| jlvillal | I updated the whiteboard | 00:00 |
|---|---|---|
| *** baoli has quit IRC | 00:05 | |
| *** milan has joined #openstack-ironic | 00:11 | |
| openstackgerrit | Merged openstack/ironic: Remove trailing backtick https://review.openstack.org/416111 | 00:17 |
| *** milan has quit IRC | 00:25 | |
| *** milan has joined #openstack-ironic | 00:31 | |
| *** trown is now known as trown|outtypewww | 00:35 | |
| *** milan has quit IRC | 00:40 | |
| *** hw_wutianwei has joined #openstack-ironic | 00:49 | |
| *** milan has joined #openstack-ironic | 00:51 | |
| *** hoangcx has joined #openstack-ironic | 00:53 | |
| *** milan has quit IRC | 00:55 | |
| *** SerenaFeng has joined #openstack-ironic | 00:57 | |
| *** phuongnh has joined #openstack-ironic | 00:57 | |
| *** ijw has quit IRC | 01:01 | |
| *** Fdaisuke_ has joined #openstack-ironic | 01:04 | |
| *** Fdaisuke has quit IRC | 01:05 | |
| *** rama_y has quit IRC | 01:09 | |
| *** milan has joined #openstack-ironic | 01:11 | |
| *** trananhkma has joined #openstack-ironic | 01:12 | |
| openstackgerrit | Cao Xuan Hoang proposed openstack/ironic-python-agent: Removes unnecessary utf-8 encoding https://review.openstack.org/413866 | 01:17 |
| *** ijw has joined #openstack-ironic | 01:22 | |
| *** milan has quit IRC | 01:25 | |
| *** Syed__ has quit IRC | 01:25 | |
| *** ijw has quit IRC | 01:27 | |
| *** milan has joined #openstack-ironic | 01:31 | |
| *** milan has quit IRC | 01:36 | |
| *** ijw has joined #openstack-ironic | 01:43 | |
| *** zhangjl has joined #openstack-ironic | 01:45 | |
| *** yufei has joined #openstack-ironic | 01:50 | |
| *** ijw has quit IRC | 01:50 | |
| *** milan has joined #openstack-ironic | 01:51 | |
| *** ijw has joined #openstack-ironic | 01:52 | |
| *** milan has quit IRC | 01:55 | |
| *** ijw has quit IRC | 01:57 | |
| *** ijw has joined #openstack-ironic | 02:05 | |
| *** ianw is now known as ianw_pto | 02:07 | |
| *** ijw has quit IRC | 02:11 | |
| *** milan has joined #openstack-ironic | 02:11 | |
| *** trananhkma has quit IRC | 02:12 | |
| *** rbudden has joined #openstack-ironic | 02:17 | |
| openstackgerrit | Tovin Seven proposed openstack/ironic: Enable OSProfiler support in Ironic https://review.openstack.org/358142 | 02:19 |
| openstackgerrit | Tovin Seven proposed openstack/ironic: Enable OSProfiler support in Ironic https://review.openstack.org/358142 | 02:20 |
| openstackgerrit | Dao Cong Tien proposed openstack/ironic: Adds clean step 'reset_bios_to_default' to iRMC drivers https://review.openstack.org/416403 | 02:21 |
| *** milan has quit IRC | 02:25 | |
| *** MattMan_1 has quit IRC | 02:31 | |
| *** MattMan_1 has joined #openstack-ironic | 02:31 | |
| *** milan has joined #openstack-ironic | 02:31 | |
| *** _Fdaisuke_ has joined #openstack-ironic | 02:39 | |
| *** milan has quit IRC | 02:40 | |
| *** Fdaisuke_ has quit IRC | 02:42 | |
| *** rama_y has joined #openstack-ironic | 02:49 | |
| *** moshele has joined #openstack-ironic | 02:50 | |
| *** milan has joined #openstack-ironic | 02:51 | |
| *** trananhkma has joined #openstack-ironic | 02:54 | |
| *** milan has quit IRC | 02:55 | |
| openstackgerrit | Merged openstack/ironic: 'updated_at' field value after node is updated https://review.openstack.org/371605 | 02:56 |
| openstackgerrit | Merged openstack/ironic: Clean up object code https://review.openstack.org/407635 | 02:56 |
| *** milan has joined #openstack-ironic | 03:11 | |
| *** milan has quit IRC | 03:25 | |
| *** milan has joined #openstack-ironic | 03:31 | |
| trananhkma | xek, morning :) are you around? | 03:35 |
| *** aparnav has joined #openstack-ironic | 03:38 | |
| *** milan has quit IRC | 03:40 | |
| *** moshele has quit IRC | 03:49 | |
| *** milan has joined #openstack-ironic | 03:51 | |
| *** milan has quit IRC | 03:55 | |
| *** links has joined #openstack-ironic | 03:56 | |
| *** SerenaFeng has quit IRC | 03:59 | |
| *** ijw has joined #openstack-ironic | 04:01 | |
| *** baoli has joined #openstack-ironic | 04:10 | |
| *** milan has joined #openstack-ironic | 04:11 | |
| *** rama_y has quit IRC | 04:14 | |
| *** baoli has quit IRC | 04:15 | |
| *** Goneri has quit IRC | 04:29 | |
| *** milan has quit IRC | 04:30 | |
| *** milan has joined #openstack-ironic | 04:31 | |
| *** milan has quit IRC | 04:48 | |
| *** milan has joined #openstack-ironic | 04:51 | |
| *** milan has quit IRC | 04:55 | |
| *** milan has joined #openstack-ironic | 05:11 | |
| openstackgerrit | paresh sao proposed openstack/proliantutils: Support to disable iscsi boot option in iLO https://review.openstack.org/406706 | 05:23 |
| *** ijw has quit IRC | 05:25 | |
| *** milan has quit IRC | 05:26 | |
| *** milan has joined #openstack-ironic | 05:31 | |
| *** g3ek has quit IRC | 05:33 | |
| *** milan has quit IRC | 05:41 | |
| *** g3ek has joined #openstack-ironic | 05:42 | |
| *** moshele has joined #openstack-ironic | 05:42 | |
| *** Goneri has joined #openstack-ironic | 05:44 | |
| *** SerenaFeng has joined #openstack-ironic | 05:44 | |
| *** milan has joined #openstack-ironic | 05:51 | |
| *** SerenaFeng has quit IRC | 05:53 | |
| openstackgerrit | paresh sao proposed openstack/proliantutils: Adds 'UefiTarget' as a boot option https://review.openstack.org/407487 | 05:56 |
| *** Goneri has quit IRC | 05:59 | |
| *** milan has quit IRC | 06:00 | |
| *** moshele has quit IRC | 06:08 | |
| *** milan has joined #openstack-ironic | 06:12 | |
| *** amotoki has joined #openstack-ironic | 06:22 | |
| *** milan has quit IRC | 06:25 | |
| *** milan has joined #openstack-ironic | 06:31 | |
| *** milan has quit IRC | 06:40 | |
| *** milan has joined #openstack-ironic | 06:51 | |
| *** pcaruana has joined #openstack-ironic | 06:51 | |
| *** e0ne has joined #openstack-ironic | 06:54 | |
| *** milan has quit IRC | 06:55 | |
| *** milan has joined #openstack-ironic | 07:11 | |
| *** mjura has joined #openstack-ironic | 07:12 | |
| *** rcernin has joined #openstack-ironic | 07:15 | |
| *** ChubYann has quit IRC | 07:16 | |
| *** e0ne has quit IRC | 07:17 | |
| *** tesseract has joined #openstack-ironic | 07:18 | |
| *** rbudden has quit IRC | 07:21 | |
| *** milan has quit IRC | 07:25 | |
| *** moshele has joined #openstack-ironic | 07:27 | |
| *** milan has joined #openstack-ironic | 07:31 | |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add soft reboot/poweroff power states. https://review.openstack.org/247904 | 07:36 |
| *** awiddersheim has quit IRC | 07:39 | |
| *** awiddersheim has joined #openstack-ironic | 07:39 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic power interface for soft reboot and soft power off https://review.openstack.org/216730 | 07:39 |
| *** milan has quit IRC | 07:41 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Ipmitool power driver for soft reboot and soft power off https://review.openstack.org/216738 | 07:41 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: iRMC power driver for soft reboot and soft power off https://review.openstack.org/216743 | 07:42 |
| joanna | morning everyone! | 07:42 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Update the existing APIs due to adding get_supported_power_states https://review.openstack.org/382194 | 07:42 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic management I/F for Inject NMI https://review.openstack.org/348191 | 07:43 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Ipmitool management driver for Inject NMI https://review.openstack.org/348721 | 07:44 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: iRMC management driver for Inject NMI https://review.openstack.org/348732 | 07:45 |
| *** SerenaFeng has joined #openstack-ironic | 07:48 | |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Support soft reboot and soft power off with timeout for OSC https://review.openstack.org/357627 | 07:49 |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add a new command "ironic node-inject-nmi" https://review.openstack.org/247905 | 07:50 |
| *** milan has joined #openstack-ironic | 07:51 | |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add a new OSC command for Inject NMI https://review.openstack.org/357661 | 07:52 |
| *** milan has quit IRC | 07:56 | |
| *** sasha21 has quit IRC | 08:04 | |
| *** milan has joined #openstack-ironic | 08:11 | |
| moshele | milan: hi can you review https://review.openstack.org/#/c/264257/ | 08:16 |
| patchbot_ | patch 264257 - ironic-inspector - Adding InfiniBand Support | 08:16 |
| vdrok | good morning ironic, joanna ! | 08:17 |
| *** milan has quit IRC | 08:26 | |
| *** milan has joined #openstack-ironic | 08:31 | |
| *** SerenaFeng has quit IRC | 08:33 | |
| *** SerenaFeng has joined #openstack-ironic | 08:33 | |
| *** pmannidi has quit IRC | 08:34 | |
| *** milan has quit IRC | 08:41 | |
| *** athomas has joined #openstack-ironic | 08:44 | |
