| opendevreview | Steve Baker proposed openstack/bifrost master: Support PXE network boot with grub https://review.opendev.org/c/openstack/bifrost/+/807220 | 00:53 |
|---|---|---|
| opendevreview | Ghanshyam proposed openstack/sushy master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/sushy/+/821637 | 01:40 |
| opendevreview | Ghanshyam proposed openstack/ironic-python-agent master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic-python-agent/+/821638 | 01:42 |
| opendevreview | Ghanshyam proposed openstack/metalsmith master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/metalsmith/+/821640 | 01:45 |
| opendevreview | Ghanshyam proposed openstack/ironic master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic/+/821641 | 01:48 |
| opendevreview | Steve Baker proposed openstack/ironic master: Use driver_internal_info methods for driver utils https://review.opendev.org/c/openstack/ironic/+/818505 | 02:34 |
| opendevreview | Steve Baker proposed openstack/ironic master: Use driver_internal_info methods for drac driver https://review.opendev.org/c/openstack/ironic/+/818506 | 02:34 |
| opendevreview | Steve Baker proposed openstack/ironic master: Use driver_internal_info methods for ilo driver https://review.opendev.org/c/openstack/ironic/+/818507 | 02:34 |
| opendevreview | Steve Baker proposed openstack/ironic master: Use driver_internal_info methods for redfish driver https://review.opendev.org/c/openstack/ironic/+/818508 | 02:34 |
| opendevreview | Steve Baker proposed openstack/ironic master: Use driver_internal_info methods for other drivers https://review.opendev.org/c/openstack/ironic/+/818509 | 02:34 |
| opendevreview | Ghanshyam proposed openstack/python-ironicclient master: Updating python testing classifier as per Yoga testing runtime https://review.opendev.org/c/openstack/python-ironicclient/+/819241 | 02:49 |
| opendevreview | Merged openstack/ironic master: Document single process Ironic https://review.opendev.org/c/openstack/ironic/+/820396 | 03:05 |
| opendevreview | Verification of a change to openstack/ironic master failed: Adoption: do not validate boot interface when local booting https://review.opendev.org/c/openstack/ironic/+/821537 | 04:47 |
| *** sshnaidm|afk is now known as sshnaidm | 06:57 | |
| rpittau | good morning ironic! o/ | 07:44 |
| arne_wiebalck | Good morning rpittau and Ironic! | 07:55 |
| rpittau | hey arne_wiebalck :) | 07:55 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-prometheus-exporter master: Use only Yoga tests https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/820321 | 07:59 |
| opendevreview | Riccardo Pittau proposed openstack/python-ironicclient master: Use only Yoga tests https://review.opendev.org/c/openstack/python-ironicclient/+/820325 | 08:01 |
| rpittau | very quick review if anyone has a moment https://review.opendev.org/c/openstack/sushy/+/821081 | 08:02 |
| *** amoralej|off is now known as amoralej | 08:15 | |
| dtantsur | morning ironic | 08:15 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-inspector master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic-inspector/+/821659 | 08:17 |
| dtantsur | rpittau: if you have a second: https://review.opendev.org/c/openstack/ironic/+/820904/ | 08:19 |
| rpittau | ack, done | 08:19 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: [trivial] add python 3.9 in classifier https://review.opendev.org/c/openstack/bifrost/+/821660 | 08:21 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-lib master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic-lib/+/821661 | 08:22 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-tempest-plugin master: [trivial] add python 3.9 in classifier https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/821664 | 08:26 |
| opendevreview | Riccardo Pittau proposed openstack/metalsmith master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/metalsmith/+/821665 | 08:27 |
| opendevreview | Verification of a change to openstack/ironic master failed: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic/+/821641 | 08:32 |
| rpittau | docs tests are foobar ? | 08:32 |
| rpittau | weird, that was passing a couple of hours ago | 08:32 |
| rpittau | and can't reproduce it locally | 08:33 |
| opendevreview | Merged openstack/metalsmith master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/metalsmith/+/821640 | 08:35 |
| opendevreview | Riccardo Pittau proposed openstack/networking-baremetal master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/networking-baremetal/+/821667 | 08:37 |
| opendevreview | Verification of a change to openstack/ironic master failed: Adoption: do not validate boot interface when local booting https://review.opendev.org/c/openstack/ironic/+/821537 | 08:37 |
