| *** pmannidi is now known as pmannidi|SA | 00:30 | |
| iurygregory | good morning Ironic | 06:20 |
|---|---|---|
| iurygregory | hey janders o/ | 06:35 |
| janders | good morning iurygregory o/ | 06:35 |
| iurygregory | good afternoon/night for you :D | 06:36 |
| arne_wiebalck | Good morning janders iurygregory and Ironic! | 06:49 |
| iurygregory | morning arne_wiebalck o/ | 06:49 |
| *** rpittau|afk is now known as rpittau | 07:08 | |
| rpittau | good morning ironic! o/ | 07:08 |
| iurygregory | morning rpittau o/ | 07:09 |
| rpittau | hey iurygregory :) | 07:09 |
| iurygregory | in about 1hr our CI might be back to normal :D (if https://review.opendev.org/c/openstack/devstack/+/805040 merges) | 07:11 |
| opendevreview | Merged x/sushy-oem-idrac master: Update default IncludeInExport value https://review.opendev.org/c/x/sushy-oem-idrac/+/800497 | 07:29 |
| arne_wiebalck | dtantsur: I could not find a section on Ironic's power sync explaining sync_power_state_workers for instance. I am happy to give it a shot, unless you are better at finding documentation :) | 07:36 |
| opendevreview | Merged x/sushy-oem-idrac stable/victoria: Revert "Add get PXE port MACs for BIOS mode" https://review.opendev.org/c/x/sushy-oem-idrac/+/799061 | 07:37 |
| opendevreview | Merged x/sushy-oem-idrac stable/victoria: Revert "Add export system configuration" https://review.opendev.org/c/x/sushy-oem-idrac/+/799062 | 07:38 |
| opendevreview | Merged x/sushy-oem-idrac stable/victoria: Update RETRY_COUNT for virtual media boot https://review.opendev.org/c/x/sushy-oem-idrac/+/797066 | 07:39 |
| opendevreview | Merged x/sushy-oem-idrac stable/victoria: Update virtual media boot-related constants https://review.opendev.org/c/x/sushy-oem-idrac/+/797065 | 07:39 |
| arne_wiebalck | dtantsur: All I find is how the sync with Nova works. | 07:42 |
| opendevreview | Merged x/sushy-oem-idrac master: Add RAID mode conversion methods https://review.opendev.org/c/x/sushy-oem-idrac/+/799506 | 07:44 |
| opendevreview | Merged x/sushy-oem-idrac master: Add RAID clear foreign config method https://review.opendev.org/c/x/sushy-oem-idrac/+/800071 | 07:44 |
| opendevreview | Aija Jauntēva proposed x/sushy-oem-idrac master: Add OEM Task extension https://review.opendev.org/c/x/sushy-oem-idrac/+/800269 | 07:53 |
| *** redrobot1 is now known as redrobot | 07:59 | |
| cenne | Good mornings ironic. | 08:04 |
| cenne | iurygregory: that's very good news! re CI | 08:05 |
| iurygregory | cenne, morning! yeah \o/ | 08:05 |
| cenne | : ) | 08:06 |
| iurygregory | Feel free to recheck your patches :D | 08:31 |
| rpittau | https://memegenerator.net/img/instances/85667456/recheck-all-the-things.jpg | 08:33 |
| dtantsur | morning ironic | 08:33 |
| rpittau | hey dtantsur :) | 08:34 |
| dtantsur | arne_wiebalck: if it's not in https://docs.openstack.org/ironic/latest/admin/tuning.html, it probably does not exist | 08:34 |
| iurygregory | rpittau,++ | 08:36 |
| iurygregory | morning dtantsur | 08:36 |
| arne_wiebalck | dtantsur: I will add it here: https://docs.openstack.org/ironic/latest/admin/power-sync.html (make the section more general) | 08:39 |
| dtantsur | arne_wiebalck: honestly, I have a feeling that "Tuning" is a better place for it, since it's about performance | 08:39 |
| dtantsur | but up to you | 08:40 |
| arne_wiebalck | dtantsur: the thing is that section describes some of the options already and I would like to avoid repetition ... maybe some cross-referecing will do? | 08:40 |
| arne_wiebalck | referencing | 08:40 |
| iurygregory | cross-referencing sounds good to me ^ | 08:41 |
| dtantsur | yep | 08:42 |
| dtantsur | folks, now that the CI behaves, could we merge https://review.opendev.org/c/openstack/ironic-lib/+/804194 and https://review.opendev.org/c/openstack/ironic-lib/+/804195/? these are holding the ironic-lib release. | 08:43 |
