Tuesday, 2019-10-15

*** e0ne has joined #openstack-ironic00:05
*** e0ne has quit IRC00:18
*** TxGirlGeek has joined #openstack-ironic00:21
*** ociuhandu has joined #openstack-ironic00:31
*** e0ne has joined #openstack-ironic00:31
*** ociuhandu has quit IRC00:35
*** e0ne has quit IRC00:40
*** e0ne has joined #openstack-ironic00:43
*** Sayali_Kutwal has joined #openstack-ironic00:46
*** e0ne has quit IRC00:47
*** e0ne has joined #openstack-ironic00:54
*** TxGirlGeek has quit IRC00:56
*** e0ne has quit IRC00:58
*** hamzy has joined #openstack-ironic01:01
*** e0ne has joined #openstack-ironic01:05
*** bnemec has joined #openstack-ironic01:12
*** e0ne has quit IRC01:12
*** e0ne has joined #openstack-ironic01:18
*** e0ne has quit IRC01:22
*** bnemec has quit IRC01:25
*** e0ne has joined #openstack-ironic01:28
*** BjoernT_ has quit IRC01:35
*** e0ne has quit IRC01:36
*** BjoernT has joined #openstack-ironic01:37
*** e0ne has joined #openstack-ironic01:42
*** bnemec has joined #openstack-ironic01:42
*** e0ne has quit IRC01:46
*** e0ne has joined #openstack-ironic01:51
*** Garyx has quit IRC01:55
*** Garyx has joined #openstack-ironic01:56
*** e0ne has quit IRC02:00
*** Garyx has quit IRC02:05
*** e0ne has joined #openstack-ironic02:06
*** Garyx has joined #openstack-ironic02:07
*** e0ne has quit IRC02:14
*** Sayali_Kutwal has quit IRC02:15
*** e0ne has joined #openstack-ironic02:17
*** ianychoi has quit IRC02:19
*** ianychoi has joined #openstack-ironic02:21
*** e0ne has quit IRC02:22
*** e0ne has joined #openstack-ironic02:23
*** e0ne has quit IRC02:27
*** BjoernT has quit IRC02:28
*** e0ne has joined #openstack-ironic02:30
*** ricolin has joined #openstack-ironic02:40
*** e0ne has quit IRC02:41
*** e0ne has joined #openstack-ironic02:44
*** gregwork has joined #openstack-ironic02:47
*** e0ne has quit IRC02:49
*** e0ne has joined #openstack-ironic02:55
*** e0ne has quit IRC03:00
*** mkrai_ has joined #openstack-ironic03:05
*** mkrai_ has quit IRC03:22
*** mkrai__ has joined #openstack-ironic03:22
*** dsneddon has quit IRC03:57
*** BjoernT has joined #openstack-ironic03:58
*** BjoernT has quit IRC04:17
*** FlorianFa has quit IRC04:18
*** tzumainn has quit IRC04:25
*** dsneddon has joined #openstack-ironic04:26
*** FlorianFa has joined #openstack-ironic04:26
*** rh-jelabarre has quit IRC04:27
*** mkrai__ has quit IRC04:45
*** hwoarang has quit IRC04:57
*** Sayali_Kutwal has joined #openstack-ironic05:09
*** hwoarang has joined #openstack-ironic05:09
*** dsneddon has quit IRC05:15
*** dsneddon has joined #openstack-ironic05:19
*** dsneddon has quit IRC05:24
*** e0ne has joined #openstack-ironic05:30
*** dsneddon has joined #openstack-ironic05:54
openstackgerritSayali Kutwal proposed openstack/sushy master: Fix typo in the section Enabling SSL  https://review.opendev.org/68862205:54
*** mkrai_ has joined #openstack-ironic05:55
*** e0ne has quit IRC05:58
*** rachit7 has joined #openstack-ironic06:06
*** igordc has quit IRC06:10
*** Sayali_Kutwal has quit IRC06:11
*** bobmel has joined #openstack-ironic06:14
*** bobmel has quit IRC06:19
*** yolanda__ has quit IRC06:20
*** yolanda has joined #openstack-ironic06:20
arne_wiebalckGood morning, ironic!06:26
*** pcaruana has joined #openstack-ironic06:30
*** mkrai_ has quit IRC06:30
*** mkrai_ has joined #openstack-ironic06:33
*** vesper11- has quit IRC06:33
openstackgerritPranjali Srivastava proposed openstack/ironic master: Improve iDrac Documentation  https://review.opendev.org/66898706:35
*** vesper11 has joined #openstack-ironic06:36
*** Sayali_Kutwal has joined #openstack-ironic06:44
*** dsneddon has quit IRC06:45
openstackgerritJerry Wang proposed openstack/ironic master: Add support for replacing WSME  https://review.opendev.org/67585306:46
*** iurygregory has joined #openstack-ironic06:51
iurygregorygood morning Ironic o/06:51
*** trident has quit IRC06:52
arne_wiebalckhey iurygregory o/06:53
iurygregoryarne_wiebalck, o/06:53
*** trident has joined #openstack-ironic06:56
*** mkrai_ has quit IRC07:09
*** jawad_axd has joined #openstack-ironic07:12
*** tesseract has joined #openstack-ironic07:12
*** rcernin has quit IRC07:16
*** dsneddon has joined #openstack-ironic07:19
*** e0ne has joined #openstack-ironic07:25
*** mkrai_ has joined #openstack-ironic07:28
*** ociuhandu has joined #openstack-ironic07:30
*** ociuhandu has quit IRC07:35
*** ociuhandu has joined #openstack-ironic07:35
openstackgerritMerged openstack/sushy master: Fix typo in the section Enabling SSL  https://review.opendev.org/68862207:39
*** ociuhandu has quit IRC07:42
*** priteau has joined #openstack-ironic07:53
*** dtantsur|afk is now known as dtantsur07:53
dtantsurmorning ironic07:53
*** bobmel has joined #openstack-ironic07:54
*** e0ne has quit IRC07:55
dtantsurthanks TheJulia!08:00
iurygregorymorning dtantsur08:00
*** Sayali_Kutwal has quit IRC08:00
*** dougsz has joined #openstack-ironic08:01
dtantsurrpioso, TheJulia, train releases must be cut from stable/train at this point. what you're trying to do doesn't seem acceptable from the release management perspective.08:02