| opendevreview | Riccardo Pittau proposed openstack/networking-generic-switch master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/networking-generic-switch/+/821669 | 08:38 |
| opendevreview | Verification of a change to openstack/sushy master failed: [trivial] fix Xena release versions https://review.opendev.org/c/openstack/sushy/+/821081 | 08:40 |
| opendevreview | Riccardo Pittau proposed openstack/python-ironic-inspector-client master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/python-ironic-inspector-client/+/821671 | 08:41 |
| opendevreview | Riccardo Pittau proposed x/sushy-oem-idrac master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/x/sushy-oem-idrac/+/821672 | 08:42 |
| opendevreview | Riccardo Pittau proposed openstack/virtualbmc master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/virtualbmc/+/821673 | 08:43 |
| opendevreview | Verification of a change to openstack/ironic master failed: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic/+/821641 | 08:52 |
| opendevreview | Merged openstack/sushy master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/sushy/+/821637 | 09:13 |
| dtantsur | TheJulia: hey, where have we left your locking rework patch? "nobody hecking reviews it"? | 09:16 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Fix Mock objects as specs https://review.opendev.org/c/openstack/ironic/+/821370 | 10:00 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Enable foreign keys in SQLite https://review.opendev.org/c/openstack/ironic/+/821219 | 10:01 |
| opendevreview | Merged openstack/ironic master: Use more granular lock in BaseDriverFactory https://review.opendev.org/c/openstack/ironic/+/820904 | 10:10 |
| opendevreview | Merged openstack/ironic master: [doc] Adoption with Nova https://review.opendev.org/c/openstack/ironic/+/821611 | 10:10 |
| opendevreview | Merged openstack/ironic-python-agent master: Re-add python 3.6/3.7 in classifier https://review.opendev.org/c/openstack/ironic-python-agent/+/821638 | 10:10 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [doc] Fix broken link in "Node Deployment" https://review.opendev.org/c/openstack/ironic/+/821685 | 10:12 |
| opendevreview | Verification of a change to openstack/ironic master failed: Allow enabling fast-track per node https://review.opendev.org/c/openstack/ironic/+/821010 | 10:17 |
| opendevreview | Verification of a change to openstack/ironic master failed: Allow enabling fast-track per node https://review.opendev.org/c/openstack/ironic/+/821010 | 11:07 |
| *** outbrito_ is now known as outbrito | 12:22 | |
| *** amoralej is now known as amoralej|lunch | 13:16 | |
| arne_wiebalck | Bare Metal SIG meeting in about 20mins with dtantsur and a "Metal3 Overview (with demo)"! All details in https://etherpad.opendev.org/p/bare-metal-sig, everyone is welcome! | 13:39 |
| TheJulia | dtantsur: nobody reviewed it and I think i eventually abandoned it | 13:41 |
| TheJulia | dtantsur: I stopped updating it at least because I was just out of time | 13:42 |
| dtantsur | I see | 13:42 |
| dtantsur | and good morning | 13:43 |
| TheJulia | good morning | 13:43 |
| dtantsur | one of my plans for pre-x-mas weeks: write a fake sushy-tools backend with a fake IPA | 13:43 |
| dtantsur | so that it simply acks all commands and implements them with time.sleep(60) | 13:43 |
| dtantsur | so that I can load-test real Ironic without having thousands of VMs | 13:43 |
| dtantsur | thoughts? | 13:43 |
| TheJulia | You could do that, but are you just focusing on cuncurrent deployment? | 13:44 |
| dtantsur | well, other things are somewhat easier to measure | 13:44 |
| dtantsur | deployments/cleanings/inspections... | 13:44 |
| dtantsur | I want to be able to answer question "what actually happens if someone starts ~ 500 virtual media deployments at ~ the same time" | 13:45 |
| TheJulia | oh jeeze | 13:48 |
| * TheJulia twitches | 13:48 | |
| TheJulia | I think that could drive us to creating a queing mechanism | 13:49 |
| TheJulia | dtantsur: https://www.youtube.com/watch?v=3fcaXmpZydI | 13:50 |
| * TheJulia is glad the migraine is gone now | 13:55 | |
| dtantsur | \o/ | 13:56 |
| TheJulia | Next doctors appointment I'm going to have to mention them | 13:56 |
| TheJulia | although they will likely just say "wear your glasses" | 13:56 |
| arne_wiebalck | Bare Metal SIG meeting starting now. | 14:00 |
| *** amoralej|lunch is now known as amoralej | 14:04 | |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Fix Mock objects as specs https://review.opendev.org/c/openstack/ironic/+/821370 | 14:09 |
