| opendevreview | Jacob Anders proposed openstack/ironic-python-agent master: Improve efficiency of storage cleaning in mixed media envs https://review.opendev.org/c/openstack/ironic-python-agent/+/818712 | 01:04 |
|---|---|---|
| opendevreview | Jacob Anders proposed openstack/ironic-python-agent master: Improve efficiency of storage cleaning in mixed media envs https://review.opendev.org/c/openstack/ironic-python-agent/+/818712 | 04:32 |
| opendevreview | Jacob Anders proposed openstack/ironic-python-agent master: Improve efficiency of storage cleaning in mixed media envs https://review.opendev.org/c/openstack/ironic-python-agent/+/818712 | 05:25 |
| rpittau | good morning ironic! Happy Friday! o/ | 07:00 |
| janders | hey rpittau o/ | 07:01 |
| janders | Happy Friday | 07:01 |
| rpittau | hey janders :) | 07:01 |
| arne_wiebalck | Good morning rpittau janders and Ironic! | 07:39 |
| rpittau | Hey arne_wiebalck :) | 07:39 |
| ajya | Morning, Ironic | 08:20 |
| ajya | dtantsur: was the issue resolved from yesterday? | 08:20 |
| iurygregory | good morning Ironic, happy Friday! | 10:33 |
| iurygregory | fyi, I will be away for some hours now, because I need to do a medical check-up today | 10:34 |
| dtantsur | morning folks | 11:18 |
| dtantsur | we survived until Friday. congratulations, keep on! | 11:19 |
| dtantsur | ajya: no, it seems to be https://github.com/dell/iDRAC-Redfish-Scripting/issues/116 | 11:20 |
| ajya | dtantsur: yeah, I'm checking if this is regression. It works so far with 2.75.75.75, upgrading to newer and see if it fails | 11:21 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Allow legacy crypto policies in CentOS Stream 9 https://review.opendev.org/c/openstack/bifrost/+/831926 | 11:28 |
| dtantsur | rpittau: I think we're doing this ^^ in test-bifrost.sh | 11:31 |
| rpittau | yeah but update-crypto-policies is not present in the base system, so it's not working | 11:31 |
| dtantsur | ahhhh | 11:32 |
| rpittau | we can probably remove it from test-bifrost | 11:32 |
| dtantsur | rpittau: then we need to install it and probably remove the code from test-bifrost.sh? ah, and restore the policy after testing, ideally. | 11:32 |
| rpittau | yeah | 11:32 |
| rpittau | so | 11:32 |
| dtantsur | meanwhile, I'll experiment with an ecdsa key | 11:32 |
| rpittau | we install it during system installation, so it's fine | 11:33 |
| rpittau | we have a trap in test-bifrost.sh to restore the policy, not sure that's enough | 11:33 |
| dtantsur | it relies on knowing the previous value | 11:33 |
| rpittau | right | 11:34 |
| rpittau | I need to double-check for the other distros then | 11:34 |
| dtantsur | it's an RH thing | 11:34 |
| dtantsur | present also in Fedora and 8 (?) | 11:34 |
| rpittau | yep | 11:34 |
| rpittau | so we can set that for all CentOS higher than 8 and Fedora and remove it from test-bifrost.sh, then restore at the end | 11:35 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: [WIP] Prefer modern SSH key algorithms https://review.opendev.org/c/openstack/bifrost/+/831771 | 11:39 |
| rpittau | heh tricky distros, CS8 and Fedora have crypto-policies-scripts installed by default | 11:46 |
| rpittau | alright, we can install crypto-policies-scripts in test-bifrost or defer all to playbooks | 11:46 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Allow legacy crypto policies in CentOS Stream 9 https://review.opendev.org/c/openstack/bifrost/+/831926 | 11:46 |
| janders | my first stab at requesting a release: https://review.opendev.org/c/openstack/releases/+/831931 rpittau iurygregory_ please let me know if I got it mostly right. Thank you! :) | 12:28 |
| dtantsur | rpittau: my ecdsa patch seems to work, but it still fails with post_failure | 12:29 |
| dtantsur | I think the infra has a problem as well | 12:29 |
| dtantsur | https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_1cc/831771/2/check/bifrost-integration-tinyipa-centos-9/1cc23a2/job-output.txt | 12:29 |