*** rachit7 has quit IRC08:04
*** bobmel has quit IRC08:05
*** bobmel has joined #openstack-ironic08:05
*** rpittau|afk is now known as rpittau08:07
rpittaugood morning ironic! o/08:07
iurygregorymorning rpittau o/08:08
rpittauhey iurygregory :)08:08
* rpittau hope to see a DIB release soon :/08:10
dtantsurrpittau: it has been released, but hasn't hit upper-constraints yet08:11
rpittauoh ok!08:11
*** dsneddon has quit IRC08:12
*** dsneddon has joined #openstack-ironic08:13
*** ociuhandu has joined #openstack-ironic08:13
*** dsneddon has quit IRC08:17
*** e0ne has joined #openstack-ironic08:17
*** ociuhandu has quit IRC08:20
*** alexmcleod has joined #openstack-ironic08:22
*** lucasagomes has joined #openstack-ironic08:24
*** ociuhandu has joined #openstack-ironic08:25
*** e0ne_ has joined #openstack-ironic08:25
*** e0ne has quit IRC08:26
iurygregorydtantsur, I was about to ask here about the No bootable device, I was looking at the logs XD08:30
*** ociuhandu has quit IRC08:31
*** ociuhandu has joined #openstack-ironic08:31
*** ociuhandu has quit IRC08:35
*** ociuhandu has joined #openstack-ironic08:35
dtantsurmm, yeah, unclear08:36
dtantsurI can only suggest trying to reproduce it locally with devstack08:37
dtantsurand checking what the node actually look like after deploy08:37
iurygregoryack o/08:37
openstackgerritMerged openstack/bifrost stable/pike: Update links for opendev  https://review.opendev.org/68855408:38
openstackgerritMerged openstack/bifrost stable/ocata: Update links for opendev  https://review.opendev.org/68855608:38
dtantsurit seems that the CI writes the image to the 2nd disk. I wonder if it may cause any problems.08:41
*** ociuhandu has quit IRC08:42
*** tssurya has joined #openstack-ironic08:45
*** dsneddon has joined #openstack-ironic08:49
*** bnemec has quit IRC08:53
*** e0ne_ has quit IRC09:09
*** dsneddon has quit IRC09:29
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Add a non-voting job with CentOS 8  https://review.opendev.org/68866109:33
*** derekh has joined #openstack-ironic09:43
*** ociuhandu has joined #openstack-ironic09:47
*** ociuhandu has quit IRC09:53
iurygregorydo we need to clone ironic-python-agent-builder before run stack.sh? .-.09:53
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866109:57
dtantsurrpittau: this ^^ may even work around the problem with DIB we have now09:57
dtantsuriurygregory: shouldn't be necessary09:57
iurygregorydtantsur, seems strange http://paste.openstack.org/show/783890/09:57
dtantsuriurygregory: I wonder if it's an ordering problem09:58
*** dsneddon has joined #openstack-ironic09:59
dtantsuriurygregory: right, we may need to move cloning IPA-builder away from build_tinyipa_image09:59
dtantsuryou're using a cached image, so it does not happen09:59
iurygregorygot it10:00
dtantsuriurygregory: if you have a time to propose a fix, would be awesome10:03
rpittaudtantsur: yep, great idea10:03
iurygregorydtantsur, ack10:04
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866110:07
*** e0ne has joined #openstack-ironic10:20
openstackgerritRiccardo Pittau proposed openstack/ironic-inspector master: [WIP] use openstacksdk for ironic module  https://review.opendev.org/67217910:25
*** ociuhandu has joined #openstack-ironic10:31
*** ociuhandu has quit IRC10:36
*** jovial[m] has joined #openstack-ironic10:42
jovial[m]Anyone got any idea why my virtual machine can't execute the init process in the centos based IPA images I got from: https://tarballs.openstack.org/ironic-python-agent/dib/files/. I get `can't execute init` in the kernel log messages, see: https://zuul.opendev.org/t/openstack/build/8a551c7c807c418b9b006f682022fa90/log/primary/system_logs/tenks/tk1-console.txt.gz#296 . `/init` seems to exist and is a symlink to10:43
jovial[m]`sbin/init` which is a symlink to systemd. Could this be some missing shared libraries or something?10:43
*** dsneddon has quit IRC10:46
*** dsneddon has joined #openstack-ironic10:56
*** dsneddon has quit IRC11:00
*** dsneddon has joined #openstack-ironic11:03
*** dsneddon has quit IRC11:08
openstackgerritzhangboye proposed openstack/ironic-inspector master: update source link in readme  https://review.opendev.org/68867811:08
dtantsurjovial[m]: can it be that you have too little RAM? real IPA images are quite heavy11:10
jovial[m]dtantsur: intersting suggestion, I will increase the ram and report back.11:11
dtantsurjovial[m]: how much do you have now?11:17
jovial[m]well, I was trying with 1GB, I haven't yet had a chance to increase.11:18
*** dsneddon has joined #openstack-ironic11:18
dtantsurjovial[m]: likely too little. I'd start with 3 GiB, make it work, and then try to bring it down. I don't think you can fit it in less than 1.5GiB, probably even 2.11:18
iurygregorydtantsur, should I add a release note?11:18
dtantsuriurygregory: for devstack changes? no, they're not production-facing.11:18
iurygregorydtantsur, ack =)11:19
*** dsneddon has quit IRC11:22
dtantsurjovial[m]: at a data point, an unpacked ramdisk currently takes 1.2GiB. This is your absolute minimum. I'm working on reducing it, but I don't expect any miracles. Various firmware blobs alone takes quite much.11:23