| *** tuanluong has joined #openstack-ironic | 08:45 | |
| *** priteau has joined #openstack-ironic | 08:49 | |
| *** milan has joined #openstack-ironic | 08:51 | |
| *** milan has quit IRC | 08:56 | |
| *** zzzeek has quit IRC | 09:00 | |
| *** zzzeek has joined #openstack-ironic | 09:00 | |
| *** derekh has joined #openstack-ironic | 09:04 | |
| *** milan has joined #openstack-ironic | 09:12 | |
| milan | morning moshele vdrok! :) | 09:19 |
| milan | morning Ironic! | 09:19 |
| *** SerenaFeng has quit IRC | 09:19 | |
| milan | moshele, will do | 09:19 |
| vdrok | morning milan :) | 09:19 |
| *** SerenaFeng has joined #openstack-ironic | 09:19 | |
| * milan relocates | 09:21 | |
| moshele | milan: my patch -1 I need to fix them first :) | 09:22 |
| moshele | milan: morning :) | 09:22 |
| milan | :) | 09:22 |
| milan | moshele, cool, I'll check | 09:23 |
| *** milan has quit IRC | 09:26 | |
| *** lucas-afk is now known as lucasagomes | 09:31 | |
| lucasagomes | morning all | 09:31 |
| vdrok | morning lucasagomes | 09:32 |
| *** ohamada has joined #openstack-ironic | 09:35 | |
| openstackgerrit | Andrey Shestakov proposed openstack/bifrost: Add support of collecting logs from baremetal nodes https://review.openstack.org/415503 | 09:43 |
| openstackgerrit | Galyna Zholtkevych proposed openstack/python-ironicclient: [WIP] ETAGs for node in ironicclient https://review.openstack.org/400335 | 09:52 |
| dtantsur | morning Ironic, morning lucasagomes, vdrok and milan | 09:57 |
| vdrok | morning dtantsur | 09:57 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Adding InfiniBand Support https://review.openstack.org/264257 | 09:58 |
| *** moshele has quit IRC | 10:06 | |
| *** moshele has joined #openstack-ironic | 10:06 | |
| *** milan has joined #openstack-ironic | 10:10 | |
| *** yufei has quit IRC | 10:15 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Remove references to ZAP* state https://review.openstack.org/416505 | 10:18 |
| *** yufei has joined #openstack-ironic | 10:18 | |
| *** yufei has quit IRC | 10:18 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Remove references to ZAP* state https://review.openstack.org/416505 | 10:20 |
| *** tuanluong has quit IRC | 10:21 | |
| *** e0ne has joined #openstack-ironic | 10:22 | |
| *** hoangcx has quit IRC | 10:35 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Tempest test tag baremetal doesn't exist https://review.openstack.org/412448 | 10:46 |
| *** SerenaFeng has quit IRC | 10:49 | |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: Advertise the correct address when using IPv6 https://review.openstack.org/411817 | 10:49 |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: iscsi: Listen for connections on both IPv4 and IPv6 ports https://review.openstack.org/411819 | 10:49 |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: Agent: Listen for connections on both IPv4 and IPv6 ports https://review.openstack.org/411818 | 10:49 |
| *** links has quit IRC | 10:57 | |
| *** mgould|a1k is now known as mgould | 10:57 | |
| mgould | morning Ironic, happy new year! | 10:57 |
| *** zhangjl has quit IRC | 10:59 | |
| dtantsur | morning mgould | 11:00 |
| *** sambetts|afk is now known as sambetts | 11:04 | |
| sambetts | o/ mgould lucasagomes dtantsur mgould vdrok milan | 11:04 |
| vdrok | morning mgould and sambetts | 11:05 |
| jroll | morning sambetts, dtantsur, mgould, vdrok, milan, lucasagomes, moshele :D | 11:05 |
| lucasagomes | sambetts, jroll o/ morning | 11:05 |
| sambetts | hey jroll | 11:05 |
| milan | morning dtantsur jroll lucasagomes mgould sambetts ! :) | 11:05 |
| vdrok | morning jroll :) | 11:06 |
| milan | finally end of PTOs, right? ;) | 11:06 |
| jroll | idk, I'm ready for another | 11:06 |
| milan | same here :D | 11:07 |
| dtantsur | morning jroll | 11:08 |
| moshele | jroll: moring | 11:08 |
| dtantsur | milan, fyi I'm cancelling the sheraton thing | 11:08 |
| milan | dtantsur, OK I feel like staying there | 11:08 |
| dtantsur | milan, ok, though I'm not sure how to expense the deposit they take.. | 11:09 |
| dtantsur | and whether we actually should | 11:09 |
| milan | dtantsur, what's the diff w/ the other hotel? | 11:09 |
| milan | * $$ | 11:09 |
| dtantsur | milan, I'm going to Hampton Inn, it's 1. cheaper, 2. has a clear WiFi and breakfast policy, 3. does not seem to withdraw money in advance (at least they haven't done it since yesterday) | 11:10 |
| * dtantsur remembers somebody was talking about cheap venues for the PTG, right? | 11:10 | |
| milan | hmmm, how much cheaper it is per night? | 11:11 |
| * milan evaluates laziness/savings ratio | 11:11 | |
| dtantsur | milan, something like $195 vs $220 | 11:11 |
| mgould | morning dtantsur sambetts vdrok jroll lucasagomes milan moshele joanna, happy new year! | 11:11 |
| lucasagomes | mgagne, hi there! Happy new year | 11:11 |
| milan | $25 sounds like the breakfast margin | 11:12 |
| milan | Happy new year mgould! :) | 11:12 |
| * milan struggles w/ construct & lldp parsing here, still not clear to me whether using it helped https://review.openstack.org/#/c/406496 | 11:13 | |
| patchbot_ | patch 406496 - ironic-inspector - Add plugin to process basic management LLDP TLVs | 11:13 |
| milan | bad thing it was my idea :-/ | 11:13 |
| sambetts | haha | 11:14 |
| milan | something doesn't add up there just cannot wrap my head around how to make it click | 11:14 |
| * milan tea | 11:14 | |
| *** phuongnh has quit IRC | 11:20 | |
| *** links has joined #openstack-ironic | 11:21 | |
| *** milan is now known as food|milan | 11:23 | |
| mgould | that's weird, Lastpass is no longer filling in my login details for review.openstack.org | 11:30 |
| mgould | did they change the URL of the single-sign-on page? | 11:31 |
| sambetts | mgould: I saw the same thing on my end too | 11:31 |
| sambetts | mgould: not sure why though | 11:31 |
| mgould | unified branding, probably | 11:33 |
| * mgould marks launchpad.net and ubuntu.com as equivalent in LastPass - success! | 11:35 | |
| *** jkilpatr has quit IRC | 11:39 | |
| *** jkilpatr has joined #openstack-ironic | 12:02 | |
| *** karthiks has quit IRC | 12:11 | |
| openstackgerrit | Merged openstack/ironic: Create a table to track loaded interfaces https://review.openstack.org/409812 | 12:18 |
| *** openstackgerrit has quit IRC | 12:33 | |
| *** hw_wutianwei has quit IRC | 12:38 | |
| *** lucasagomes is now known as lucas-hungry | 12:39 | |
| *** openstackgerrit has joined #openstack-ironic | 12:40 | |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: Advertise the correct address when using IPv6 https://review.openstack.org/411817 | 12:40 |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: iscsi: Listen for connections on both IPv4 and IPv6 ports https://review.openstack.org/411819 | 12:40 |
| openstackgerrit | Derek Higgins proposed openstack/ironic-python-agent: Agent: Listen for connections on both IPv4 and IPv6 ports https://review.openstack.org/411818 | 12:40 |
| *** Nisha_Agarwal has joined #openstack-ironic | 12:45 | |
| *** moshele has quit IRC | 13:03 | |
| openstackgerrit | Galyna Zholtkevych proposed openstack/python-ironicclient: Fix OSC port create and port set commands https://review.openstack.org/384345 | 13:11 |
| *** moshele has joined #openstack-ironic | 13:14 | |
| *** trown|outtypewww is now known as trown | 13:17 | |
| openstackgerrit | Merged openstack/ironic-inspector: Tempest test tag baremetal doesn't exist https://review.openstack.org/412448 | 13:17 |
| dtantsur | \o/ | 13:18 |
| *** Nisha_Agarwal has quit IRC | 13:23 | |
| vdrok | lucas-hungry: oh, by the way, this finally passed | 13:24 |
| vdrok | https://review.openstack.org/414091 | 13:24 |
| patchbot_ | patch 414091 - wsme - Fix pep8 issues and switch to py35 in tox.ini | 13:24 |
| *** bfournie has quit IRC | 13:24 | |
| openstackgerrit | Aparna proposed openstack/proliantutils: Add Sanitize disk erase support using SSA https://review.openstack.org/398202 | 13:25 |
| *** jheroux has joined #openstack-ironic | 13:27 | |
| *** aparna has joined #openstack-ironic | 13:28 | |
| *** aparnav has quit IRC | 13:30 | |
| *** aparna has quit IRC | 13:33 | |
| sambetts | food|milan: taking a look at that patch regarding the TLVs construct certainly makes things look complicated at first glance, but I don't know if it would look any less complicated if we'd written native python code to do the same sort of switching logic that we can do with the structs, however there is someparts which are very difficult to understand like the get_autoneg_cap function (again | 13:40 |
| sambetts | that might not be related to construct, but its diffuclt to tell) | 13:40 |
| *** food|milan is now known as milan | 13:41 | |
| milan | sambetts, yeah it is complicated | 13:41 |