| dtantsur | oh, that's our item. hmmm | 12:31 |
| dtantsur | symlink has no referent: \"/home/zuul/logs/all/README\" | 12:32 |
| dtantsur | rpittau: if you don't mind, I'd probably go with the ecdsa path, it seems more future-proof to me | 12:32 |
| ajya | dtantsur: after further testing I cannot reproduce the issue anymore. It does not look to be related to firmware version. I saw the error only once when I started. Any ideas what could be pre-condition for the failure? State of the system? Previously executed commands? | 12:36 |
| ajya | I'm trying both with the Redfish py script and Ironic deployment. | 12:36 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: Fix the CI on CentOS 9 https://review.opendev.org/c/openstack/bifrost/+/831771 | 12:39 |
| dtantsur | ajya: I'll ask | 12:39 |
| dtantsur | ajya: are you testing on FC430? | 12:40 |
| ajya | dtantsur: no, I don't have it, I'm testing on R630. I was able to get the error once, not sure how specific this is to FC430 | 12:43 |
| ajya | could it be similar workflow that janders was seeing? | 12:44 |
| dtantsur | ajya: do you know the difference between the FC and the R modules? | 12:46 |
| dtantsur | * models | 12:46 |
| rpittau | dtantsur: fine for me :) | 12:52 |
| ajya | dtantsur: I think they're different by how they are mounted (R - rack), we haven't tested FC models, but don't think they are somehow different from operational point of view and that they are 430 vs 630. They both are 13G, and support same iDRAC versions. | 12:54 |
| janders | ajya that is an interesting angle... may be worth checking if BootSoureOverride requests fail as well in that "corrupt" iDRAC state where we can no longer attach/detach vMedia | 12:54 |
| ajya | yeah, and is idrac reset helping to get it back to working? | 12:55 |
| janders | yes | 12:55 |
| ajya | I stopped seeing the issue when downgraded firmware that reboots idrac and cleans it state | 12:55 |
| ajya | I had the issue first I tried and I didn't check what's the state of the system was | 12:56 |
| janders | actually I take back that "yes" | 12:56 |
| janders | will let others answers this, not sure | 12:56 |
| ajya | sure, in summary, to me it looks there is a bug but need to understand how to reproduce it consistently, other than that there is no regression in newer idracs | 12:57 |
| ajya | the bug could be similar to the one that janders was seeing - something somewhere corrupts the state of idrac and only idrac reboot helps | 12:58 |
| janders | getting late here - gotta run | 12:59 |
| janders | see you on Monday Ironic - have a good weekend everyone o/ | 12:59 |
| janders | sorry to leave in the middle of a good discussion | 13:00 |
| janders | 2300hrs here, alarm clock set to 530 for a bike ride before it gets stinking hot | 13:00 |
| janders | was still 35C at 1730hrs today so need all the morning cool I can get | 13:01 |
| rpittau | janders: enjoy the ride and the weekend! o/ | 13:03 |
| ajya | janders: have a nice weekend | 13:05 |
| ajya | dtantsur: about poweredge naming convention - https://www.dell.com/support/kbdoc/en-us/000137343/how-to-identify-which-generation-your-dell-poweredge-server-belongs-to here it's form factor and CPU differences. | 13:07 |
| dtantsur | thanks! | 13:12 |
| dtantsur | ajya: the response: I tried a new blade and it worked the first time but then successive times it failed with the 503 - I think in that old git issue someone reported it worked the initial first time and then never worked again. | 13:14 |
| ajya | dtantsur: yup, saw that and for me it worked several times. Does iDRAC reboot get it working again? | 13:20 |
| dtantsur | ajya: I think no | 13:41 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: Fix the CI on CentOS 9 https://review.opendev.org/c/openstack/bifrost/+/831771 | 13:47 |
| *** iurygregory_ is now known as iurygregory | 14:28 | |
| iurygregory | I'm back o/ | 14:28 |
| dtantsur | welcome back iurygregory :) | 14:30 |
| * MahnoorAsghar o/ | 14:31 | |