openstackgerritIury Gregory Melo Ferreira proposed openstack/ironic master: Setup ipa-builder before building ramdisk  https://review.opendev.org/68868111:27
*** dsneddon has joined #openstack-ironic11:32
*** dsneddon has quit IRC11:37
openstackgerritKaifeng Wang proposed openstack/ironic-inspector master: Add libffi-devl to bindep  https://review.opendev.org/68868611:48
kaifengdtantsur: interesting that normally there would be some VFS panic when ram is insufficient, but the log above has no hint of that.11:50
openstackgerritKaifeng Wang proposed openstack/ironic-inspector master: Add libffi-dev to bindep  https://review.opendev.org/68868611:52
openstackgerritIlya Etingof proposed openstack/ironic-python-agent-builder master: Add static network configuration support  https://review.opendev.org/68840211:53
*** kaifeng is now known as kaifeng|afk11:55
openstackgerritRachit Kapadia proposed openstack/ironic master: DRAC: Drives conversion from JBOD to RAID  https://review.opendev.org/67909311:56
*** goldyfruit has quit IRC11:58
*** mkrai_ has quit IRC12:00
*** dsneddon has joined #openstack-ironic12:00
iurygregorynode failed to deploy locally D:12:00
openstackgerritIlya Etingof proposed openstack/ironic master: Add `get_node_network_data` to non-Neutron NetworkInterface  https://review.opendev.org/68826612:08
openstackgerritIlya Etingof proposed openstack/ironic master: Make redfish boot interface using network interface metadata  https://review.opendev.org/68827212:08
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866112:10
iurygregorydtantsur, locally it fails with "Asynchronous exception: Node failed to deploy. http://paste.openstack.org/show/783903/ end o the tempest.log12:11
rpiosoGood morning, ironicers12:12
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866112:13
jovial[m]dtantsur: thanks you were right about the ram, It boots OK with 4GiB, going to see how low I can go.12:13
iurygregorymorning rpioso12:14
rpiosodtantsur: Would modifying the branch of the change for the new sushy release from master to stable/train address your comment? Any issue with the commits I suggested basing it on?12:15
rpiosoiurygregory: Hey! :-)12:16
*** rh-jelabarre has joined #openstack-ironic12:18
*** belmoreira has joined #openstack-ironic12:24
dtantsurrpioso: I'm not sure what exactly you suggest.. any train release must be cut from a commit on stable/train (or before it was branched). if it's the case - all is fine.12:25
dtantsurI got an impression that you're trying to cut a train release from master *after* stable/train was branched. this is not allowed since it messes up the history for future train releases (and train is a stable release, so not features allowed).12:26
dtantsurif what you're trying is to cut a release from somewhere before cutting stable/train, it may or may not work. you need to get the release team assistance in this case.12:27
dtantsurrpioso: I hope this ^^ addresses your case12:27
*** yaawang has quit IRC12:29
rpiosodtantsur: All three commits identified in https://review.opendev.org/#/c/688551/, the one presently proposed and the two I suggested, exist on stable/train. master and stable/train are presently almost identical.12:29
patchbotpatch 688551 - releases - Release sushy 1.10.0 from during Train - 1 patch set12:29
*** yaawang_ has joined #openstack-ironic12:29
rpiosodtantsur: Thank you! Is that assistance available in #openstack-release?12:32
dtantsurrpioso: got it. you still need to chat with the release team on whether it's acceptable at all or not. yes, #openstack-release.12:32
dtantsurrpioso: before you go, what exactly are you trying to solve with this release? downstreams are expected to consume the latest releases from stable branches anyway..12:32
*** hamzy has quit IRC12:34
rpiosodtantsur: driver_requirements.txt identifies a sushy release which we have known does not work with Dell EMC gear. Couldn’t a packager pick that broken release?12:35
dtantsurrpioso: potentially, although they're supposed to watch out for such problems.12:35
dtantsurin any case, we don't update driver-requirements for bug fixes as a rule12:36
*** rloo has joined #openstack-ironic12:36
*** hamzy has joined #openstack-ironic12:37
dtantsuralthough in this case... I don't know. I would get the release team guidance on this as well.12:38
*** rachit7 has joined #openstack-ironic12:39
rpiosodtantsur: The original bug was fixed about three months ago. The existing bug is that driver requirements wasn’t updated to account for that fix.12:39
rpiosoA good deal of time passed between the original fix and next release containing it.12:40
dtantsurserves us as a reminder to release often12:41
rpiosoHeh12:41
rpioso+112:41
jrollI still can't figure out which versions of sushy are bugged for train, and why releasing 1.10.0 fixes it12:42
jrollI think I've read every reference12:42
jrollthere's nothing declaring "version x of sushy does not work with ironic train"12:42
jroll(I assume train is the problem, anyway)12:42
dtantsurcurrently train is 1.9.0 and 2.0.0. the bug fix happened between them.12:42