| milan | sambetts, i'm considering embedding the values now | 13:41 |
| milan | as the TLVs are supposed to have the type, subtype& value fields | 13:42 |
| milan | so maybe that way it's going to be simpler | 13:42 |
| milan | * less nested | 13:42 |
| milan | so we have a chassis_id TLV that has chassis_id == 'chassis_id' | 13:43 |
| milan | * type == 'chassis_id' | 13:43 |
| milan | value = '192.168.1.1' | 13:43 |
| milan | subtype = 'IANA_address' | 13:43 |
| milan | family = 'ipv4' | 13:43 |
| milan | or PortID w/ type = 'port_id' | 13:43 |
| *** fritzstauffacher has joined #openstack-ironic | 13:44 | |
| milan | value = '2222' | 13:44 |
| milan | subtype = 'you got the picture' | 13:44 |
| milan | should allow us to "dump" them in a regular fashion | 13:44 |
| milan | without the if statements | 13:45 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Register/unregister hardware interfaces for conductors https://review.openstack.org/412631 | 13:45 |
| * milan will post the review in a bit | 13:45 | |
| jroll | dtantsur: ^ I think that's good to go now | 13:45 |
| sambetts | yeah :/ /me is still learning construct so I'll look forward to seeing the review | 13:45 |
| milan | sambetts, same here O:-) | 13:46 |
| milan | it just felt like a good idea in the beginning :D | 13:46 |
| *** sasha2 has joined #openstack-ironic | 13:47 | |
| *** dprince has joined #openstack-ironic | 13:48 | |
| sambetts | milan: heh, I really like the Struct objects, with the cc.Switch(...) parts, it makes it good for processing cases where we have differnt subtypes that need different processing, instead of having if subtype == 1: elif subtype == 2: elif subtype == 3: etc etc | 13:48 |
| milan | sambetts, exactly, it should parse a TLV into a flat container w/ values set according to the subtype&type combination | 13:50 |
| sambetts | yeah :) | 13:50 |
| milan | so one can process a TLV record w/ correct data in it | 13:50 |
| *** lucas-hungry is now known as lucasagomes | 13:51 | |
| dtantsur | jroll, will check in a few, thanks | 13:51 |
| lucasagomes | vdrok, oh hah great1 | 13:51 |
| jroll | thanks, no rush | 13:51 |
| lucasagomes | vdrok, finally, I still don't know what happened in the gate there | 13:51 |
| lucasagomes | vdrok, cause it totally worked for me locally | 13:51 |
| lucasagomes | anyway... glad gate is now happy | 13:52 |
| milan | sambetts, I regret a bit we parse the TLVs raw data into the type/value pairs, better off if it was the blob imo | 13:52 |
| vdrok | lucasagomes: yup, same for me | 13:52 |
| milan | sambetts, we could just tlv.parse(the_blob) | 13:52 |
| sambetts | milan: yeah, tbh we might be able to change that if raw raw data might be easier | 13:52 |
| milan | sambetts, yeah, maybe this patch should do that | 13:53 |
| sambetts | milan: its a IPA change required | 13:53 |
| milan | so tlv.parse(the_blob).type == 'chassis_id' | 13:53 |
| sambetts | jroll: whats the policy on changing the inventory in IPA? | 13:53 |
| milan | sambetts, aaah :-/ | 13:53 |
| jroll | sambetts: make it work with one version on each side, pls :) | 13:54 |
| jroll | sambetts: it does have a version if we need to change the ironic side | 13:54 |
| milan | heh | 13:54 |
| milan | :D | 13:54 |
| lucasagomes | vdrok, cdent just approved it :-) | 13:55 |
| vdrok | thanks! | 13:56 |
| sambetts | milan: IPA processes the raw data into the (INT: BIN) type: value tuples we receive (my fault cos I wrote it before we were going to process the whole thing) | 13:56 |
| milan | sambetts, it would simplify the hole thing a lot if it was just a list of blobs, right? | 13:57 |
| milan | no parsing at all | 13:57 |
| milan | in the IPA | 13:57 |
| milan | people parsing what they need if use their own plugin in inspector | 13:58 |
| milan | or using the default one we provide | 13:58 |
| milan | *list of blobs maybe just the PDU struct | 13:59 |
| milan | even simpler | 13:59 |
| dtantsur | so, the CI is back up, right? we can update the whiteboard then | 14:00 |
| *** Nakato has quit IRC | 14:00 | |
| *** Nakato has joined #openstack-ironic | 14:00 | |
| milan | sambetts, single blob parsing w/ http://paste.openstack.org/show/593863/ : tlvs = pdu.parse(blob).tlvs | 14:01 |
| milan | then inventory['lldp'] = tlvs | 14:02 |
| milan | but maybe I'm oversimplifying it | 14:04 |
| sambetts | milan: it'll require ripping a pretty large chuck on code out of IPA https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/netutils.py and also its included in the IPA inventory for anyone, not just inspection | 14:05 |
| *** bfournie has joined #openstack-ironic | 14:08 | |
| *** baoli has joined #openstack-ironic | 14:09 | |
| *** baoli has quit IRC | 14:09 | |
| milan | sambetts, https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/netutils.py#L120,#L139 replace w/ just store the whole packet as a PDU (hex-encoded) | 14:09 |
| *** baoli has joined #openstack-ironic | 14:10 | |
| milan | sambetts, actually, are the TLVs received as packets or as PDU struct? | 14:11 |
| sambetts | milan: raw packets from a socket | 14:12 |
| milan | so each packet == tlv | 14:12 |
| sambetts | milan: no, each packet contains multiple TLVs | 14:12 |
| sambetts | milan: and you have to chop up the binary to process it | 14:13 |
| milan | sambetts, OK then packet == pdu | 14:13 |
| milan | it should be this format: http://paste.openstack.org/show/593863/ | 14:13 |
| milan | nope? | 14:13 |
| milan | * the packet | 14:13 |
| sambetts | heh maybe, in my head PDU == power distribution unit XD https://en.wikipedia.org/wiki/Link_Layer_Discovery_Protocol#Frame_structure | 14:14 |
| milan | ** the packet payload | 14:14 |
| milan | heh :) | 14:14 |
| milan | LLDP Data Unit (LLDPDU) | 14:16 |
| milan | my fault | 14:16 |
| milan | :) | 14:16 |
| sambetts | :-P | 14:16 |
| sambetts | basically the ethernet packet payload is a TLVs one after each other and you have to use the L part to work out the how many bits to read next | 14:17 |
| *** Goneri has joined #openstack-ironic | 14:17 | |
| sambetts | the process manually is read the type, read the legth, read the val, read the type read the length read the val, until you run out of binary | 14:18 |
| sambetts | to read | 14:18 |
| *** dtantsur is now known as dtantsur|brb | 14:18 | |
| *** hwoarang has quit IRC | 14:19 | |
| *** priteau has quit IRC | 14:19 | |
| *** hwoarang has joined #openstack-ironic | 14:20 | |
| milan | sambetts, http://www.ieee802.org/1/files/public/docs2002/lldp-protocol-01.pdf call it the PDU :-/ what ever | 14:20 |
| milan | sambetts, yeah, that sounds like a job for the construct to parse | 14:20 |
| milan | the stream | 14:20 |
| sambetts | milan: yeah | 14:20 |
| milan | so if we had the lldp.py in some shared lib, we | 14:21 |
| milan | could reuse in IPA | 14:21 |
| milan | and inspector plugins referring to constants | 14:22 |
| *** karthiks has joined #openstack-ironic | 14:22 | |
| milan | * use in IPA reuse in Inspector | 14:22 |
| milan | but lemme post the review first | 14:23 |
| bfournie | morning milan and sambetts - are you discussing the construct lib and lldp? | 14:23 |
| milan | bfournie, morning! :) yup | 14:23 |
| bfournie | I'm not sure IPA would get much benefit out of the construct lib as it doesn't care about format of the TLVs | 14:24 |
| milan | bfournie, tl;dr we're discussing where the parsing should happen | 14:24 |
| sambetts | bfournie: yeah, we're talking about whether IPA should even do the processing that it currently does or not, because if we had the raw binary data unprocessed then its easier to just throw the whole lot at construct | 14:24 |
| karthiks | milan, Can you please have a look at https://review.openstack.org/#/c/396147/ | 14:24 |
| patchbot_ | patch 396147 - ironic-inspector-specs - Spec for retrieving NUMA node information | 14:24 |
| milan | karthiks, sure | 14:24 |
| karthiks | milan, Thanks. | 14:25 |
| * milan 's review queue grows | 14:25 | |
| sambetts | milan: we'd have to deal with the issue that IPA isn't versioned right now and Mitaka inspector needs to continue to work with Newton IPA and visa-versa :/ | 14:26 |
| bfournie | sambetts, milan - yeah, could use the construct lib to process entire TLV packet in inspector, although would require IPA changes then | 14:26 |
| sambetts | *cough* plug review https://review.openstack.org/#/c/341086/ *cough* | 14:26 |
| patchbot_ | patch 341086 - ironic-specs - Add spec for IPA API versioning support | 14:26 |
| milan | :D | 14:26 |
| * milan .track(^^) | 14:27 | |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add new dbsync command online_data_migrations https://review.openstack.org/408556 | 14:29 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add version column https://review.openstack.org/412397 | 14:29 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add RPC and object version pinning https://review.openstack.org/407491 | 14:29 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Rolling upgrade procedure documentation https://review.openstack.org/416565 | 14:29 |