| dtantsur | hi MahnoorAsghar | 14:31 |
| * MahnoorAsghar hello! | 14:31 | |
| iurygregory | ty, medical check-up is a pain... 2hrs waiting and 2min talking with the doctor lol | 14:32 |
| dtantsur | I can imagine :D | 14:32 |
| opendevreview | Merged openstack/ironic-python-agent bugfix/8.3: Stop defaulting the standalone job to netboot; use the correct boot mode https://review.opendev.org/c/openstack/ironic-python-agent/+/829722 | 14:33 |
| opendevreview | Ameya Raut proposed openstack/ironic-tempest-plugin master: Fix for verifying introspected data https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/831958 | 14:50 |
| opendevreview | Harald Jensås proposed openstack/ironic-lib master: Remove python version 2 pkgs from bindep https://review.opendev.org/c/openstack/ironic-lib/+/830319 | 14:57 |
| opendevreview | Ruby Loo proposed openstack/ironic stable/xena: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831851 | 15:05 |
| dtantsur | iurygregory: if you have a second: https://review.opendev.org/c/openstack/bifrost/+/831771 | 15:14 |
| opendevreview | Ruby Loo proposed openstack/ironic stable/xena: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831851 | 15:20 |
| opendevreview | Ruby Loo proposed openstack/ironic stable/xena: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831851 | 15:21 |
| opendevreview | Ruby Loo proposed openstack/ironic master: Fix rebuilds using anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831964 | 15:29 |
| rpittau | bye everyone. have a great weekend! o/ | 15:57 |
| * MahnoorAsghar Happy Friday ^-^ | 15:57 | |
| iurygregory | dtantsur, sure, I just got back from lunch :D | 15:58 |
| dtantsur | o/ | 16:40 |
| arne_wiebalck | anyone using redfishtool to communicate with redfish endpoints, or is everyone relying on curl? | 16:52 |
| iurygregory | arne_wiebalck, to test if something in your HW is working? I've used sushy itself or curl =) | 16:54 |
| arne_wiebalck | iurygregory: no, this is for users | 16:54 |
| iurygregory | oh | 16:54 |
| arne_wiebalck | iurygregory: to give them commands they could use to get the power status, for instance | 16:54 |
| iurygregory | we have the sushy-cli (but we dropped) | 16:54 |
| iurygregory | had* | 16:54 |
| arne_wiebalck | iurygregory: but this is already where I am stuck: I would think that getting the power status is the very first thing one tries, but I do not see this supported by redfishtool | 16:55 |
| iurygregory | weird =( | 16:55 |
| iurygregory | it would make sense | 16:55 |
| iurygregory | https://github.com/DMTF/Redfishtool this one right? | 16:56 |
| arne_wiebalck | yep | 16:56 |
| arne_wiebalck | mraineri: any hint how to get the power state with redfishtool? I did not see this in the documentation, but maybe I just missed it ... | 16:56 |
| mraineri | Let me check... | 16:59 |
| mraineri | Are you looking for a ComputerSystem power state? | 17:00 |
| mraineri | There's also power state reported at the Chassis level | 17:00 |
| arne_wiebalck | mraineri: ComputerSystem ... I guess :) | 17:03 |
| arne_wiebalck | mraineri: not sure what the difference is tbh | 17:03 |
| aman | I'm trying to get a list of cleaning step that are executed while running automatic cleaning on a node. But node's driver_internal_info just returns a blank [] for clean_steps, is there a way to get these steps? | 17:03 |
| mraineri | You should be able to get the full system info (including power state) with.... | 17:04 |
| arne_wiebalck | mraineri: I can get the whole Power item, but the current state does not seem to be part of it | 17:04 |
| mraineri | redfishtool -r <ip> -u <username> -p <password> Systems -I <id> | 17:04 |
| arne_wiebalck | mraineri: thanks, let me try ... | 17:04 |
| arne_wiebalck | mraineri: yep, that is a start, I was missing to pass the id | 17:05 |
| arne_wiebalck | mraineri: I was looking for the equivalent of 'ipmitool chassis power status' | 17:06 |