| iurygregory | dtantsur, ++ | 08:46 |
| iurygregory | lol I saw the old +2 in https://review.opendev.org/c/openstack/ironic-lib/+/804195/ I've +W but I've removed the +W already, +2 for now | 08:48 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [doc] Update power sync documentation https://review.opendev.org/c/openstack/ironic/+/805168 | 09:04 |
| fmuyassarov | Good morning Ironic. Is there a way I can instruct the IPA to start some systemd daemons when it is initially booted on the node | 09:05 |
| dtantsur | fmuyassarov: I think a custom DIB element is the way to go | 09:16 |
| dtantsur | you can create a systemd service with Before: ironic-python-agent | 09:16 |
| fmuyassarov | dtantsur, thanks. I didn't know that it's possible with DIB element. I thought whatever we write in DIB element will be executed only while building IPA image and not when it is booted later on. | 09:18 |
| dtantsur | fmuyassarov: your element can add systemd services or really anything to the final image | 09:23 |
| dtantsur | you can check how it's done in the main element | 09:23 |
| arne_wiebalck | fmuyassarov: FWIW, we do what dtantsur suggests to disable/enable services in our image and it is pretty easy to use | 09:26 |
| fmuyassarov | arne_wiebalck, dtantsur that's good to know. I will go with DIB element than. Thank you! | 09:45 |
| janders | dtantsur iurygregory rpittau I wanted to briefly go back to the discussion we had some time back about the default values of Inserted and WriteProtected fields in insert_media request payload | 10:05 |
| janders | reference: | 10:05 |
| janders | https://review.opendev.org/c/openstack/sushy/+/802452 | 10:06 |
| janders | and | 10:06 |
| * dtantsur is on a meeting, will respond slowly | 10:06 | |
| janders | https://review.opendev.org/c/openstack/sushy/+/802690 | 10:06 |
| janders | dtantsur: noted! | 10:06 |
| janders | the question is - what do we do with this - and what is the "right thing" to do | 10:06 |
| janders | changing defaults per se won't change much cause Ironic is calling insert_media with true,true | 10:07 |
| * janders is finding link to that part of the code | 10:07 | |
| * iurygregory looks | 10:08 | |
| janders | https://opendev.org/openstack/ironic/src/branch/master/ironic/drivers/modules/redfish/boot.py#L199 | 10:08 |
| janders | if we also changed Ironic behaviour, we will need to make sure the PATCH-based attachment is still fed with Inserted=True otherwise older Lenovos will break as we learned | 10:09 |
| janders | so - what would be the best way about this: | 10:09 |
| janders | 1) do nothing | 10:09 |
| janders | 2) change ironic and sushy to match Redfish schema defaults - and have hacks in place for PATCH-based attachments (e.g. hardcode Inserted=True if use_patch) | 10:10 |
| janders | 3) another approach? | 10:10 |
| janders | when you have time, have a look and let me know what you think | 10:10 |
| janders | now that we have the supermicro fix merged in master and backported to stable, I'm not too worried, but it seems kinda reasonable to align Ironic/sushy defaults with RF schema defaults if it doesn't wreak havoc | 10:12 |
| iurygregory | Inserted True is the default by redfish right? | 10:12 |
| janders | I believe so | 10:12 |
| iurygregory | and the only scenario so far we need this to be sent is with PATCH afaik | 10:12 |
| janders | correct | 10:12 |
| iurygregory | I think its fine to change their defaults to None, and we keep in mind that Inserted is required when doing Patch (at least for Lenovo) I don't think WriteProtected is required (we need to test this scenario) | 10:15 |
| iurygregory | at least I saw this in the lenovo support https://datacentersupport.lenovo.com/us/en/products/servers/thinksystem/sr650/7x06/solutions/ht508834-redfish-internal-error-when-mounting-remote-image-on-thinksystem-lenovo-thinksystem | 10:16 |