| milan | no way to deprecate API unless IPA is versioned I guess :-/ | 14:29 |
| * milan .grab(tea) # again. | 14:31 | |
| *** thiagop has joined #openstack-ironic | 14:32 | |
| thiagop | (coming back from the dead to say) Good morning, Ironicers! | 14:32 |
| *** rbartal has quit IRC | 14:32 | |
| jroll | thiagop: oh wow, welcome back \o/ | 14:33 |
| thiagop | How's it going around here jroll ? Did I missed a whole lot? | 14:34 |
| jroll | thiagop: of course you did! things are good :) | 14:34 |
| *** rbartal has joined #openstack-ironic | 14:35 | |
| *** lindycoder has joined #openstack-ironic | 14:35 | |
| thiagop | jroll: great to know :) | 14:37 |
| vdrok | morning thiagop :) | 14:40 |
| thiagop | hiya vdrok | 14:40 |
| milan | morning thiagop! :) | 14:40 |
| thiagop | milan: o/ | 14:41 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add version column https://review.openstack.org/412397 | 14:42 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Add new dbsync command online_data_migrations https://review.openstack.org/408556 | 14:42 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Rolling upgrade procedure documentation https://review.openstack.org/416565 | 14:43 |
| milan | bfournie, sambetts so, hypothetically, we have a library that parses the LLDP as a stream in IPA and stores the struct in the inventory: lldp{'flags': ..., 'ttl':..., 'tlvs': [...]} ; how about allowing the user to extend the basic set we'd provide through plugins in IPA? | 14:43 |
| milan | *tlvs: [{'type': 'chassis_id', 'value': '192.168.0.1', 'subtype': 'IANA_address', 'family': 'ipv4'}, ....] | 14:46 |
| milan | imo it makes more sense to have an inspector plugin that decides the fate of a BM based on this info than to parse it only | 14:47 |
| *** vinsh has joined #openstack-ironic | 14:48 | |
| *** athomas has quit IRC | 14:49 | |
| *** jcoufal has joined #openstack-ironic | 14:49 | |
| *** baoli has quit IRC | 14:52 | |
| *** athomas has joined #openstack-ironic | 14:55 | |
| *** ppiela has joined #openstack-ironic | 14:57 | |
| *** dtantsur|brb is now known as dtantsur | 14:57 | |
| *** baoli has joined #openstack-ironic | 14:58 | |
| *** links has quit IRC | 14:59 | |
| *** dyasny has joined #openstack-ironic | 15:00 | |
| *** yufei has joined #openstack-ironic | 15:00 | |
| *** vinsh has quit IRC | 15:02 | |
| *** dyasny has quit IRC | 15:04 | |
| *** baoli has quit IRC | 15:05 | |
| * jroll steps away for a while | 15:07 | |
| *** baoli has joined #openstack-ironic | 15:08 | |
| NobodyCam | Good Morning and Happy New year Ironic'ers | 15:11 |
| dtantsur | hey NobodyCam, how were holidays for you? | 15:11 |
| sambetts | Hey NobodyCam | 15:11 |
| NobodyCam | hey hey dtantsur they where awesome. I hope yours was good too! | 15:12 |
| NobodyCam | hey morning sambetts :) | 15:12 |
| dtantsur | pretty nice, yeah :) | 15:12 |
| NobodyCam | Now its email catch up time! :p | 15:12 |
| *** g3ek has quit IRC | 15:13 | |
| *** zzzeek has quit IRC | 15:13 | |
| *** zzzeek has joined #openstack-ironic | 15:14 | |
| *** g3ek has joined #openstack-ironic | 15:14 | |
| bfournie | Hi milan, not sure if I understand your proposal^. We do have a plan to provide separate inspector plugins to deal with different TLV sets, e.g. LLDP-MED or vendor specific TLVs per https://review.openstack.org/#/c/374381/ | 15:15 |
| patchbot_ | patch 374381 - ironic-inspector-specs - Add LLDP processing hook and new CLI commands (MERGED) | 15:15 |
| thiagop | NobodyCam: long time no see! | 15:16 |
| thiagop | NobodyCam: how's it going? | 15:16 |
| milan | morning NobodyCam! :) | 15:16 |
| milan | bfournie, yeah I know :-/ | 15:17 |
| *** dyasny has joined #openstack-ironic | 15:18 | |
| NobodyCam | hey hey thiagop happy new year. :) going well, I was out in Roswell New Mexico for new year drinking :) | 15:18 |
| NobodyCam | hey hey milan :) happy new year | 15:18 |
| thiagop | NobodyCam: is this the place with a lot of UFO sighting, right? | 15:19 |
| milan | happy new year NobodyCam! :) | 15:19 |
| NobodyCam | thiagop: yep :) hehehehe | 15:19 |
| NobodyCam | milan: :) | 15:20 |
| thiagop | NobodyCam: awesome | 15:20 |
| NobodyCam | :) thank you, | 15:21 |
| NobodyCam | we had a blast :) | 15:21 |
| NobodyCam | though it was a bit cool.. as in cold! | 15:21 |
| *** vinsh has joined #openstack-ironic | 15:22 | |
| thiagop | I bet | 15:23 |
| *** ijw has joined #openstack-ironic | 15:23 | |
| *** rbartal has quit IRC | 15:24 | |
| dtantsur | sambetts, pxe_ucs doesn't allow disabling SSL validation, does it? | 15:24 |
| *** srobert has joined #openstack-ironic | 15:27 | |
| *** g3ek has quit IRC | 15:27 | |
| *** ijw has quit IRC | 15:29 | |
| *** Guest66666 has quit IRC | 15:29 | |
| sambetts | dtantsur: :/ I don't see anything in the code to do it, the ucsm driver makes me sad... | 15:29 |
| sambetts | dtantsur: can I ask what your using for? | 15:29 |
| *** Guest66666 has joined #openstack-ironic | 15:29 | |
| dtantsur | sambetts, I was approached by the guy with an invalid certificate (default?) on the BMC. that's all I know :) | 15:30 |
| *** absubram has joined #openstack-ironic | 15:30 | |
| *** mjura has quit IRC | 15:30 | |
| dtantsur | | last_error | Failed to get power state for node 50cec054-2c5d-4077-8df7-f89a9d562da6. | | 15:31 |
| openstackgerrit | Bradley Jones proposed openstack/ironic-specs: Cisco Nexus POAP Driver https://review.openstack.org/416588 | 15:31 |
| dtantsur | | | Error: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate | | 15:31 |
| dtantsur | | | verify failed (_ssl.c:579)> | 15:31 |
| *** baoli has quit IRC | 15:31 | |
| milan | bfournie, sambetts reading https://specs.openstack.org/openstack/ironic-inspector-specs/specs/lldp-reporting.html#data-model-impact the example data; we're just aggregating this into a dictionary here but someone might have a better idea how to process that information (dumping the parsed data and doing aggregation w/ eg jq as necessary) | 15:32 |
| milan | bfournie, sambetts I know we've got the spec, just wondering here | 15:33 |
| *** rbartal has joined #openstack-ironic | 15:34 | |
| *** g3ek has joined #openstack-ironic | 15:37 | |
| TheJulia | good morning | 15:37 |
| *** srobert has quit IRC | 15:38 | |
| mat128 | good morning TheJulia | 15:38 |
| jlvillal | Good morning all :) | 15:38 |
| milan | morning TheJulia! :) | 15:38 |
| milan | morning jlvillal, mat128! :) | 15:39 |
| *** Sukhdev has joined #openstack-ironic | 15:39 | |
| * jlvillal is happy to see the Nova patch just got approved | 15:39 | |
| sambetts | dtantsur: ah :/ not sure I'm afraid, in my CI for the ucsm driver I didn't have to do anything special to make it work with Ironic | 15:39 |
| TheJulia | jlvillal: awesome! | 15:39 |
| jlvillal | TheJulia: It has been about two weeks or more with that issue. Holidays didn't help the speed ;) | 15:39 |
| sambetts | dtantsur: unfortunatly although I maintain the Ironic driver I'm not an expert in maintaining the UCSM itself :/ | 15:40 |
| dtantsur | I see, nevermind | 15:40 |
| dtantsur | maybe I'll get more details from them | 15:40 |
| TheJulia | thiagop: no UFOs, lots of super cheezey places, and only a few good places to get breakfast. | 15:40 |
| jlvillal | sambetts: Crazy! And everyone thinks I'm an expert on all Intel hardware ;) | 15:40 |
| *** vinsh_ has joined #openstack-ironic | 15:40 | |
| dtantsur | morning TheJulia | 15:40 |
| sambetts | jlvillal: you know exactly how the 7700k works right ;) | 15:41 |
| TheJulia | jlvillal: I think that is standard for that time of year. :) | 15:41 |
| jlvillal | sambetts: :D | 15:41 |
| jlvillal | TheJulia: Yep | 15:41 |
| *** rama_y has joined #openstack-ironic | 15:41 | |
| *** thiagop has quit IRC | 15:42 | |
| *** thiagop has joined #openstack-ironic | 15:42 | |
| *** vinsh has quit IRC | 15:43 | |
| *** thiagop has quit IRC | 15:44 | |
| bfournie | milan: yes, sure, user could use jq to postprocess the processed lldp data. Here is an example command on a baremetal system that has the lldp_basic plugin installed that will give processed data for one interface - http://paste.openstack.org/show/593870/ | 15:45 |
| *** Nisha_Agarwal has joined #openstack-ironic | 15:47 | |
| milan | bfournie, yup, just imagine the data was in the original form | 15:47 |
| milan | but json | 15:47 |
| * milan pastes | 15:47 | |
| sambetts | lucasagomes, dtantsur: do you remember what that website was with the concurrency code exercises on it? | 15:49 |
| * dtantsur does not | 15:49 | |
| *** rama_y has quit IRC | 15:50 | |