dtantsurbumping a major version on a stable branch is a clear problem, so I guess rpioso is trying to find a way out12:43
dtantsurI'd leave the downstreams to figure it our, but we're in a known disagreement on this topic :)12:43
*** S4ren has joined #openstack-ironic12:43
jrollcurrently ironic in train depends on >=1.9.0, though, so what does releasing 1.10.0 do? or do we plan to pin driver-requirements to <2.0.0?12:43
dtantsurjroll: it's the other way around: 2.0.0 works, 1.9.0 breaks Dell hardware12:43
dtantsur1.10.0, if released, will also work and won't be major12:44
jrollif 2.0.0 works, why not just continue using that in train?12:44
dtantsurI'm fine with ^^^12:44
jrollthat is in upper-constraints for train: https://opendev.org/openstack/requirements/src/branch/stable/train/upper-constraints.txt#L43112:44
jrolland 2.0.0 was released in train: https://opendev.org/openstack/releases/src/branch/master/deliverables/train/sushy.yaml#L1312:45
jrollif dell hardware works in ironic train with sushy 2.0.0, why are we doing anything?12:45
jrollI'm even more confused now12:46
jrollrpioso: ^^12:46
rpiosojroll: Because the sushy entry in driver-requirements suggests 1.9.0 is good. It’s not for Dell EMC and perhaps other vendors.12:47
jrollrpioso: what does releasing 1.10.0 accomplish? have we agreed that we can bump driver-requirements in stable/train to >=1.10.0? I don't think that's in line with stable policy, either12:48
rpiosoAs far as I understand, packagers could select that broken 1.9.0 version.12:49
rpiosoI thought we had agreed this past Thursday.12:50
jrollI was genuinely asking, fwiw12:51
jrollI was barely here thursday :)12:51
* jroll goes to logs12:51
dtantsurI don't remember agreeing either (it doesn't mean that I necessary disagree)12:52
jrolloh, I was in that conversation, heh12:52
rpiosoIt may have been Friday. Weekends! :-)12:52
jrollhttp://eavesdrop.openstack.org/irclogs/%23openstack-ironic/%23openstack-ironic.2019-10-10.log.html#t2019-10-10T20:38:5112:52
dtantsursince we're obeying the official policy, it will need a sign-off from the stable team12:52
dtantsur(not only our stable team, but also the global one)12:53
jrollI said I wonder if we can cut 1.9.1 from the history, and that I wonder if dtantsur knows the rules on bumping driver-reqs in stable :)12:53
dtantsurgenerally, it's discouraged. we've done it in the past though to rule out clearly broken versions.12:53
jrollTheJulia agreed that 1.9.1 would be best12:53
dtantsurthe definition of "clearly broken" is not entirely defined. somewhere between "ironic doesn't even start" and "causes mild inconveniences"12:53
jrollright12:54
dtantsur1.9.1 would be the best, if we did not have any feature patches between 1.9.0 and the fix12:54
jrollI suspect we did, if TheJulia decided to make it 1.10.012:54
dtantsurraising the requirement would still need a stable team approval, but I assume it would be easier to get12:54
rpiosojroll: Right! 1.9.1 it was.12:55
jrollso to answer my original question, we haven't agreed that we can bump driver-requirements in stable/train to >=1.10.012:56
jrollthis sounds like something to discuss on the ML12:56
rpiosodtantsur: In this case, the redfish hardware type cannot be used to manage Dell EMC and perhaps other vendors’ hardware.12:57
* rpioso naively thinks that fits the definition of broken12:58
*** mbeierl has joined #openstack-ironic12:58
dtantsurI'm inclined to agree. But I won't be the final authority in this particular question.12:59
rpiosoTheJulia: Interesting backscroll ^^^13:00
rpiosocdearborn: ^^^13:01
jrollso, can we bring this to the ML with all of the details, to bring in the stable folks?13:01
*** dsneddon has quit IRC13:01
*** cdearborn has joined #openstack-ironic13:02
jrollrpioso: ^13:02
rpiosocdearborn: Good morning!13:03
cdearbornrpioso: Morning13:03
rpiosocdearborn: There’s backscroll above regarding new sushy release for Train.13:04
*** mbeierl has quit IRC13:05
jrollI take it we aren't going to take this to the mailing list, then. ok.13:05
dtantsurI'd like it to go to the ML13:05
dtantsurwe haven't consistently done it, but we really should13:06
rpiosojroll: Seems reasonable assuming TheJulia hasn’t already consulted with them.13:06
jrollI don't think I can follow this topic in IRC, so I'm just gonna wait for that before I comment any further13:06
dtantsuresp. since we'll need an explicit approval, not a lazy consensus13:06
jrollrpioso: seems reasonable to me anyway, to ensure everyone is aware and it's well-documented. remember not everybody reads IRC, especially this channel. also remember the ML has operators on it.13:07
jrolloperators and packagers, I should say13:07
rpiosojroll: Gotcha. Who should send it?13:07
jrollrpioso: you seem to have the most context, to be able to add all the details13:07
rpiosoBear in mind other vendors could be impacted, too.13:08
jrollsure, even more reason to involve the whole community13:08
openstackgerritRiccardo Pittau proposed openstack/ironic-inspector master: [WIP] use openstacksdk for ironic module  https://review.opendev.org/67217913:09
rpiosojroll: Which tags should be used on the subject line?13:09