| opendevreview | Merged openstack/sushy master: [trivial] fix Xena release versions https://review.opendev.org/c/openstack/sushy/+/821081 | 14:19 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Fix Mock objects as specs https://review.opendev.org/c/openstack/ironic/+/821370 | 14:34 |
| arne_wiebalck | Thanks a lot dtantsur, very instructive, and cool live demo! | 14:37 |
| dtantsur | I hope you liked it :) | 14:37 |
| arne_wiebalck | big time :) | 14:37 |
| dtantsur | okay, now I can use a big cup of tea | 14:41 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Fix Mock objects as specs https://review.opendev.org/c/openstack/ironic/+/821370 | 15:31 |
| cvstealth | Hello folks, I was curious if support for port groups works in combination with the network interface being neutron? When looking at the neutron port for a deployed instance I only see one NIC and grabbing the network_data.json out of Nova metadata there is only a single interface from the port group that is defined in Ironic for the particular node. | 16:02 |
| dtantsur | cvstealth: I think it *only* works with the neutron network interface.. but you need to create the port the way you expect it? | 16:03 |
| cvstealth | What I'm seeing is that the instance comes up with only a single interface that properly DHCPs but I don't see any parts of the bond being assembled. I was under the impression those pieces came down from the network_data in the metadata payload. | 16:06 |
| dtantsur | yeah, Nova should pass them there | 16:08 |
| cvstealth | Just wanted to make sure that was the case and I didn't need to throw anything in the userdata of the Nova instance to properly get it to bond. | 16:08 |
| dtantsur | I'm not exactly expert on this part of ironic though.. TheJulia do you remember better? | 16:08 |
| * TheJulia is summoned | 16:09 | |
| TheJulia | yeah, it has to be articulated in neutron for the instance, that information should get shipped down. This is where I wish sambetts was still around. | 16:10 |
| cvstealth | During the deploy process it does create 2 neutron ports for the nodes, so the standalone ports attribute does look to be functional and my ports seemed to be defined properly. Once it hits the switch_to_tenant_network step then I only see 1 port defined in Neutron. | 16:10 |
| TheJulia | the two neutron ports are being created by what? ironic? | 16:11 |
| TheJulia | for deployment yes? | 16:11 |
| TheJulia | So switch_to_tenant_network uses the VIF port record which is supplied by the user which needs to articulate the desired config AIUI | 16:11 |
| cvstealth | All port creation is being done by Ironic, I'm not pre-creating any type of ports to be assigned to the instance. | 16:12 |
| TheJulia | ahh, there is the problem | 16:12 |
| TheJulia | you need to create vifs as you need them, and then request them to be attached | 16:12 |
| dtantsur | yep | 16:12 |
| TheJulia | you can use ironic's portgroups in terms of articulating switch config and bonding so ironic doesn't do something stupid port attachment wise, and that info can be shipped to neutron, but under normal circumstances it will still be a single vif to a port | 16:12 |
| TheJulia | portgroups *are* a bit weird in that they serve as both an information source and possible configuration mapping | 16:13 |
| cvstealth | I need to do a bit more research on that, at a high level I need to create a neutron port with a bonding type of configuration then it gets consume when the user facing instance is created? | 16:17 |
| TheJulia | and kind of yes | 16:19 |
| TheJulia | err | 16:19 |
| TheJulia | s/and// | 16:19 |
| TheJulia | that neutron port is represented as a vif id or in other words the neutron port uuid | 16:19 |
| TheJulia | you feed that to nova when you ask for the instance | 16:19 |
| dtantsur | meanwhile, we have a potential transient ut failure: https://de2803ac5f1730f38621-c66b78121e2092583b9dfc27e6e3a4e3.ssl.cf1.rackcdn.com/821010/3/check/openstack-tox-cover/7d3b2ba/testr_results.html | 16:22 |
| TheJulia | I've seen that once before | 16:23 |
| TheJulia | happening more fequently? | 16:23 |
| dtantsur | it's the first time I see it. just making sure to mention it. | 16:23 |
| cvstealth | TheJulia: Is there a difference between the static port assignment that is fed into the Nova instance vs just booting the instance with "--network XXXX"? Sorry for the ignorance on the topic just didn't see it mentioned on the port groups documentation for creating the vif ahead of time. | 16:29 |
| TheJulia | cvstealth: there is, just sayign the network only creates a port and attaches it | 16:31 |
| TheJulia | at least, int he nova code aiui | 16:31 |