| lucasagomes | sambetts, hmm I remember that, can't remmeber the url now | 15:50 |
| lucasagomes | 1 sec I'm on a call, but I will try to find it later | 15:51 |
| lucasagomes | sambetts, actually here: https://deadlockempire.github.io/ | 15:51 |
| lucasagomes | history ftw! | 15:51 |
| sambetts | woop | 15:51 |
| openstackgerrit | Bradley Jones proposed openstack/ironic-specs: Cisco Nexus POAP Driver https://review.openstack.org/416588 | 15:51 |
| sambetts | \o/ | 15:51 |
| sambetts | lucasagomes: thanks!!! | 15:51 |
| NobodyCam | oh nice | 15:52 |
| milan | bfournie, sambetts such as http://paste.openstack.org/show/593873/ as I don't think we know better that the data user how to accumulate&process it | 15:53 |
| *** Syed__ has joined #openstack-ironic | 15:53 | |
| sambetts | milan: I don't know if we'd need to save the ttl and length and flag etc right? | 15:54 |
| milan | sambetts, maybe not | 15:55 |
| bfournie | milan. sambetts: agree, I don't think its necessary to store the entire processed packet, I don't see the use case | 15:55 |
| milan | bfournie, you mean parsed&aggregated? | 15:56 |
| *** yufei has quit IRC | 15:56 | |
| bfournie | milan: I mean no need to store parsed type/subtype | 15:56 |
| milan | bfournie, but then how do you know what you're looking at? | 15:57 |
| milan | would have to "if/try" parse things to infer their meaning, no? | 15:58 |
| bfournie | milan: well we are storing it in as type/value. I don't mean to not identify it, but for OUI TLVs, no need to store '127' for example | 15:58 |
| milan | maybe I don't follow | 15:58 |
| milan | bfournie, you need it the same for the end user to be able to tell it's not the "basic" TLV nope? | 15:59 |
| *** milan is now known as milan|mgt | 16:00 | |
| bfournie | milan: again, not sure why they would care if they are consuming the output of the lldp data to get switch configuration. If they really want to parse/decode the lldp packets they are there in raw format. What we're providing is a processed output for users to consume, so they don't have to find where in in the 802.1AB spec '127' is a LLDP OUI type, for example. They should not have to open up a spec to use this data. | 16:02 |
| *** dsneddon_afk is now known as dsneddon | 16:04 | |
| *** rcernin has quit IRC | 16:09 | |
| *** rbartal has quit IRC | 16:15 | |
| *** ijw has joined #openstack-ironic | 16:22 | |
| *** ricardoas has joined #openstack-ironic | 16:23 | |
| *** ijw has quit IRC | 16:27 | |
| *** athomas has quit IRC | 16:32 | |
| *** trown is now known as trown|lunch | 16:39 | |
| vdrok | good night everyone! | 16:40 |
| *** athomas has joined #openstack-ironic | 16:40 | |
| lucasagomes | vdrok, see ya, g'night | 16:41 |
| NobodyCam | night vdrok | 16:43 |
| *** srart has quit IRC | 16:47 | |
| *** palmertime has joined #openstack-ironic | 16:49 | |
| *** topol has left #openstack-ironic | 16:52 | |
| openstackgerrit | Merged openstack/ironic-specs: Remove references to ZAP* state https://review.openstack.org/416505 | 16:55 |
| *** ijw has joined #openstack-ironic | 16:56 | |
| palmertime | Im having a time out issue with Baremetal deploy with driver pxe_ipmitool. After install the systems power off but fail to power back on. Is there a way to increase the tries or timeout for this? Maybe increase the time between tries? Here is the error log entry and the manually repeated ipmitool command. https://gist.github.com/palmertime/20f9ffad925fcaa276b764df927b0abf | 16:58 |
| *** srart has joined #openstack-ironic | 16:59 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Move default image logic from DevStack to Ironic https://review.openstack.org/414227 | 16:59 |
| lucasagomes | hey folks, can someone take a look at ^ please | 16:59 |
| *** rama_y has joined #openstack-ironic | 17:00 | |
| lucasagomes | the devstack counterpart patch is the: https://review.openstack.org/#/c/414604/ | 17:00 |
| patchbot_ | patch 414604 - openstack-dev/devstack - Remove default image logic for Ironic from DevStack | 17:00 |
| JayF | lucasagomes: ordering will matter for that, right? Or will DEFAULT_IMAGE_NAME not be set in any other devstack plugins? | 17:01 |
| *** moshele has quit IRC | 17:01 | |
| lucasagomes | JayF, so, AFAIK it won't be set in others plugins but you raise a good question here | 17:02 |
| JayF | I'm just wondering if that's a lot safer if we use an ironic-namespaced value | 17:02 |
| JayF | since we no longer rely on anything outside the plugin for that | 17:02 |
| lucasagomes | true, lemme take a look | 17:04 |
| * milan|mgt relocates, bfournie talk to you in a bit | 17:05 | |
| *** milan|mgt has quit IRC | 17:05 | |
| lucasagomes | JayF, so far the only place it's set seems to be: https://review.openstack.org/#/c/375095/ | 17:05 |
| patchbot_ | patch 375095 - openstack-dev/devstack - Ironic: Set default images for UEFI deployment | 17:05 |
| lucasagomes | I wonder if we could get rid of this DEFAULT_IMAGE_NAME var for all, and just have our own thing inside our plugin | 17:06 |
| * lucasagomes will test locally | 17:06 | |
| JayF | lucasagomes: let me ask the question differently: is there any virt... yeah that | 17:06 |
| JayF | basically wondering why we'd use a generic name | 17:06 |
| *** ijw has quit IRC | 17:06 | |
| *** zhugaoxiao has quit IRC | 17:07 | |
| *** srart has quit IRC | 17:07 | |
| *** zhugaoxiao has joined #openstack-ironic | 17:07 | |
| lucasagomes | JayF, yeah maybe we don't need it | 17:08 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Add new capabilities to ironic https://review.openstack.org/338138 | 17:12 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Register/unregister hardware interfaces for conductors https://review.openstack.org/412631 | 17:14 |
| jroll | zomg so much code | 17:14 |
| *** ijw has joined #openstack-ironic | 17:15 | |
| *** srart has joined #openstack-ironic | 17:17 | |
| soliosg | Rephrasing question, when should I use two underscores in module or test method? | 17:18 |
| openstackgerrit | Merged openstack/proliantutils: Support to disable iscsi boot option in iLO https://review.openstack.org/406706 | 17:19 |
| soliosg | Sorry, some keyboard shortcut in weechat pasted that comment, it's out of context | 17:21 |
| *** tesseract has quit IRC | 17:21 | |
| *** ijw has quit IRC | 17:23 | |
| JayF | jroll: https://review.openstack.org/#/c/412621/ should be fine as a backport, right? Just checking since I know you wrote those chains, | 17:23 |
| patchbot_ | patch 412621 - ironic (stable/mitaka) - Turn off tempest's multitenant network tests | 17:23 |
| *** 09SAAA525 has joined #openstack-ironic | 17:25 | |
| *** e0ne has quit IRC | 17:31 | |
| mgould | soliosg: in Python? http://stackoverflow.com/a/12629901 | 17:32 |
| *** 09SAAA525 has quit IRC | 17:33 | |
| *** mgould is now known as mgould|afk | 17:40 | |
| *** trown|lunch is now known as trown | 17:41 | |
| *** ijw has joined #openstack-ironic | 17:44 | |
| *** baoli has joined #openstack-ironic | 17:45 | |
| *** ijw has quit IRC | 17:45 | |
| *** rpioso has joined #openstack-ironic | 17:53 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: WIP/DNM: Testing gate https://review.openstack.org/407220 | 17:53 |
| *** moshele has joined #openstack-ironic | 17:53 | |
| *** derekh has quit IRC | 17:58 | |
| *** trown is now known as trown|brb | 17:59 | |
| *** trown|brb is now known as trown | 17:59 | |
| *** Sukhdev has quit IRC | 18:01 | |
| *** Sukhdev has joined #openstack-ironic | 18:02 | |
| *** dprince has quit IRC | 18:06 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic-specs: Intel Node Manager sensor collection support https://review.openstack.org/413249 | 18:07 |
| *** ohamada has quit IRC | 18:09 | |
| *** dtantsur is now known as dtantsur|afk | 18:17 | |
| jroll | JayF: yes | 18:18 |
| jroll | JayF: in general, the devstack/ directory is almost always okay to backport | 18:18 |
| JayF | given it was disabling tests, I wanted to be 100% suer | 18:19 |
| JayF | so we weren't removing tests in one place that weren't readded elsewhere | 18:19 |
| jroll | JayF: it disables tests we don't use :) | 18:19 |
| jroll | as weird as that sounds | 18:19 |
| JayF | I mean, I trust you about it. I don't know as much about our tempest tests as I should | 18:19 |
| *** dprince has joined #openstack-ironic | 18:20 | |
| lucasagomes | folks, I'm calling it a day | 18:20 |
| JayF | o/ | 18:21 |
| lucasagomes | JayF, tomorrow I let you know about the tests with a unique name ofr the default image | 18:21 |
| JayF | Thanks :) | 18:21 |
| JayF | have a good day! | 18:21 |
| lucasagomes | you too! See ya | 18:21 |
| *** lucasagomes is now known as lucas-afk | 18:21 | |
| NobodyCam | have a good night lucas-afk | 18:21 |
| lucas-afk | o/ | 18:21 |