*** mbeierl has joined #openstack-ironic13:09
jroll[ironic][stable]?13:10
* TheJulia makes coffee13:10
iurygregory maybe [requirements] also not sure =)13:10
jrollyeah, I've no clue13:12
rpiosoiurygregory, jroll: Thank you! I’ll look around to refine.13:13
rpittauisn't this something for governance also ?13:13
jrollrpittau: I don't think so, why would it be?13:14
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866113:17
rpittaujroll: my bad, I actually meant releases as tag, but I was reading stuff on governance and I wrote that :/13:19
TheJuliaWhat I find most disturbing by only shipping oldest is the security implications...13:19
jrollrpittau: ah yeah, [release] might be good. /cc rpioso13:19
jrollTheJulia: "only shipping oldest"?13:19
jrolloh, if we pin <213:19
jrollyeah13:20
TheJuliaThe argument that we should do this because packagers may ship 1.9.013:20
TheJuliabut reality is packagers are also fighting time constraints and don't have solid visibility13:20
jrollah. yes, that would be bad if they never updated13:20
jrollwait, I'm not supposed to talk about this here, waiting for the email version :)13:21
*** goldyfruit has joined #openstack-ironic13:22
*** goldyfruit has quit IRC13:24
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Temporary raise diskimage-builder to the latest version  https://review.opendev.org/68870813:24
dtantsurrpittau: ^^ ugly, but...13:24
rpittaudtantsur: well as a temp fix is ok13:25
dtantsurif it gets approved, I'll propose a revert, and we can recheck it until it passes :)13:25
rpiosojroll, rpittau: Ack re: [release].13:27
rpiosodtantsur, jroll, rpittau, TheJulia: I’ll send an email to the ML today, if that’s agreeable.13:29
dtantsur++13:29
rpittaurpioso: sure :)13:29
jrollthanks ++13:29
*** jtwill98 has joined #openstack-ironic13:30
*** goldyfruit has joined #openstack-ironic13:32
*** pcaruana has quit IRC13:32
TheJuliarpioso: That would be perfet, thanks13:32
*** pranjali has joined #openstack-ironic13:32
TheJuliadtantsur: ack13:32
*** tzumainn has joined #openstack-ironic13:33
TheJuliadtantsur: re findfs, dunno. I kind of like the way the existing code was written looping through and evaluation. because we have more to troubleshoot. I was really thinking findfs is just a "well, the world is moderately broken fallback.13:38
dtantsurfindfs is also part of linux-utils, it should probably use the same code as lsblk13:39
dtantsuror rather: it also uses libblkid13:39
*** bobmel has quit IRC13:39
openstackgerritIury Gregory Melo Ferreira proposed openstack/ironic master: Setup ipa-builder before building ramdisk  https://review.opendev.org/68868113:39
iurygregorydtantsur, thanks for the review o/13:39
dtantsurI'm mildly worried about too many code paths. But only mildly.13:39
TheJuliait does, I'd really like to find out why lsblk is basically crippled, but *shrug*13:41
*** bnemec has joined #openstack-ironic13:43
*** Goneri has joined #openstack-ironic13:44
*** bobmel has joined #openstack-ironic13:44
rpittauTheJulia, dtantsur, I dunno if we tried it yet but it could be worth to set the debug parameters for lsblk and libblkid to get more info, if we want to investigate more13:45
TheJuliadtantsur: btw, I think I technically spotted a bug with what the uuid code presently does, but the idea behind the fallback does kind of fix that for the "raw filesystem on a whole disk" edge case13:45
*** pranjali has quit IRC13:49
iurygregorybrb, commuting to home13:51
dtantsurhmm. okay, we can start with a fallback to make it more backportable13:52
dtantsurthen consider a full switch13:52
*** jawad_axd has quit IRC13:52
TheJuliaI'll put a note in and for later consideration.13:52
TheJuliathat is once I get back to it13:52
*** jawad_axd has joined #openstack-ironic13:53
*** iurygregory has quit IRC13:56
*** jawad_ax_ has joined #openstack-ironic13:57
*** ociuhandu has joined #openstack-ironic13:58
*** jawad_axd has quit IRC13:58
*** jawad_ax_ has quit IRC14:01
*** pcaruana has joined #openstack-ironic14:04
arne_wiebalckTheJulia: For manual inspection, what's the recommended/obvious way to try it? Is there a way to install the IPA on an active instance?14:10
openstackgerritDmitry Tantsur proposed openstack/ironic master: [WIP] Wire in in-band cleaning for PXE boot  https://review.opendev.org/68872014:12
TheJuliaarne_wiebalck: you can install it, but you only want to run the special command noted in the release notes14:15
openstackgerritIlya Etingof proposed openstack/ironic-python-agent-builder master: Add static network configuration support  https://review.opendev.org/68840214:15
arne_wiebalckTheJulia: I understood the special command is to trigger ironic to talk to the IPA inside the instance, no?14:16
TheJuliawell, it reaches outbound from the instance14:16
TheJuliabut that would really only be for active instances14:17
TheJuliaif not active, the inspect state machine is best14:17
arne_wiebalckI thought the idea was to inspect active instances?14:17
TheJuliaYes, but it only publishes out14:17
TheJuliafor that case, and you need to tell ironic-inspector to process the data.14:18
TheJuliait is a config file flag14:18