| cvstealth | On the deployed instance with the --network path, doing a node vif list it does show the neutron port for the user facing instance. | 16:31 |
| TheJulia | cvstealth: yeah, I would expect that | 16:31 |
| TheJulia | cvstealth: nova has ?5? different paths it can take on networking for an instance | 16:32 |
| TheJulia | one of being just give us a network, and we'll create a plain neutron port for it | 16:32 |
| TheJulia | others being give me a port based paths | 16:32 |
| TheJulia | I think ther eis also a "just give me an instance" and it guesses, or maybe that is configurable | 16:33 |
| TheJulia | I've never tried that. | 16:33 |
| cvstealth | ok thanks I'll give that a try to see what amount of broken pieces I get, thanks for the help. | 16:33 |
| dtantsur | the snmp failure makes no sense to me... | 16:58 |
| *** amoralej is now known as amoralej|off | 17:26 | |
| TheJulia | dtantsur: cross thread unhappiness? | 17:29 |
| dtantsur | possibly | 17:29 |
| TheJulia | it happens, we see it from time to time with sqlalchemy | 17:29 |
| TheJulia | At like the worst possible time | 17:30 |
| dtantsur | and our unit tests absolutely don't expect foreign keys to work.... | 17:31 |
| dtantsur | Failed: 1611 | 17:34 |
| dtantsur | well, this is going to be enjoyable (no) | 17:35 |
| dtantsur | `node_id=fake_db_allocation['id']` ehmm, no, I don't think so :) | 17:37 |
| TheJulia | well, we're backing in a lot of the data to set the scenario | 17:46 |
| TheJulia | also, I believe since they would have used sqllite for base unit testing, we never could just expect it to magically already be a thing | 17:49 |
| arne_wiebalck | heh, seems we have a node with a redfish endpoint where the etag value is not allowed to be inside double quotes (not allowed meaning: -H 'If-Match: abc' works, while -H 'If-Match: "abc"' does not) ... I think this is against how etags are supposed to be used. | 17:52 |
| arne_wiebalck | so far I have only dealt with 2 node types and redfish, but the universe of interpretations and issues seems to unfold quite nicely already | 17:53 |
| TheJulia | Out of curiosity, anyone up for reviewing https://review.opendev.org/c/openstack/sushy/+/821324 ? | 17:53 |
| TheJulia | so umm.. hmm | 17:54 |
| dtantsur | arne_wiebalck: UGH. U.G.H. | 17:54 |
| arne_wiebalck | Ironic is ofc not happy about this :) | 17:55 |
| * arne_wiebalck Ironic goes ugh as well. | 17:55 | |
| TheJulia | ugh | 17:55 |
| TheJulia | rfc7232 says yes, required | 17:55 |
| dtantsur | arne_wiebalck: time for the dmtf forum \o/ | 17:55 |
| TheJulia | Section 2.3 of RFC7232, last sentence of the first paragraph | 17:56 |
| arne_wiebalck | "An entity-tag consists of an opaque quoted string, possibly prefixed by a weakness indicator." | 17:57 |
| arne_wiebalck | TheJulia: yes | 17:57 |
| arne_wiebalck | dtantsur: I think this is for the vendor to fix, let's see what they say ... maybe this is fixed already in a newer version of the f/w. | 17:58 |
| dtantsur | arne_wiebalck: yep. if they don't react, use dmtf-shaming! | 17:59 |
| arne_wiebalck | dtantsur: ha! | 17:59 |
| arne_wiebalck | ok, have a nice evening everyone, see you tomorrow o/ | 18:02 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Enable foreign keys in SQLite https://review.opendev.org/c/openstack/ironic/+/821219 | 18:11 |
| dtantsur | see you tomorrow o/ | 18:11 |
| rpittau | good night! o/ | 18:15 |
| NobodyCam | Good AfterNoon Ironic'ers | 20:35 |
| *** dking is now known as Guest8470 | 21:19 | |
| *** Guest8470 is now known as dking | 21:19 | |
| TheJulia | Good afternoon NobodyCam! | 21:53 |
| TheJulia | how goes the... wet stuff falling from the sky | 21:54 |
| cvstealth | TheJulia / dtantsur : Thanks again for the help from earlier, I was able to get the configuration working by pre-allocating a Neutron port and using a configdrive. | 22:08 |
| jamesdenton | Are "openstack baremetal node inspect" and "openstack baremetal introspection {start}" related in some way? Like, would the process be to set it to 'inspect' then perform an introspection? Is this what potentially triggers a network vlan change with NGS? | 22:13 |
| TheJulia | cvstealth: \o/ | 22:18 |
| TheJulia | janders: so... "baremetal introspection" call inspector directly | 22:19 |
| TheJulia | "baremetal node inspect" call ironic and ironic drives it | 22:19 |
| TheJulia | jamesdenton: to do introspection networks with NGS, you'd need to use it through ironic and not inspector since inspector largely ifmemory assumes only thinks about power turning on | 22:20 |
| TheJulia | .,.. I think | 22:20 |
| stevebaker[m] | arne_wiebalck dtantsur : The metal3 SIG video is liiive https://www.youtube.com/watch?v=rjSC6cJ9YY8 | 23:32 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!