| openstackgerrit | Chris Krelle proposed openstack/ironic-specs: Intel Node Manager sensor collection support https://review.openstack.org/413249 | 18:27 |
| soliosg | mgould|afk: thanks for the link (name mangling is clearer now) | 18:29 |
| openstackgerrit | Annie Lezil proposed openstack/ironic-python-agent: [WIP]Reboot and Poweroff fails with coreos IPA image https://review.openstack.org/412594 | 18:35 |
| *** ijw has joined #openstack-ironic | 18:39 | |
| *** ijw has quit IRC | 18:44 | |
| openstackgerrit | Merged openstack/ironic: Turn off tempest's multitenant network tests https://review.openstack.org/412621 | 18:44 |
| *** moshele has quit IRC | 18:48 | |
| *** Nisha_Agarwal has quit IRC | 18:49 | |
| *** pcaruana has quit IRC | 18:52 | |
| *** milan has joined #openstack-ironic | 18:55 | |
| mariojv | JayF: IPA will definitely not be used with a BSD based ramdisk, right? i'm wondering if this change would have the same problem that we encountered w/ certbot: https://review.openstack.org/#/c/411818/4/ironic_python_agent/agent.py | 18:57 |
| patchbot_ | patch 411818 - ironic-python-agent - Agent: Listen for connections on both IPv4 and IPv... | 18:57 |
| jroll | fyi, grenade job is busted | 19:00 |
| jroll | this should fix it: https://review.openstack.org/#/c/416705/ | 19:00 |
| patchbot_ | patch 416705 - openstack-dev/devstack (stable/newton) - Make Nova/Ironic communication use Identity v3 | 19:00 |
| JayF | mariojv: IPA is not "alone" | 19:01 |
| JayF | mariojv: IPA is a ramdisk shipped in either a tinycorelinux or coreos container (or unsupported dib, but for sake of argument it's excluded) | 19:01 |
| JayF | mariojv: so right now, IPA only runs in linux. At some later date it could be BSD, but it'd require testing, etc, just like wedo for the other images | 19:01 |
| JayF | mariojv: so it's explicitly unsupported | 19:01 |
| JayF | does that logic make sesne? | 19:01 |
| mariojv | OK - i suppose my question should have been about whether BSD container should be supported | 19:01 |
| mariojv | yes | 19:01 |
| mariojv | thanks | 19:02 |
| JayF | the container is defined in the dockerfile in the repo :) | 19:02 |
| mariojv | ah, ok | 19:02 |
| mariojv | i haven't done a ton of creating new agent images, surprisingly enough | 19:02 |
| *** rbudden has joined #openstack-ironic | 19:03 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: DNM: testing 416705 https://review.openstack.org/416707 | 19:04 |
| jroll | put the grenade thing on the whiteboard, btw | 19:04 |
| jlvillal | jroll: Thanks. I wasn't sure if grenade was broken. I suspected so. | 19:05 |
| * sambetts -> home | 19:05 | |
| *** sambetts is now known as sambetts|afk | 19:05 | |
| jroll | jlvillal: if you see something, say something :P | 19:06 |
| jroll | nn sambetts|afk | 19:06 |
| jlvillal | jroll: I had a test job going, but yeah | 19:06 |
| *** e0ne has joined #openstack-ironic | 19:13 | |
| *** ijw has joined #openstack-ironic | 19:15 | |
| *** ChubYann has joined #openstack-ironic | 19:19 | |
| *** ijw has quit IRC | 19:24 | |
| *** e0ne has quit IRC | 19:27 | |
| *** ijw has joined #openstack-ironic | 19:28 | |
| *** e0ne has joined #openstack-ironic | 19:29 | |
| *** dprince has quit IRC | 19:30 | |
| *** dprince has joined #openstack-ironic | 19:31 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Refactor Lenovo OEM support https://review.openstack.org/416718 | 19:32 |
| *** sabeen has joined #openstack-ironic | 19:34 | |
| *** bnemec has quit IRC | 19:35 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Refactor Lenovo OEM support https://review.openstack.org/416718 | 19:35 |
| *** dtantsur|afk has quit IRC | 19:36 | |
| *** aignatov has quit IRC | 19:36 | |
| *** raginbajin has quit IRC | 19:36 | |
| *** Madasi has quit IRC | 19:36 | |
| *** jschwarz has quit IRC | 19:36 | |
| *** lifeless has quit IRC | 19:36 | |
| *** vgadiraj has quit IRC | 19:36 | |
| *** aslezil has quit IRC | 19:36 | |
| *** yarkot has quit IRC | 19:36 | |
| *** pshige_________ has quit IRC | 19:36 | |
| *** eglute has quit IRC | 19:36 | |
| *** mikal has quit IRC | 19:36 | |
| *** mrda has quit IRC | 19:36 | |
| *** bnemec has joined #openstack-ironic | 19:36 | |
| *** mikal has joined #openstack-ironic | 19:36 | |
| *** eglute has joined #openstack-ironic | 19:36 | |
| *** jschwarz has joined #openstack-ironic | 19:36 | |
| *** dtantsur has joined #openstack-ironic | 19:36 | |
| *** mrda has joined #openstack-ironic | 19:36 | |
| *** pshige_________ has joined #openstack-ironic | 19:36 | |
| *** Madasi has joined #openstack-ironic | 19:36 | |
| *** lifeless has joined #openstack-ironic | 19:36 | |
| *** aignatov has joined #openstack-ironic | 19:36 | |
| *** raginbajin has joined #openstack-ironic | 19:38 | |
| *** yarkot has joined #openstack-ironic | 19:39 | |
| * jroll meetings for the rest of the day, then out, see y'all tomorrow | 19:45 | |
| mat128 | cya | 19:48 |
| *** aslezil has joined #openstack-ironic | 19:49 | |
| *** dprince has quit IRC | 19:50 | |
| *** dprince has joined #openstack-ironic | 19:50 | |
| *** SerenaFeng has joined #openstack-ironic | 19:52 | |
| openstackgerrit | Jesse J. Cook proposed openstack/ironic: Add API and conductor methods for [un]rescue https://review.openstack.org/350831 | 19:56 |
| jcook | JayF mariojv ^ | 19:56 |
| JayF | jcook: awesome, will look now | 19:57 |
| *** vgadiraj has joined #openstack-ironic | 19:58 | |
| mariojv | thanks | 19:59 |
| mariojv | looking at that & spec update if it hasn't merged yet after my current review | 19:59 |
| *** athomas has quit IRC | 20:02 | |
| *** Sukhdev has quit IRC | 20:11 | |
| JayF | If we can get some reviews on https://review.openstack.org/#/c/407236/ -- just some minor spec updates for rescue that need some core attention to land | 20:11 |
| patchbot_ | patch 407236 - ironic-specs - Update rescue spec | 20:11 |
| mariojv | +1ed it | 20:13 |
| JayF | yeah, I put a +1 on the spec and the new api patch for rescue just now | 20:14 |
| JayF | it looks really good | 20:14 |
| jroll | JayF: one quick thing inline | 20:15 |
| mariojv | re-reviewing api/conductor patch now | 20:16 |
| jroll | or jcook I guess | 20:16 |
| JayF | jroll: look at line 75 | 20:16 |
| jroll | someone deleted too many words | 20:16 |
| JayF | jroll: we now remove rescue password when it rescues | 20:16 |
| * jroll looks | 20:16 | |
| jroll | oh weird | 20:16 |
| *** SerenaFeng has quit IRC | 20:16 | |
| jroll | can a rescue fail after that point somehow? | 20:17 |
| jroll | probably not | 20:17 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-specs: Update rescue spec https://review.openstack.org/407236 | 20:17 |
| jroll | I guess another rescue would pass the password in | 20:17 |
| JayF | exactly | 20:17 |
| JayF | so we hold the secret for the smallest amount of time possible | 20:17 |
| jroll | +2 | 20:17 |
| JayF | even though it'll be blocked by default from being pulled from the api | 20:17 |
| *** zackf has quit IRC | 20:26 | |
| *** zackf has joined #openstack-ironic | 20:26 | |
| jcook | sorry was lunching back | 20:29 |
| JayF | it's no problem, we fixed it up | 20:30 |
| JayF | lunch is important :P | 20:30 |
| jcook | cool | 20:30 |
| jcook | :D | 20:30 |
| *** zackf has quit IRC | 20:31 | |
| milan | night Ironic! :) milan -> rest | 20:34 |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Refactor Lenovo OEM support https://review.openstack.org/416718 | 20:35 |
| *** d0ugal has quit IRC | 20:39 | |
| mat128 | JayF: commented inline about an additional security risk that this spec brings in | 20:40 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Adding InfiniBand Support https://review.openstack.org/264257 | 20:41 |
| *** milan has quit IRC | 20:41 | |
| palmertime | Im having a time out issue with Baremetal deploy with driver pxe_ipmitool. After install the systems power off but fail to power back on. Is there a way to increase the tries or timeout for this? Maybe increase the time between tries? Here is the error log entry and the manually repeated ipmitool command. https://gist.github.com/palmertime/20f9ffad925fcaa276b764df927b0abf | 20:41 |
| mat128 | palmertime: https://git.openstack.org/cgit/openstack/ironic/tree/etc/ironic/ironic.conf.sample#n1744 | 20:42 |
| mat128 | exactly what you are looking for :) retry_timeout and min_command_interval | 20:42 |
| palmertime | mat128: Perfect, thanks for the info! I'll try that out. | 20:43 |
| *** Goneri has quit IRC | 20:44 | |
| mariojv | jcook: left some comments on https://review.openstack.org/#/c/350831 | 20:44 |
| patchbot_ | patch 350831 - ironic - Add API and conductor methods for [un]rescue | 20:44 |
| mariojv | actually, that validate() method may need to be added to the base classes too | 20:46 |
| mariojv | i'll also add that as a follow-up comment | 20:47 |