* arne_wiebalck needs a t-shirt, erm diagram14:18
TheJuliaarne_wiebalck: you can modify the one on the link :)14:18
arne_wiebalcklol14:18
arne_wiebalckwell, I would need to understand first :-S14:19
*** iurygregory has joined #openstack-ironic14:19
TheJuliahttps://www.customink.com/designs/pixieboots/tny0-00c0-tckg/share?pc=EMAIL-40778&utm_campaign=shared%20design&utm_source=share%20link&utm_medium=shared%20design&utm_content=shared%20desktop14:19
arne_wiebalckI thought the manual inspection would trigger ironic/inspector to get in touch with an IPA inside an active instance14:19
openstackgerritChristopher Dearborn proposed openstack/ironic master: Update sushy library version  https://review.opendev.org/68798314:19
TheJuliachange the type of shirt on the screen on the left, because that is a women's vneck14:20
arne_wiebalckwe do individual (i.e. no group) orders, right?14:23
TheJuliathat is currently my plan. I really have no desrie to try and complicate entry with the current US-China trade situation14:23
TheJuliare manual introspection triggered from with-in the instance. IPA's inspection action is actually a default action in the code right after starting the agent (which is why there is a special command now for with-in the agent, so the IPA API is not started up because that could be bad^(TM)14:24
arne_wiebalckorder minimum is 614:24
TheJuliaarne_wiebalck: you can do individuals if you find an individual shirt you like14:25
arne_wiebalckTheJulia: ok, will check ...14:25
arne_wiebalckTheJulia: manual inspection: so, no running IPA needed, just the repo and the command launched?14:26
TheJuliaarne_wiebalck: yup14:26
arne_wiebalckTheJulia: how does the comamnd now where to send things?14:26
arne_wiebalckcommand14:26
arne_wiebalckknow14:26
arne_wiebalck(sorry)14:27
TheJuliaarne_wiebalck: it will first try to use mdns, if that fails for discovery, you have to feed it an inspector api url14:27
openstackgerritIlya Etingof proposed openstack/ironic-python-agent-builder master: Add static network configuration support  https://review.opendev.org/68840214:28
arne_wiebalckTheJulia: I think that is in the docs ... sorry.14:28
arne_wiebalckTheJulia: Ok, I'll try, thanks!14:28
*** bobmel has quit IRC14:31
arne_wiebalckTheJulia: re "the magic shirt": no shipping outside the US14:32
TheJuliagah :(14:32
iurygregory$33 =X14:32
TheJuliaarne_wiebalck:  let me know what size you want and I'll start building an order I gues14:32
arne_wiebalckTheJulia: I'll use your file and go with another shop.14:32
TheJuliaiurygregory: same^14:32
TheJuliaarne_wiebalck: okay14:32
iurygregoryI was able to pu the czech republic zipcode arne_wiebalck .-.14:32
TheJuliaoh!14:32
TheJuliaone moment14:33
iurygregoryput*14:33
TheJuliahttps://usercontent.irccloud-cdn.com/file/HiJcFE0T/Ironic_bear_archv2_white.png14:33
arne_wiebalckTheJulia: great, thanks14:33
TheJuliawhite on transparent for a dark shirt14:33
*** ociuhandu has quit IRC14:33
arne_wiebalckTheJulia: nice14:33
iurygregoryTheJulia, what was the ETA to delivery for you?14:34
* iurygregory only saw now: Unrecognized ZIP Code Change Applicable tax may apply during checkout. WUT14:35
TheJuliaiurygregory: I guess it really is US only14:40
*** dtantsur is now known as dtantsur|brb14:40
dtantsur|brbsad14:40
iurygregoryyeah14:40
TheJuliaiurygregory: it is trying to apply tax to the purcase in the site :(14:40
* dtantsur|brb -> short walk14:40
iurygregoryarne_wiebalck, if you find one that ships to europe let me know XD14:41
arne_wiebalckiurygregory: I'll order from Europe :)14:41
*** dsneddon has joined #openstack-ironic14:41
*** rachit7 has quit IRC14:42
rpittauiurygregory, arne_wiebalck, I'm trying with https://www.shirtinator.eu but the quality of the image doesn't look the best14:46
iurygregoryrpittau, =(14:46
iurygregory#sad14:46
*** dsneddon has quit IRC14:46
TheJuliait could also be that their fast scaling alogorithm for the page doesn't try to scale the image very well. Maybe try uploading a locally scaled up version to a higher resolution?14:48
*** mkrai_ has joined #openstack-ironic14:54
iurygregoryAnyone saw this http://paste.openstack.org/show/783903/ ?14:54
* iurygregory trying to reproduce the failure from the CI locally 14:55
TheJuliano, but https://github.com/ansible/ansible/issues/19720 seems to indicate parser error, so likely jinja2 processing issue14:57
TheJuliaor incompatability14:57
iurygregorythis only happens when we are using python314:58
iurygregory=(14:58
iurygregorypython2.7 CI is fine and tests locally are fine also14:59
*** TxGirlGeek has joined #openstack-ironic14:59
iurygregoryin the CI with python3 ansible fails to build the image https://openstack.fortnebula.com:13808/v1/AUTH_e8fd161dc34c421a979a9e6421f823e9/zuul_opendev_logs_d4d/687521/4/check/ironic-standalone/d4dbd44/controller/logs/ironic-bm-logs/node-4_no_ansi_2019-10-14-10%3A48%3A36_log.txt.gz14:59
TheJuliahmmm15:00
iurygregorymaybe something in the ansible module is not python3 friendly .-.15:02
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Don't try to put a bootloader in place when bootable  https://review.opendev.org/67933315:02