| mraineri | Be forwarned though, redfishtool was brought on early to help people transition from ipmitool; as things have grown, we've been trying to encourage folks to use Redfish Tacklebox | 17:06 |
| * arne_wiebalck googles | 17:07 | |
| mraineri | We actually haven't done something that does power info for systems.... The power/reset tool literally just invokes the Reset action for systems | 17:07 |
| mraineri | Can certainly add more system info stuff to get live state (beyond the FRU data) | 17:08 |
| arne_wiebalck | mraineri: yep, was just checking this :-D | 17:09 |
| arne_wiebalck | mraineri: so, I should rather go with Redfish Tacklebox than redfishtool ? | 17:09 |
| mraineri | Yeah, and I'd love to get feedback on it and requests for more functionality | 17:10 |
| arne_wiebalck | mraineri: you're years ahead :) ipmitool is still the main tool here, so our users will look for similar functionality | 17:11 |
| arne_wiebalck | mraineri: the three main things are: | 17:11 |
| arne_wiebalck | - get the power state | 17:11 |
| arne_wiebalck | - reset | 17:11 |
| arne_wiebalck | - reset the BMC | 17:11 |
| mraineri | Tacklebox has the system reset and BMC reset covered | 17:12 |
| arne_wiebalck | mraineri: cool | 17:12 |
| mraineri | redfishtool can do all three (but you need to comb the output of the system command to get the power state) | 17:12 |
| arne_wiebalck | mraineri: right | 17:13 |
| arne_wiebalck | mraineri: I can file a request if you point me to a good place and if that is helpful | 17:13 |
| mraineri | https://github.com/DMTF/Redfish-Tacklebox/issues | 17:13 |
| arne_wiebalck | mraineri: thanks a lot! | 17:14 |
| mraineri | Sure thing! | 17:14 |
| iurygregory | aman, you can check the conductor logs to identify the clean steps also | 17:16 |
| aman | iurygregory, yes that's what I am doing right now, i do see a mention of clean_steps.yaml in ironic.conf, will that file contain the steps? | 17:24 |
| arne_wiebalck | mraineri: sorry, if you have time for another quick one: a System's ID is basically a free identifier, right? (I see some implementations seem to use the serial number, but most simply use '1') | 17:31 |
| mraineri | That's right | 17:32 |
| arne_wiebalck | mraineri: thanks! | 17:35 |
| iurygregory | aman, humm let me check that part | 17:35 |
| opendevreview | Ruby Loo proposed openstack/ironic stable/xena: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831851 | 17:38 |
| arne_wiebalck | Resending [Bare Metal SIG] Short term change of plans for next week: our speaker is not available and we have to postpone his talk by a month. Anyone wants to step in spontaneously? | 17:40 |
| iurygregory | aman, this is under the ansible section on it, so it won't work. You can check if you have the steps you want to run with a priority set , our docs have information on how you can override priorities https://docs.openstack.org/ironic/latest/admin/cleaning.html | 17:40 |
| iurygregory | arne_wiebalck, ouch =( | 17:41 |
| aman | iurygregory, thank you. I will take a look at it! | 17:44 |
| opendevreview | Julia Kreger proposed openstack/sushy master: Fix session authentication issues https://review.opendev.org/c/openstack/sushy/+/821324 | 17:51 |
| opendevreview | Julia Kreger proposed openstack/ironic master: project scoped manager support https://review.opendev.org/c/openstack/ironic/+/818299 | 18:27 |
| arne_wiebalck | bye everyone, see you next week o/ (and let me know if you have a cool topic you would like to present next Tuesday in the Bare Metal SIG meeting :-)) | 18:33 |
| JayF | arne_wiebalck: Next month is the 8 year anniversary of Ironic being released as part of OpenStack. Perhaps a kinda look back/retrospective? But might be weird for the BM SIG | 18:56 |
| opendevreview | Ruby Loo proposed openstack/ironic stable/wallaby: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/831852 | 19:10 |