| jkilpatr | If I wanted to gather performance metrics on ironic what should I watch most carefully? I've already got timings on whole operations and system metrics like processes and cpu usage. | 20:48 |
| *** ijw has quit IRC | 20:49 | |
| mariojv | jkilpatr: i'd suggest taking a look at this if you haven't yet http://docs.openstack.org/developer/ironic/deploy/metrics.html | 20:51 |
| jkilpatr | mariojv, thanks. | 20:52 |
| mariojv | np | 20:53 |
| *** d0ugal has joined #openstack-ironic | 20:55 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic-inspector: Update documentation to deploy Ironic Inspector with DevStack https://review.openstack.org/410456 | 21:01 |
| *** jcoufal has quit IRC | 21:10 | |
| openstackgerrit | Matt Riedemann proposed openstack/ironic: DNM: Test Ironic in Newton with the placement service https://review.openstack.org/416734 | 21:10 |
| soliosg | How can I delete a node not seen by nova anymore? Shown by 'openstack baremetal node list' but not by 'openstack server list' | 21:16 |
| *** vgadiraj has quit IRC | 21:16 | |
| *** e0ne has quit IRC | 21:17 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Refactor Lenovo OEM support https://review.openstack.org/416718 | 21:18 |
| *** jjohnson2 has joined #openstack-ironic | 21:19 | |
| mrda | Morning Ironic | 21:23 |
| *** fritzstauffacher has quit IRC | 21:24 | |
| JayF | mat128: did you read security impact? | 21:27 |
| mat128 | JayF: I found something new :) | 21:29 |
| mat128 | oh | 21:29 |
| *** ijw has joined #openstack-ironic | 21:29 | |
| JayF | mat128: Are you certain? I read your comment and we addressed it pretty clearly | 21:30 |
| mat128 | just read your comment in the review | 21:30 |
| mat128 | I havent read the whole spec | 21:30 |
| mat128 | just the parts you changed | 21:30 |
| mat128 | my bad :( | 21:30 |
| JayF | yeah I occassionally fall into that trap too | 21:30 |
| JayF | no problem :) | 21:30 |
| mat128 | "This means deployers may need to ensure no secret information (such as custom cleaning steps or firmwares) are not present in the rescue ramdisk." | 21:31 |
| mat128 | exactly this | 21:31 |
| * mat128 failed hard on this one | 21:31 | |
| *** dsneddon is now known as dsneddon_afk | 21:31 | |
| JayF | at our downstream version of this, for Rackspace, not only do we use a different ramdisk, but we wrote a small IPA clone in go to keep the rescue image small (and instead fill it with a useful operator tooling container) | 21:32 |
| JayF | now I think that was probably not the best idea, but we know about this problem in particular :) | 21:32 |
| mat128 | I didnt think you were running this feature already | 21:33 |
| mat128 | good to know | 21:33 |
| JayF | Our version is not as good as this, honestly | 21:33 |
| JayF | but we've had working rescue downstream in onmetal for a long time | 21:33 |
| JayF | sorta a pattern we followed for a while: have a crappy implementation of something downstream in onmetal; learn lessons; upstream it | 21:33 |
| mat128 | :) | 21:33 |
| *** Goneri has joined #openstack-ironic | 21:34 | |
| *** mjturek has joined #openstack-ironic | 21:41 | |
| *** zackf has joined #openstack-ironic | 21:50 | |
| openstackgerrit | Bertrand Lallau proposed openstack/ironic: Remove netaddr useless requirement https://review.openstack.org/416743 | 21:59 |
| *** mriedem has joined #openstack-ironic | 22:01 | |
| mriedem | hey, is this a known NoValidHost that people see in ironic grenade jobs? http://logs.openstack.org/23/415523/1/check/gate-grenade-dsvm-ironic-ubuntu-xenial/1a25aa5/logs/new/screen-n-sch.txt.gz#_2017-01-04_16_39_36_078 | 22:01 |
| mriedem | warnings before that about compute hosts being down | 22:02 |
| *** sabeen has quit IRC | 22:03 | |
| JayF | mriedem: I believe our gate is currently broken pending a fix in devstack | 22:06 |
| JayF | mriedem: one second | 22:06 |
| JayF | mriedem: https://review.openstack.org/#/c/416705/ | 22:07 |
| patchbot_ | patch 416705 - openstack-dev/devstack (stable/newton) - Make Nova/Ironic communication use Identity v3 | 22:07 |
| palmertime | mat128: That answered my question and helped to extend the time between checks. However, didn't solve the problem. After the image is installed that the system is powered off, it does not start again. | 22:08 |
| *** dprince has quit IRC | 22:09 | |
| mriedem | JayF: so the ironic grenade job is keystone v3 only? | 22:09 |
| mriedem | looks like the backport is failing on newton | 22:10 |
| palmertime | mat128: I should add … after install it changes the boot order, powers off the system and executes the power on command but it doesn't power on. | 22:10 |
| *** trown is now known as trown|outtypewww | 22:11 | |
| palmertime | mat128: I think i'm going to look for a setting to wait longer between poweroff and poweron. | 22:11 |
| *** lindycoder has quit IRC | 22:12 | |
| JayF | mriedem: we were using a deprecated mechanism to configure nova, and apparently when it was removed from nova, it broke the "master" end of grenade | 22:16 |
| JayF | mriedem: jroll wrote the fix, but I can rework it if it's not the right way to go | 22:16 |
| mriedem | i haven't dug into the newton failures yet | 22:17 |
| mriedem | might be unrelated, but scary that many jobs would fail in unrelated ways | 22:17 |
| JayF | the newton failures on that patch appeared to be found by e-r, at least one of them | 22:17 |
| JayF | tbh, this is anecdotal as hell, but since coming back for the holidays, we've had more unexplained failures | 22:18 |
| JayF | I think probably since the resource tracking stuff hit, and we had that set of breakage | 22:18 |
| *** rbudden has quit IRC | 22:18 | |
| JayF | when I see 'em fail I try to dig and explain it | 22:18 |
| *** dyasny has quit IRC | 22:20 | |
| *** pmannidi has joined #openstack-ironic | 22:21 | |
| *** pmannidi has quit IRC | 22:22 | |
| *** pmannidi has joined #openstack-ironic | 22:22 | |
| *** jkilpatr has quit IRC | 22:25 | |
| mriedem | well the placement api stuff shouldn't be causing issues right now since ironic isn't running with it on master, | 22:25 |
| mriedem | and wasn't in newton | 22:25 |
| mriedem | http://logs.openstack.org/05/416705/1/check/gate-grenade-dsvm-forward-ubuntu-trusty/fac7ba1/logs/new/screen-n-api.txt.gz#_2017-01-04_19_49_36_411 is clearly a known unrelated issue | 22:27 |
| *** jheroux has quit IRC | 22:27 | |
| *** baoli has quit IRC | 22:29 | |
| *** baoli has joined #openstack-ironic | 22:30 | |
| mriedem | the other voting job that failed was using libvirt, so yeah it's probably a safe recheck | 22:33 |
| *** baoli has quit IRC | 22:33 | |
| mriedem | yup, http://logs.openstack.org/05/416705/1/check/gate-tempest-dsvm-neutron-linuxbridge-ubuntu-xenial/37a1be9/logs/screen-n-cpu.txt.gz#_2017-01-04_19_31_20_883 | 22:34 |
| mriedem | rebuild failed which is a cast, so server never got back to ACTIVE status | 22:34 |
| *** baoli has joined #openstack-ironic | 22:36 | |
| openstackgerrit | Annie Lezil proposed openstack/ironic-python-agent: [WIP]Reboot and Poweroff fails with coreos IPA image https://review.openstack.org/412594 | 22:37 |
| *** baoli has quit IRC | 22:40 | |
| jlvillal | JayF: FYI: I pushed up a test stable/newton devstack patch to see if it is broken. https://review.openstack.org/#/c/416763/ | 22:46 |
| patchbot_ | patch 416763 - openstack-dev/devstack (stable/newton) - WIP/DNM: Check if stable/newton builds | 22:46 |
| mriedem | jlvillal: me too, in a different context :) https://review.openstack.org/#/c/416734/ | 22:49 |
| patchbot_ | patch 416734 - ironic (stable/newton) - DNM: Test Ironic in Newton with the placement service | 22:49 |
| jlvillal | mriedem: :) | 22:50 |
| *** openstack has joined #openstack-ironic | 22:55 | |
| palmertime | Manual execution of ipmitool works without issue | 22:55 |
| *** Goneri has quit IRC | 22:55 | |
| JayF | palmertime: but other ipmi commands, during other parts of the install, work fine? | 22:56 |
| palmertime | yes | 22:56 |
| JayF | that's very strange. | 22:56 |
| *** bfournie has quit IRC | 22:56 | |
| JayF | If you log at debug, I'm fairly sure it'll log the exact commandlines it's using | 22:56 |
| palmertime | It's like it logs the execution of up/on but doesn't actually execute | 22:57 |
| JayF | well, ipmi is a notoriously bad protocol :) I've had machines before where on/off commands return but don't actually work on the node | 22:57 |
| JayF | I'd log at debug, get the exact commands and ordering, and try to duplicate that yourself | 22:57 |
| palmertime | Ok, Set debug in ironic.conf? | 22:58 |
| JayF | my hunch is that there's some quirk of your hardware where when we send the timeout disable raw bytes or the change boot device it's causing your ipmi to act strangely | 22:58 |
| JayF | yeah, log level to debug | 22:58 |
| JayF | let me make sure it's debug level, it might not be that high | 22:58 |