*** ociuhandu has joined #openstack-ironic15:07
*** ociuhandu has quit IRC15:10
TheJuliasure sounds that way15:11
*** ociuhandu has joined #openstack-ironic15:11
TheJuliaxXraphXx: You around?15:11
xXraphXxTheJulia, hi, checking log history15:12
*** gyee has joined #openstack-ironic15:13
arne_wiebalckTheJulia: 679333 affects partition images only, no?15:13
*** bobmel has joined #openstack-ironic15:14
*** bobmel has quit IRC15:15
*** bobmel has joined #openstack-ironic15:17
TheJuliaarne_wiebalck: really whole disk images, but one can make a bootable partition only image... in theory15:18
arne_wiebalckTheJulia: I thought that whole disk images don't install the bootloader anyway ..?15:19
TheJuliaThat got changed along the line15:19
* arne_wiebalck remembers some if statement ...15:20
TheJuliaand is semi-necessary for UEFI15:20
TheJuliasome vendors's firmware is broken in otherwords15:20
arne_wiebalckTheJulia: this is the "if" I was thinking of: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/agent_base_vendor.py#L83415:23
* TheJulia tries to remember the case we still end up there sometimes15:25
*** jawad_axd has joined #openstack-ironic15:28
xXraphXxiurygregory, hi, regarding the ansible module, using it along with python3. I'm not sure to see logs correctly about your issue. Here is some diff I found, that may be totally unrelated, just made a dumb diff between my version and the upstream one regarding ansible:15:33
*** jawad_axd has quit IRC15:33
xXraphXxcommented callback line in drivers/modules/ansible/playbooks/ansible.cfg:15:33
xXraphXxcallback_whitelist = ironic_log15:33
iurygregoryxXraphXx, the failure in CI is https://zuul.opendev.org/t/openstack/build/d4dbd443550346f9ba02edd4aa998f1815:33
iurygregorythe one I pasted was what I got while trying to reproduce the CI failure locally .-.15:34
TheJuliaoh, so no callback mod... that seems like maybe where this is going sideways15:34
*** igordc has joined #openstack-ironic15:37
*** bobmel has quit IRC15:39
*** dsneddon has joined #openstack-ironic15:41
* iurygregory tries to set callback on the patch to see how CI will run15:45
*** dsneddon has quit IRC15:46
*** jawad_axd has joined #openstack-ironic15:49
*** lucasagomes has quit IRC15:50
*** belmoreira has quit IRC15:53
*** jawad_axd has quit IRC15:53
TheJuliasomething about that seems... familiar15:54
iurygregoryops, its comment the callback .-.15:56
iurygregorynot sure if this would help15:56
iurygregory# uncomment if you have problem with ramdisk locale on ansible >= 2.115:56
iurygregory#module_set_locale=False15:56
iurygregorynot sure about this one15:57
*** dsneddon has joined #openstack-ironic15:58
rpittaubye, have a nice evening! o/16:03
*** rpittau is now known as rpittau|afk16:03
openstackgerritRiccardo Pittau proposed openstack/ironic-inspector master: [WIP] use openstacksdk for ironic module  https://review.opendev.org/67217916:05
*** dsneddon has quit IRC16:07
*** jawad_axd has joined #openstack-ironic16:10
arne_wiebalckbye everyone! o/16:12
TheJuliagoodnight arne_wiebalck16:12
* arne_wiebalck will be off for the rest of the week16:12
TheJuliaarne_wiebalck: enjoy16:12
arne_wiebalckTheJulia: thanks!16:12
*** dtantsur|brb is now known as dtantsur16:13
*** jawad_axd has quit IRC16:14
*** e0ne has quit IRC16:16
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Allow changing DIB_RELEASE and add a job with CentOS 8  https://review.opendev.org/68866116:17
*** dsneddon has joined #openstack-ironic16:23
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Last resort fallback to find a partition  https://review.opendev.org/68857516:24
*** dsneddon has quit IRC16:29
*** jawad_axd has joined #openstack-ironic16:30
*** igordc has quit IRC16:31
openstackgerritIlya Etingof proposed openstack/ironic-python-agent-builder master: Add static network configuration support  https://review.opendev.org/68840216:34
*** jawad_axd has quit IRC16:34
*** dsneddon has joined #openstack-ironic16:35
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Last resort fallback to find a partition  https://review.opendev.org/68857516:35
TheJuliadtantsur: ^^^ with story and bugfix reno so we can backport16:35
*** henriqueof has joined #openstack-ironic16:36
dtantsurthx! I'll try to check it the first thing tomorrow16:36
openstackgerritDmitry Tantsur proposed openstack/ironic master: [WIP] Wire in in-band cleaning for PXE boot  https://review.opendev.org/68872016:36
TheJuliadtantsur: no problem, I'm going to go back to hacking on some kexec support today. It just happens to be on the top of my braincells as a performance enhancement16:36
dtantsurkexec \o/16:37
*** dtantsur is now known as dtantsur|afk16:37
dtantsur|afkg'night16:37
TheJuliagoodnight!16:37
*** mkrai_ has quit IRC16:38
etingofdhcpless boot seem to work (in devstack)16:38
TheJulia\o/16:38
rpiosoetingof: Wow! \o/16:39
rpiosoetingof: Is that against physical bare metal?16:39
*** tssurya has quit IRC16:40
etingofrpioso, no, however it should work - it's not much different in this regard16:40
*** dsneddon has quit IRC16:40
rpiosoetingof: So sweet!16:40