| janders | We could also set WriteProtected to True if use_patch - just to follow the principle of minimum change | 10:19 |
| janders | (I think Ironic and Sushy have been passing WriteProtected=True since vMedia support was added) | 10:20 |
| iurygregory | yeah | 10:20 |
| janders | also don't think it's needed but maybe it's safer not to change it if there is no need? | 10:20 |
| janders | (no need to make the change) | 10:20 |
| janders | OK - now how about the Ironic bit - do we change it not to send Inserted=True WriteProtected=True (and just send nothing)? | 10:21 |
| janders | (if we keep it as is that won't cause issues per se, but will be confusing, as sushy will drop these) | 10:22 |
| iurygregory | It would be better if we don't send Inserted/WriteProtected I think | 10:22 |
| janders | and then just hardcode Inserted|WriteProtected to True in use_patch conditional? | 10:23 |
| janders | (ugly, but I can't think of a better solution) | 10:23 |
| iurygregory | because we don't really send the info in SMC... | 10:23 |
| iurygregory | so it's a bit weird having hardcoded in Ironic while sushy will do some magic to remove before calling the BMC | 10:23 |
| janders | yeah agreed | 10:23 |
| janders | feels like one day someone will be reading the code to make some change or debug something and will go what da... | 10:24 |
| iurygregory | lets wait a bit to see what others think since they are in another meeting hehe | 10:24 |
| iurygregory | janders, yeah! | 10:24 |
| janders | ok! I think this is a good plan, thank you iurygregory | 10:24 |
| iurygregory | we will be like "wait we send True True, but sushy didn't send this information" | 10:24 |
| iurygregory | etc :D | 10:24 |
| janders | yeah | 10:25 |
| janders | how to track the flow of information through all the layers | 10:25 |
| janders | things magically disappearing doesnt help | 10:26 |
| jeeva | is there anywhere i can get information on the physical aspect of Ironic, in regards to physical routers & Switches configurations. We're currently running openstack deployed via kolla-ansible. | 10:27 |
| iurygregory | jeeva, I'm not familiar how kolla does the deployment, but we have some general documentation about Ironic and maybe the installation guide can give some information for you https://docs.openstack.org/ironic/latest/#installation-guide | 10:29 |
| jeeva | i've read through pretty much all the docs but nothing address physical switches/router setups | 10:32 |
| jeeva | unless im missing something | 10:32 |
| jeeva | got about 200 nodes i want to move from hypervisors to ironic | 10:33 |
| arne_wiebalck | jeeva: so, instead of virtual machines you want to hand out physical instances to users directly ? | 10:35 |
| jeeva | yes, since they have nvidia tesla V100 / A100's in them | 10:35 |
| jeeva | and will give a minor speed performance | 10:35 |
| arne_wiebalck | jeeva: GPUs work with virtual machines as well, no? | 10:36 |
| jeeva | yes they do | 10:36 |
| jeeva | but 1-3% speed increase going baremetal is worth it for us | 10:36 |
| * arne_wiebalck is not arguing against the idea, just likes to understand Ironic use cases | 10:37 | |
| arne_wiebalck | jeeva: yep, this is about the same we see | 10:37 |
| jeeva | we analyze meerkat/ square kilometer array data | 10:37 |
| jeeva | "Star data" | 10:37 |
| arne_wiebalck | jeeva: and why we are converting some of our hosts from hypervisors to bare metal instances | 10:37 |
| jeeva | however our team cant really find any concrete details about what we should do with our networking | 10:38 |
| jeeva | for ironic | 10:38 |
| jeeva | because our network component is managed by the university's Network engineers, so we need to put on paper what they need to do for us regarding that component | 10:39 |
| jeeva | everything else we self-manage | 10:39 |
