC | 19:46 | |
*** baoli has joined #openstack-ironic | 19:47 | |
*** xhku has quit IRC | 19:51 | |
*** lucasagomes has quit IRC | 19:51 | |
*** wajdi_ has joined #openstack-ironic | 19:51 | |
*** vsaienko has joined #openstack-ironic | 19:54 | |
*** mbound has joined #openstack-ironic | 19:54 | |
*** _vsaienko has joined #openstack-ironic | 19:54 | |
_vsaienko | jlvillal nova resource tests are passed http://logs.openstack.org/09/311109/16/experimental/gate-grenade-dsvm-partial-ironic/f0412ab/logs/grenade.sh.txt.gz | 19:54 |
jlvillal | vsaienko1: Sweet!!!! | 19:55 |
_vsaienko | jroll, jlvillal we need to enable cinder-api for grenade job or skip singer tests http://logs.openstack.org/09/311109/16/experimental/gate-grenade-dsvm-partial-ironic/f0412ab/logs/grenade.sh.txt.gz | 19:55 |
*** wajdi has quit IRC | 19:55 | |
NobodyCam | jroll: have you seen this: http://logs.openstack.org/20/316820/3/check/gate-ironic-pep8/fe88fcc/console.html#_2016-05-16_16_37_10_674 | 19:55 |
jlvillal | _vsaienko: That is awesome though. I mean we hit another issue. But have hopefully fixed one. | 19:56 |
*** wajdi has joined #openstack-ironic | 19:56 | |
*** lucasagomes has joined #openstack-ironic | 19:57 | |
_vsaienko | jlvillal yes, it is good news. I'm checking cinder volume issues locally probably the easiest way is to enable c-api service | 19:58 |
*** wajdi_ has quit IRC | 19:59 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic: Setup for using the Grenade 'early_create' phase https://review.openstack.org/316234 | 20:00 |
jlvillal | _vsaienko: Or maybe it should not be running if cinder-api is not present? | 20:00 |
*** wajdi has quit IRC | 20:01 | |
klindgren | hello - I am have aptch to fix an issue with tCOREOS ipa image where the image that being used doesn't contain the IPMI drivers for the kernel. https://review.openstack.org/#/c/316306/ | 20:05 |
openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Fix agentless error on M4 systems https://review.openstack.org/317072 | 20:05 |
klindgren | Jenkins failed on it - when I look at the logs its saying out of disk space. I assume this is related to resources available for the vm's in gate - and some nuance around amount of things that can be in memory at once? | 20:06 |
*** baoli has quit IRC | 20:07 | |
*** baoli has joined #openstack-ironic | 20:08 | |
*** krtaylor has joined #openstack-ironic | 20:08 | |
*** baoli has quit IRC | 20:09 | |
*** baoli has joined #openstack-ironic | 20:09 | |
*** piet has quit IRC | 20:10 | |
openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Remove errant IBM nextscale id https://review.openstack.org/317074 | 20:10 |
*** Sukhdev has quit IRC | 20:10 | |
_vsaienko | jlvillal: we need to skip cinder volume test, as booting instance from volume is not supported yet https://github.com/openstack-dev/grenade/blob/master/projects/70_cinder/resources.sh#L121-L126 | 20:10 |
_vsaienko | jlvillal: going to prepare patch to grenade | 20:11 |
jlvillal | jroll: rloo NobodyCam _vsaienko : For the grenade work, I have this patch. I think it is ready for merge. https://review.openstack.org/#/c/316234/ Reviews welcome :) | 20:11 |
jlvillal | _vsaienko: Thanks. Please add me to review if you remember | 20:11 |
*** rama_y has quit IRC | 20:12 | |
*** joprovost has quit IRC | 20:13 | |
NobodyCam | jlvillal: Ack will try and take alook in few | 20:13 |
jlvillal | NobodyCam: You rock :) | 20:13 |
*** dmk0202 has joined #openstack-ironic | 20:14 | |
jlvillal | _vsaienko: Is it a problem that https://review.openstack.org/#/c/311109/ depended on https://review.openstack.org/#/c/316703/ | 20:15 |
*** rama_y has joined #openstack-ironic | 20:19 | |
jroll | _vsaienko: nice work making progress, grenade patch sounds fine to me | 20:20 |
jroll | _vsaienko: I guess just skip the entire 70_cinder file if not enabled | 20:21 |
jroll | NobodyCam: I hadn't seen that til now, thanks for the heads up | 20:21 |
jroll | (and for the review:)) | 20:21 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 20:22 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: WIP: grenade checking https://review.openstack.org/311109 | 20:22 |
_vsaienko | jlvillal it shouldn't be a problem | 20:23 |
_vsaienko | jlvillal, jroll please review https://review.openstack.org/#/c/317076/ | 20:23 |
jlvillal | _vsaienko: Okay. I wasn't sure if you were doing the same thing twice. | 20:23 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/316936 | 20:26 |
*** baoli has quit IRC | 20:29 | |
*** sinval has quit IRC | 20:30 | |
*** xavierr has quit IRC | 20:30 | |
*** gabriel-bezerra has quit IRC | 20:30 | |
*** thiagop has quit IRC | 20:30 | |
*** clenimar has quit IRC | 20:30 | |
*** clenimar has joined #openstack-ironic | 20:32 | |
jlvillal | _vsaienko: +1 from me on the patch | 20:32 |
*** krtaylor has quit IRC | 20:32 | |
*** gabriel-bezerra has joined #openstack-ironic | 20:32 | |
rloo | JayF, jroll, alineb: wrt https://review.openstack.org/#/c/301526/ the metrics patch. I really don't like the API. Do you have a patch/example that uses it so I can understand why you have all these prefix/name parts and why they can be strings/ints/lists/blah blah. | 20:33 |