| arne_wiebalck | JayF: that is a very nice idea! And I don't think this is weird at all ... "the milestones of Ironic's evolution since the beginning" would be pretty neat, plus maybe some dead ends to spice it up :) maybe we could activate some contributors from back then ... or all PTLs ... | 20:27 |
| JayF | When is the BM SIG, usually, timewise? | 20:30 |
| JayF | Trying to see if I could help, but I think it's unlikely. Current job has no incentive for me to do this on their time, and I'd probably have to get some kinda approval even if it was on my time. | 20:31 |
| iurygregory | JayF, 2PM UTC =) | 20:34 |
| * TheJulia feels exhausted | 20:35 | |
| JayF | that's... 6am local time, I think? Yeah, probably not likely, but I'm happy to contribute knowledge or stories if needed :) | 20:38 |
| JayF | Gotta make sure "How Jay+friends heroically saved Ironic users from iSCSI" gets written in the history books :P /s | 20:39 |
| mraineri | arne_wiebalck: All merged now; keep the requests coming :) | 20:56 |
| arne_wiebalck | JayF: thanks for the offer to help! | 21:10 |
| arne_wiebalck | JayF: we could do an Ironic birthday special and change the time | 21:10 |
| arne_wiebalck | JayF: next week is probably too short notice | 21:10 |
| arne_wiebalck | JayF: but maybe for April or May? | 21:11 |
| JayF | Hmm. No promises, I'll see what sorta things I have to do on my end to get approval to do it. | 21:11 |
| arne_wiebalck | JayF: sure | 21:11 |
| arne_wiebalck | JayF: I guess we could also ask some more people who have been around back then | 21:11 |
| arne_wiebalck | mraineri: great, thanks! | 21:12 |
| arne_wiebalck | mraineri: getting SEL logs is one, but I see this is covered | 21:12 |
| JayF | rloo and dtantsur might be neck and neck for longest-tenured-active-contributor, maybe they'd have some insights :D | 21:12 |
| arne_wiebalck | mraineri: setting and getting the boot device is another, haven't checked that one yet | 21:12 |
| arne_wiebalck | JayF: I bet they have :) | 21:13 |
| JayF | I think a look back is good, and usually really feel good too -- because you can focus on what exists, and how far you've come vs the "pain point" sorta conversations that sometimes can dominate | 21:13 |
| rloo | I don't know anything, but NobodyCam has been around longer ;) | 21:13 |
| arne_wiebalck | oh, seems it is getting traction already :-D | 21:13 |
| * TheJulia is all for getting NobodyCam involved | 21:14 | |
| rloo | aeva i think was the instigator. you could get dan smith to say how he still regrets allowing it in nova as a beta or something... | 21:14 |
| * TheJulia thinks it is time to go take the strong migraine meds | 21:15 | |
| arne_wiebalck | we could do a panel like discussion ... I bet we would only have to ask one question and the discussion would flow by itself | 21:15 |
| rloo | (I remember when TheJulia mentioned that they'd like to be PTL one day...) | 21:15 |
| mraineri | arne_wiebalck: There is one-time boot override support; can certainly add something for the permanent boot order | 21:16 |
| mraineri | I was intentially avoiding that since it doesn't seem well-supported (at least at the time I last looked at it) | 21:16 |
| mraineri | But I can do a new inspection of the systems available to me | 21:17 |
| arne_wiebalck | mraineri: next boot is what I had in mind, not sure permanent is needed (for us, at least) | 21:19 |
| mraineri | Ah, okay; that should be covered with rf_boot_override.py | 21:20 |
| mraineri | So, a command like this will set the system to boot from PXE on the next boot, and then reset the system so it attempts that type of boot | 21:21 |
| mraineri | rf_boot_override.py -u root -p root -r https://192.168.1.100 -t Pxe -reset | 21:22 |
| mraineri | And this command will show you the current boot override configuration | 21:22 |
| mraineri | rf_boot_override.py -u root -p root -r https://192.168.1.100 | 21:22 |
| arne_wiebalck | awesome, will try this next week on our systems - thanks mraineri ! | 21:34 |
| arne_wiebalck | bye everyone, see you next week o/ | 21:36 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!