| iurygregory | have you check the common considerations for network? https://docs.openstack.org/ironic/latest/install/refarch/common.html#networking | 10:41 |
| iurygregory | we also have some small information regarding small cloud with trusted tennats https://docs.openstack.org/ironic/latest/install/refarch/small-cloud-trusted-tenants.html#networking | 10:41 |
| jeeva | thanks, will have to see how this clashes with MAAS | 10:43 |
| arne_wiebalck | jeeva: if time permits, mgoddard may able to point you into the right direction when it comes to kolla and Ironic deployments | 10:51 |
| *** pmannidi|SA is now known as pmannidi|AFK | 10:55 | |
| *** pmannidi|AFK is now known as pmannidi | 10:56 | |
| *** pmannidi is now known as pmannidi|AFK | 11:27 | |
| opendevreview | Merged openstack/ironic-lib master: json_rpc: list_opts() should return tuple https://review.opendev.org/c/openstack/ironic-lib/+/804194 | 11:37 |
| opendevreview | Merged openstack/ironic-lib master: json_rpc: setup.cfg is missing entrypoint for json_rpc's list_opts https://review.opendev.org/c/openstack/ironic-lib/+/804195 | 11:37 |
| cenne | Hi, would this be considered about ready for merge before the freeze? https://review.opendev.org/c/openstack/ironic/+/800084 :) | 11:41 |
| iurygregory | cenne, we don't really have a freeze :D and this week we are only doing release for libs | 11:42 |
| iurygregory | sushy, ironic-lib | 11:42 |
| cenne | oh okay. cool. Thanks :) | 11:42 |
| iurygregory | I will try to take a look at your patch later today | 11:43 |
| dtantsur | we will try to get it in. as iurygregory said, we don't have formal freezes, other than some openstack-wide library freezes | 11:43 |
| dtantsur | cenne: but note that next week is the freeze for ironicclient, so you may want to get that in shape asap | 11:43 |
| cenne | dtantsur: it's ready https://review.opendev.org/c/openstack/python-ironicclient/+/803970 | 11:44 |
| dtantsur | ah, right, I've seen that and started thinking about 'secure boot on/off' | 11:44 |
| dtantsur | you see, we have two different precedences: `power on/off` and `maintenance set/unset` | 11:44 |
| dtantsur | I guess 'on/off' makes more sense, since it's a complex action, closer to power | 11:45 |
| dtantsur | any opinions? iurygregory? | 11:45 |
| dtantsur | but then we have `boot mode set`... | 11:46 |
| iurygregory | on/off makes sense "secure boot on" "secure boot off" | 11:48 |
| iurygregory | or enabled/disabled | 11:48 |
| dtantsur | okay, I think I won't break my head on this, will just go with the cenne's proposal | 11:49 |
| cenne | On module function level it should btw be able to take even True/False etc. There's conversion to bool happening in the end. | 11:53 |
| cenne | on cli on/off is slightly easier to type :P | 11:54 |
| dtantsur | yep | 12:02 |
| cenne | Re openstacksdk, the tests seem to fail on some 'cloud' tests though. | 12:03 |
| cenne | It's failing even on older patch that zuul had already blessed earlier, so I think it's def unrelated. | 12:03 |
| cenne | https://review.opendev.org/c/openstack/openstacksdk/+/802101 | 12:04 |
| dtantsur | yeah, I think there was a CI failure | 12:04 |
| dtantsur | see https://review.opendev.org/c/openstack/openstacksdk/+/804530/ and linked patches | 12:05 |
| cenne | ack | 12:05 |
| janders | see you tomorrow Ironic o/ | 12:25 |
| arne_wiebalck | bye janders o/ | 12:28 |
| cenne | bye janders o/ | 12:30 |
| mgoddard | jeeva: hi | 12:36 |
| TheJulia | good morning | 13:26 |
| opendevreview | Aija Jauntēva proposed x/sushy-oem-idrac stable/wallaby: Fix IncludeInExport allowed value validation https://review.opendev.org/c/x/sushy-oem-idrac/+/805145 | 13:29 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Add better error messages for invalid conf molds https://review.opendev.org/c/openstack/ironic/+/805222 | 13:32 |
| dtantsur | morning TheJulia | 13:46 |
| arne_wiebalck | Hey TheJulia o/ | 13:47 |