rloo | and why we need all/most of those configs, vs passing that info when a logger is created? | 20:33 |
*** sinval has joined #openstack-ironic | 20:33 | |
*** thiagop has joined #openstack-ironic | 20:34 | |
*** xavierr has joined #openstack-ironic | 20:34 | |
*** moshele has joined #openstack-ironic | 20:36 | |
*** sdake has quit IRC | 20:37 | |
*** sdake has joined #openstack-ironic | 20:37 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic: Update resources subnet CIDR https://review.openstack.org/317082 | 20:39 |
*** gabriel-bezerra has quit IRC | 20:40 | |
*** gabriel-bezerra has joined #openstack-ironic | 20:41 | |
*** _vsaienko has quit IRC | 20:43 | |
*** vsaienko has quit IRC | 20:43 | |
*** moshele has quit IRC | 20:47 | |
*** CIA has quit IRC | 20:51 | |
jlvillal | jroll: Not sure if it makes difference in your thought process but the resources.sh file is executed by grenade. It is not sourced by grenade. So any code we pull in doesn't bubble up to grenade. | 20:51 |
jlvillal | But if it is better, I can make it a separate file too | 20:52 |
jroll | jlvillal: ah, ok | 20:52 |
* jroll +2 | 20:52 | |
jlvillal | jroll: Thanks :) | 20:53 |
jroll | np | 20:54 |
*** baoli has joined #openstack-ironic | 20:55 | |
*** sabeen has quit IRC | 20:56 | |
*** Sukhdev has joined #openstack-ironic | 21:01 | |
*** trown is now known as trown|outtypewww | 21:03 | |
xavierr | nick xavierr_not_found | 21:04 |
NobodyCam | wow: first time I've seen this error: http://logs.openstack.org/26/316826/3/check/gate-ironic-python27-db/df84926/console.html#_2016-05-16_16_34_34_673 | 21:17 |
*** nicodemos has joined #openstack-ironic | 21:17 | |
jlvillal | NobodyCam: I wish I could say the same.... | 21:17 |
NobodyCam | :( | 21:17 |
jlvillal | NobodyCam: But I see that internally. I don't think I have seen it in the gate. | 21:18 |
jroll | sounds like a bad mirror? | 21:18 |
*** sabeen has joined #openstack-ironic | 21:18 | |
NobodyCam | yea | 21:18 |
*** jaybeale has quit IRC | 21:20 | |
*** jaybeale_ has joined #openstack-ironic | 21:20 | |
*** jaybeale_ has quit IRC | 21:25 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove backward compat for Liberty cleaning https://review.openstack.org/316827 | 21:29 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove [conductor]/clean_nodes config option https://review.openstack.org/316826 | 21:29 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "message" attribute support from IronicException https://review.openstack.org/316825 | 21:29 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove "periodic_interval" config option https://review.openstack.org/316824 | 21:29 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove deprecated driver_periodic_task https://review.openstack.org/316820 | 21:29 |
jroll | that was annoying | 21:30 |
*** Goneri has quit IRC | 21:33 | |
*** sabeen has quit IRC | 21:40 | |
*** nicodemos has quit IRC | 21:49 | |
mrda | Morning Ironic | 21:49 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Eliminate warnings about rm in api-ref build https://review.openstack.org/316793 | 21:51 |
jroll | ohai mrda | 21:52 |
jroll | good night ironic :P | 21:52 |
mrda | Night jroll :) | 21:53 |
*** sabeen has joined #openstack-ironic | 21:55 | |
thiagop | I'm going too | 22:04 |
thiagop | good night everyone | 22:04 |
*** thiagop has quit IRC | 22:04 | |
*** dmk0202 has quit IRC | 22:10 | |
*** [1]cdearborn has quit IRC | 22:11 | |
jlvillal | jroll: Ciao | 22:11 |
*** Sukhdev has quit IRC | 22:20 | |
*** Sukhdev has joined #openstack-ironic | 22:21 | |
*** penick has quit IRC | 22:27 | |
*** piet has joined #openstack-ironic | 22:36 | |
jlvillal | Woot! http://paste.openstack.org/show/497278/ | 22:41 |
*** ametts has quit IRC | 22:41 | |
*** piet has quit IRC | 22:50 | |
*** piet has joined #openstack-ironic | 22:52 | |
openstackgerrit | Merged openstack/ironic: Remove "discoverd" configuration group https://review.openstack.org/316821 | 22:53 |
*** rloo has quit IRC | 22:53 | |
openstackgerrit | Merged openstack/ironic: Remove [agent]/manage_tftp option https://review.openstack.org/316822 | 22:54 |
*** Sukhdev has quit IRC | 23:01 | |
*** sabeen has quit IRC | 23:04 | |
*** piet has quit IRC | 23:07 | |
*** rcernin has quit IRC | 23:08 | |
openstackgerrit | Merged openstack/ironic: Remove deprecated [pxe]/http_* options https://review.openstack.org/316823 | 23:12 |
*** mbound has quit IRC | 23:14 | |
jlvillal | sambetts|afk: FYI: https://review.openstack.org/#/c/247331/ another instance of CiscoCI going crazy... | 23:26 |
*** Sukhdev has joined #openstack-ironic | 23:38 | |
*** spandhe has joined #openstack-ironic | 23:41 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic: Fix shutdown.sh for grenade https://review.openstack.org/317139 | 23:42 |
*** ChrisAusten has quit IRC | 23:43 | |
jlvillal | Today was a very good day for grenade work! :) | 23:44 |
openstackgerrit | John L. Villalovos proposed openstack/ironic: Fix shutdown.sh & upgrade.sh for grenade https://review.openstack.org/317139 | 23:48 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!