etingofrpioso, you do not want to see how many -L and -R my ssh command line has16:40
*** ociuhandu_ has joined #openstack-ironic16:40
etingofif you get a chance to test my patch chain against a bare metal - that would be awesome16:41
etingofhowever the missing piece is neutron integration16:41
*** S4ren has quit IRC16:41
rpiosoetingof: LOL, re: your ssh command16:41
etingofonce it's in place, the testing would be easier and make more sense16:41
rpiosoetingof: I'm all in on testing it against bare metal. It may be a few days. Does that work?16:43
etingofrpioso, absolutely! I hope to wrap up by then16:44
rpiosoetingof: :-)16:44
*** ociuhandu has quit IRC16:44
rpiosoetingof: And congratulations!16:45
etingof(-;16:45
*** derekh has quit IRC16:47
*** ociuhandu_ has quit IRC16:47
*** dougsz has quit IRC16:55
*** ociuhandu has joined #openstack-ironic16:56
*** ociuhandu has quit IRC17:01
*** e0ne has joined #openstack-ironic17:02
*** dsneddon has joined #openstack-ironic17:03
*** priteau has quit IRC17:03
*** e0ne has quit IRC17:08
*** dsneddon has quit IRC17:13
*** alexmcleod has quit IRC17:13
*** goldyfruit_ has joined #openstack-ironic17:15
*** dsneddon has joined #openstack-ironic17:17
*** goldyfruit has quit IRC17:18
*** ricolin has quit IRC17:20
*** dsneddon has quit IRC17:22
*** ociuhandu has joined #openstack-ironic17:31
*** ociuhandu has quit IRC17:35
*** goldyfruit___ has joined #openstack-ironic17:36
*** goldyfruit_ has quit IRC17:39
*** e0ne has joined #openstack-ironic17:42
*** e0ne has quit IRC17:47
*** ociuhandu has joined #openstack-ironic17:48
*** jawad_axd has joined #openstack-ironic17:50
*** dsneddon has joined #openstack-ironic17:55
*** ociuhandu has quit IRC17:56
TheJuliarunning out for lunch and then a doctor appt, bbl18:03
*** goldyfruit_ has joined #openstack-ironic18:05
*** dsneddon has quit IRC18:06
*** openstackgerrit has quit IRC18:07
*** goldyfruit___ has quit IRC18:07
*** priteau has joined #openstack-ironic18:10
*** ociuhandu has joined #openstack-ironic18:26
*** igordc has joined #openstack-ironic18:36
*** dsneddon has joined #openstack-ironic18:46
*** ociuhandu has quit IRC18:55
*** iurygregory has quit IRC18:56
*** e0ne has joined #openstack-ironic18:59
*** henriqueof has quit IRC19:21
*** trident has quit IRC19:29
*** trident has joined #openstack-ironic19:33
*** tesseract has quit IRC19:49
*** e0ne has quit IRC19:55
*** jtwill98 has quit IRC20:19
*** jawad_axd has quit IRC20:21
*** pcaruana has quit IRC20:25
*** pcaruana has joined #openstack-ironic20:34
*** ociuhandu has joined #openstack-ironic20:46
*** pcaruana has quit IRC20:51
*** Goneri has quit IRC21:10
NobodyCamGood Afternoon Ironic Folks21:21
jrollmorning NobodyCam21:24
jrolletingof: that is awesome \o/21:24
NobodyCamHey hey jroll Happy not Monday ! :)21:24
jrollheh, truth!21:25
*** ociuhandu has quit IRC21:40
*** TxGirlGeek has quit IRC21:46
TheJuliablah, what a fun day!21:59
TheJuliaNobodyCam: the snow birds are arriving... and the traffic is getting disturbing.21:59
NobodyCamhehehehe oh ya21:59
jrollyou're welcome :D22:00
jrollkidding, our snow birds go to FL22:00
TheJuliaheh22:00
NobodyCam:p22:00
TheJuliaYeah, mainly washington plates... but I spotted a Florida plate the other day!22:01
NobodyCamWe get a lot from WA and Canada22:01
*** goldyfruit_ has quit IRC22:08
clarkbthe ones from oregon probably never updated their plates to oregon after moving here from california. Stealth mode snow birds22:09
*** jainiti has quit IRC22:28
NobodyCamLoL22:29
TheJuliaclarkb: heh22:29
TheJuliaclarkb: this would explain a lot....22:29
*** rcernin has joined #openstack-ironic22:30
clarkbwe have a lot of california plates up here22:30
*** openstackgerrit has joined #openstack-ironic22:32
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Crazy idea: kexec  https://review.opendev.org/68858122:32
JayFTheJulia: back in the day, we had a kexec POC we did downstream in OnMetal... basically we didn't use it in the end because (at least at the time) (1) Hardware support was not guaranteed. Some devices would just not initialize properly after kexec. and (2) you had to use *identical* kernel versions22:38
JayFTheJulia: I wonder how much of that has changed, but figured I'd pass along that22:38
*** bnemec has quit IRC22:38
TheJuliaJayF: I was totally thinking a "use at your own risk" kind of feature, specifically folks that use it a lot already want it :)22:38
JayFyour code appears to bake in that kernel assumption by re-using the ramdisk/kernel22:39
JayFTheJulia: in-band reboot to image --> hardware manager? lol22:39
JayFTheJulia: so it can be customized via plugin? :D22:39
TheJuliaI also wanted to be able to restart the agent....22:39
TheJuliadidn't really think of that, I was just kind of thinking of hacking it out as I would do it and see what some of those larger operators thought22:40
JayFYeah, just passing it on :). We tried it, got it to work-ish, but hit enough real-world problems that we punted on it22:40
TheJulia:)22:42
*** verdurin has quit IRC23:11
*** verdurin has joined #openstack-ironic23:19

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!