| iurygregory | good morning TheJulia welcome back =) | 13:49 |
| iurygregory | just a reminder I've sent the email with the midcycle schedule already - so if anyone think we should change something please let me know today | 13:52 |
| iurygregory | oh I should probably mention we will be using the meetpad https://meetpad.opendev.org/ironic | 13:52 |
| dtantsur | could I get a 2nd +2 on https://review.opendev.org/c/openstack/bifrost/+/802532 https://review.opendev.org/c/openstack/sushy-tools/+/803598 and https://review.opendev.org/c/openstack/sushy-tools/+/803596 please? | 14:52 |
| dtantsur | of these 3 only the 1st is non-trivial | 14:52 |
| opendevreview | Aija Jauntēva proposed x/sushy-oem-idrac master: Add OEM Task extension https://review.opendev.org/c/x/sushy-oem-idrac/+/800269 | 15:00 |
| iurygregory | dtantsur, done sorry for the long time the bifrost one was quite big :D | 15:37 |
| dtantsur | no worries, thanks much :) | 15:37 |
| iurygregory | ironic-cores one thing to keep in mind, the deadline for PTL Nomination is Aug 24 (ie Tuesday next week) | 15:40 |
| rpittau | bye, good night! o/ | 15:51 |
| *** rpittau is now known as rpittau|afk | 15:51 | |
| arne_wiebalck | bye everyone o/ | 16:04 |
| *** sshnaidm is now known as sshnaidm|afk | 16:09 | |
| opendevreview | Merged openstack/sushy-tools master: Use correct status code with werkzeug exceptions https://review.opendev.org/c/openstack/sushy-tools/+/803598 | 16:21 |
| opendevreview | Merged openstack/sushy-tools master: Support HTTP basic auth https://review.opendev.org/c/openstack/sushy-tools/+/803596 | 16:29 |
| opendevreview | Merged openstack/bifrost master: Move Nginx code to a new role bifrost-nginx-install https://review.opendev.org/c/openstack/bifrost/+/802532 | 16:35 |
| opendevreview | Dmitry Tantsur proposed openstack/sushy-tools master: Handle HTTP errors when downloading virtual media https://review.opendev.org/c/openstack/sushy-tools/+/805255 | 17:07 |
| opendevreview | Dmitry Tantsur proposed openstack/sushy-tools master: Support credentials for virtual media https://review.opendev.org/c/openstack/sushy-tools/+/802493 | 17:09 |
| opendevreview | Dmitry Tantsur proposed openstack/sushy-tools master: Support credentials for virtual media https://review.opendev.org/c/openstack/sushy-tools/+/802493 | 17:20 |
| opendevreview | Verification of a change to openstack/sushy-tools master failed: Add lower-constraints job to current development branch https://review.opendev.org/c/openstack/sushy-tools/+/804376 | 17:28 |
| dtantsur | o/ | 17:34 |
| TheJulia | stevebaker: you around? | 20:16 |
| stevebaker | TheJulia: hi! | 20:37 |
| TheJulia | I am starting to think grub is the root of all evil | 20:40 |
| stevebaker | TheJulia: I concur | 20:46 |
| stevebaker | TheJulia: oh I didn't realise this was already proposed for train, nice https://review.opendev.org/c/openstack/puppet-ironic/+/804430 | 20:51 |
| TheJulia | Well, last week :) | 20:52 |
| stevebaker | its taken me a week to get back to baseline after pfizer-1 | 20:54 |
| opendevreview | Merged openstack/ironic stable/wallaby: Retry stdlib ssl.SSLError https://review.opendev.org/c/openstack/ironic/+/803675 | 21:02 |
| TheJulia | stevebaker: yeouch :( | 21:48 |
| TheJulia | I'm not looking forward to moderna-3 | 21:48 |
| stevebaker | hey, its better than the alternative. I can't wait for pfizer-2 now that delta is spreading here | 21:51 |
| TheJulia | ugh :( | 21:51 |
| stevebaker | we had a good run | 21:55 |
| TheJulia | yeah | 21:59 |
| stevebaker | It will be interesting to see if an elimination strategy can work with delta. The whole country went into hard lockdown as soon as case 1 was found | 22:08 |
| janders | good morning Ironic o/ | 22:50 |
| *** pmannidi|AFK is now known as pmannidi | 23:21 | |
| *** pmannidi is now known as pmannidi|brb | 23:31 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