| JayF | sec | 22:58 |
| jlvillal | mriedem: Have you ever seen something like this? http://logs.openstack.org/20/407220/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv/0779f94/logs/screen-n-cond.txt.gz?level=TRACE | 22:59 |
| JayF | palmertime: yeah, it's debug level: http://logs.openstack.org/94/412594/13/check/gate-tempest-dsvm-ironic-ipa-partition-agent_ipmitool-coreos-src-ubuntu-xenial/eb52cbd/logs/screen-ir-cond.txt.gz#_2017-01-04_18_57_06_623 (as an example of what I'm talking about) | 22:59 |
| mriedem | jlvillal: hmm, compute.free_disk_gb is None | 22:59 |
| jlvillal | mriedem: Yeah... | 22:59 |
| mriedem | is that master? | 23:00 |
| jlvillal | mriedem: And maybe I should have asked you over in #openstack-nova | 23:00 |
| *** amotoki has quit IRC | 23:00 | |
| jlvillal | Yep | 23:00 |
| mriedem | it's pretty quiet over there | 23:00 |
| mriedem | plus i'm never in this channel.. :) | 23:00 |
| jlvillal | mriedem: https://review.openstack.org/#/c/407220/ | 23:00 |
| patchbot_ | patch 407220 - ironic - WIP/DNM: Testing gate | 23:00 |
| JayF | palmertime: https://github.com/openstack/ironic/blob/master/etc/ironic/ironic.conf.sample#L380 should be all you need to set it as debug | 23:00 |
| openstackgerrit | Jesse J. Cook proposed openstack/ironic: Add API and conductor methods for [un]rescue https://review.openstack.org/350831 | 23:00 |
| jcook | mariojv ^ | 23:01 |
| mriedem | jlvillal: is http://logs.openstack.org/20/407220/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv/0779f94/logs/screen-n-cond.txt.gz?level=TRACE#_2017-01-04_19_05_16_434 coming from something that ironic hooks into the nova scheduler? | 23:02 |
| *** amotoki has joined #openstack-ironic | 23:02 | |
| jlvillal | mriedem: I'm not sure. And I just started a 1:1 with my manager now :( | 23:03 |
| *** rbudden has joined #openstack-ironic | 23:03 | |
| mriedem | ask them if they know | 23:03 |
| JayF | mriedem: I'm looking now, sec | 23:03 |
| jlvillal | mriedem: Try to sync up tomorrow :) | 23:03 |
| JayF | gate troubleshooting by comittee | 23:03 |
| mriedem | i don't see "Failed to compute_task_build_instances" anywhere in nova | 23:03 |
| JayF | mriedem: https://github.com/openstack/nova/blob/master/nova/scheduler/ironic_host_manager.py it's in the nova tree | 23:03 |
| JayF | mriedem: that's coming from tempest, I think? | 23:04 |
| mariojv | thanks jcook i'll take a look tomorrow | 23:04 |
| mriedem | yeah i see that https://github.com/openstack/nova/blob/master/nova/scheduler/ironic_host_manager.py#L46 | 23:04 |
| *** ijw has quit IRC | 23:05 | |
| mriedem | http://logs.openstack.org/20/407220/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv/0779f94/logs/screen-n-sch.txt.gz?level=TRACE#_2017-01-04_19_05_16_400 | 23:07 |
| mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22_update_from_compute_node%5C%22%20AND%20message%3A%5C%22TypeError%3A%20unsupported%20operand%20type(s)%20for%20*%3A%20'NoneType'%20and%20'int'%5C%22%20AND%20tags%3A%5C%22screen-n-sch.txt%5C%22&from=7d | 23:10 |
| JayF | mriedem: could that be caused by the other bug that I pointed out? where nova/ironic can't communicate? | 23:10 |
| mriedem | wtf, that doesn't look good | 23:10 |
| mriedem | which one? keystone v3? | 23:11 |
| palmertime | JayF: Ok, that the boot device command is giving me problems. | 23:11 |
| palmertime | JayF: Error: Unable to establish IPMI v2 / RMCP+ session | 23:11 |
| palmertime | JayF: Will start searching around with that error. Thanks for the help | 23:11 |
| mriedem | does gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv only use keystone v3? | 23:12 |
| palmertime | JayF: This looks interesting, https://access.redhat.com/solutions/1981303 | 23:13 |
| mriedem | https://review.openstack.org/#/c/405597/ just merged today so that's not it | 23:13 |
| patchbot_ | patch 405597 - nova - conf: remove deprecated ironic options (MERGED) | 23:13 |
| *** sabeen has joined #openstack-ironic | 23:13 | |
| *** vgadiraj has joined #openstack-ironic | 23:20 | |
| mriedem | JayF: found an older bug that looks related https://bugs.launchpad.net/nova/+bug/1515870 | 23:20 |
| openstack | Launchpad bug 1515870 in OpenStack Compute (nova) "server can not launch while there's a new nova compute node registed failed" [Low,In progress] - Assigned to Eric Brown (ericwb) | 23:20 |
| mriedem | it's not the ironic host manager in this case, | 23:20 |
| mriedem | but same type of failure | 23:20 |
| mriedem | i suspect the compute node is getting created in the database before update_available_resource has filled out it's actual resource info | 23:21 |
| mriedem | checking to see when that thing is created in the n-cpu service | 23:23 |
| mriedem | right before that failure i see this http://logs.openstack.org/20/407220/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv/0779f94/logs/screen-n-cpu.txt.gz#_2017-01-04_19_03_55_573 | 23:26 |
| *** amotoki has quit IRC | 23:26 | |
| mriedem | that's not the same compute node though, that's compute node 1, the thing that blows up is compute node 8 | 23:28 |
| mriedem | hmm, i don't know, would probably need jroll or jaypipes to take a look at this | 23:30 |
| *** amotoki has joined #openstack-ironic | 23:31 | |
| palmertime | JayF: Well, first attempt with the setting recommended here, https://access.redhat.com/solutions/1981303 seemed to have helped. | 23:32 |
| JayF | palmertime: nice! I'm glad it helped. | 23:33 |
| *** bfournie has joined #openstack-ironic | 23:41 | |
| mriedem | JayF: well, i can see where free_disk_gb is set on the ComputeNode record after the CN is created, which could maybe explain how the scheduler pulled that from the db and blew up on it | 23:42 |
| JayF | mriedem: so is that an ironic problem? ironic driver problem? nova problem? | 23:42 |
| JayF | I'm not super familiar with that part of nova, honestly | 23:42 |
| mriedem | you and me both | 23:42 |
| mriedem | the resource tracker is really jaypipes' domain | 23:43 |
| JayF | it's nice to know something is going on though | 23:43 |
| JayF | should we file a bug for this and get an e-r going? | 23:43 |
| JayF | so other folks know what's going on | 23:43 |
| mriedem | well i'd re-use https://bugs.launchpad.net/nova/+bug/1515870 as it's a similar failure in the non-ironic host manager | 23:44 |
| openstack | Launchpad bug 1515870 in OpenStack Compute (nova) "server can not launch while there's a new nova compute node registed failed" [Medium,Confirmed] | 23:44 |
| mriedem | same type of race though it looks like | 23:44 |
| mriedem | the thing that puzzles me is how often this hits in logstash, | 23:44 |
| mriedem | but it should be a pretty small window, we create the compute node record in the db, then set the free_disk_gb field after that and update it in the db | 23:44 |
| mriedem | so the scheduler would have to pull the record in between those two things | 23:44 |
| JayF | well, unless with ironic | 23:45 |
| JayF | adding/removing nodes as they clean | 23:45 |
| JayF | is causing that race to happen multiple times | 23:45 |
| JayF | that'd be possible, no? | 23:45 |
| mriedem | once free_disk_gb is set to something in the db i'm not seeing anything that would reset that back to None though | 23:47 |
| mriedem | unless it's just a brand new compute node object everytime | 23:47 |
| mriedem | with non-ironic things that's not really an issue, we create the compute node record once when nova-compute starts up | 23:47 |
| mriedem | looks like the non-Ironic case was fixed https://review.openstack.org/#/c/352317/ | 23:49 |
| patchbot_ | patch 352317 - nova - Avoid update resource if compute node not updated (MERGED) | 23:49 |
| mriedem | that same fix wouldn't work for ironic because updated_at is set in this failure | 23:50 |
| mriedem | http://logs.openstack.org/20/407220/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-multinode-ubuntu-xenial-nv/0779f94/logs/screen-n-sch.txt.gz#_2017-01-04_19_05_16_398 | 23:50 |
| mriedem | updated_at=2017-01-04T19:04:51Z | 23:50 |
| mriedem | fails at: 2017-01-04 19:05:16.400 | 23:51 |
| JayF | that makes sense then | 23:52 |
| JayF | so my hunch is maybe correct then/ | 23:52 |
| JayF | sorry I'm doing two things at once :/ | 23:52 |
| mriedem | JayF: ok i created a new bug https://bugs.launchpad.net/nova/+bug/1654102 | 23:55 |
| openstack | Launchpad bug 1654102 in OpenStack Compute (nova) "Ironic: TypeError: unsupported operand type(s) for *: 'NoneType' and 'int' - during select_destinations()" [Undecided,New] | 23:55 |
| *** ijw has joined #openstack-ironic | 23:55 | |
| mriedem | it's not causing voting jobs to fail so an e-r query wouldn't probably be accepted | 23:57 |
| JayF | is it just the multinode jobs doing this? | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!