Thursday, 2018-08-09

*** gyee has quit IRC00:00
*** r-daneel has quit IRC00:10
*** moshele has quit IRC00:35
*** jesusaur has quit IRC00:42
bdoddTheJulia: Yes, when I updated my master branch, rebased and did git review again it seemed to resolve the merge failure.00:46
*** ddurst- is now known as ddurst01:04
*** phuongnh has joined #openstack-ironic01:09
*** rmcallis has joined #openstack-ironic02:39
*** rmcall has quit IRC02:41
openstackgerritKaifeng Wang proposed openstack/ironic-inspector master: Support manipulating iptables filter under IPv6 network  https://review.openstack.org/59007602:55
*** rmcall has joined #openstack-ironic02:56
*** rmcallis has quit IRC02:57
*** jesusaur has joined #openstack-ironic03:38
*** e0ne has joined #openstack-ironic04:11
*** e0ne has quit IRC04:24
*** gkadam has joined #openstack-ironic04:28
openstackgerritNguyen Van Trung proposed openstack/ironic master: Update docs to portgroup with creating windows images  https://review.openstack.org/58561304:46
*** jesusaur has quit IRC04:46
*** jesusaur has joined #openstack-ironic04:48
*** galaxyblr has joined #openstack-ironic04:55
*** galaxyblr has quit IRC05:51
openstackgerritKaifeng Wang proposed openstack/ironic master: Reformat instructions related with various OS  https://review.openstack.org/59010905:57
*** hamzy_ has quit IRC06:29
*** hamzy_ has joined #openstack-ironic06:29
*** pcaruana has joined #openstack-ironic06:38
*** etingof has quit IRC06:42
*** hshiina has joined #openstack-ironic06:56
*** ajya has joined #openstack-ironic07:15
*** rcernin has quit IRC07:16
openstackgerritAija Jaunteva proposed openstack/sushy master: Add support for loading resources from archive file  https://review.openstack.org/58914707:32
*** ajya has quit IRC07:38
*** ajya has joined #openstack-ironic07:39
*** juhak has left #openstack-ironic07:43
*** etingof has joined #openstack-ironic07:45
*** skazi has quit IRC07:54
*** lucasagomes has joined #openstack-ironic08:12
openstackgerritMerged openstack/ironic-specs master: Follow-up to owner field specificaiton  https://review.openstack.org/58835608:17
*** tssurya has joined #openstack-ironic08:19
*** skazi has joined #openstack-ironic08:23
*** dougsz_ has joined #openstack-ironic08:26
*** ibmko has quit IRC08:28
*** dtantsur|afk is now known as dtantsur08:33
dtantsurgood chilly (sarcasm) morning ironic08:33
*** skazi has quit IRC08:36
etingofgood morning, dtantsur o/08:37
etingofwish today is the pinnacle of this heat madness, it's gonna cool down afterwards08:38
dtantsuryep. well, it's nice and cool today - "just" 32 :D08:44
*** priteau has quit IRC08:47
etingofthey promised 36 over here today \o/08:47
*** priteau has joined #openstack-ironic08:47
*** mgoddard has joined #openstack-ironic09:00
dtantsurugh, I'm sorry for you :(09:11
*** jaganathan has joined #openstack-ironic09:28
openstackgerritShivanand Tendulker proposed openstack/ironic master: A minor update to documentation of `ilo` hardware type  https://review.openstack.org/59020309:33
*** derekh has joined #openstack-ironic09:35
openstackgerritShivanand Tendulker proposed openstack/ironic master: A minor update to documentation of `ilo` hardware type  https://review.openstack.org/59020309:58
*** serlex has joined #openstack-ironic09:58
*** gkadam has quit IRC10:13
*** gkadam has joined #openstack-ironic10:13
* lucasagomes is sorry for dtantsur and etingof 10:18
dtantsurhey lucasagomes, how is it going? chilly at your place? :)10:18
lucasagomes17C here today (-:10:18
dtantsurhah, nice10:18
lucasagomesno rain which is great too10:19
lucasagomesbut lets see, it won't last long I think10:19
openstackgerritNguyen Van Trung proposed openstack/ironic master: Fix not exist deploy image within irmc-virtual-media booting  https://review.openstack.org/58970210:33
*** hshiina has quit IRC10:34
*** Tomatosoup- has joined #openstack-ironic10:42
Tomatosoup-hello10:43
Tomatosoup-I'm trying to inspect (fake) baremetal node using ironic-inspector, where by fake i mean that it is a virtual machine with virtualbmc configured10:44
Tomatosoup-using ipmitool I can check the chassis power status of the VM with virtualbmc10:45
Tomatosoup-however `ipmitool power on` is failing by no obvious reason10:45
Tomatosoup->Set Chassis Power Control to Up/On failed: Command not supported in present state10:45
Tomatosoup-thus, also ironic inspection fails10:46
Tomatosoup-anyone handled such issue, or configured successfully virtualbmc?10:46
etingofTomatosoup-, this indicates a libvirt error, could you check vbmc log?10:49
Tomatosoup-etingof, where may i find vbmc logs?10:51
etingofTomatosoup-, no traces in journalctl/syslog?10:55
etingofTomatosoup-, to enable file logging you need to create an ini file with the name /etc/virtualbmc/virtualbmc.conf containing smth like:  [log]\nlogfile=/var/log/virtualbmc/vbmc.log10:55
etingofTomatosoup-, alternatively, you can try to check out libvirt logs10:56
Tomatosoup-already checked, libvirt logs just contain entries like >2018-08-09 10:55:36.052+0000: shutting down, reason=failed10:57
Tomatosoup-creating vbmc conf file as you suggested10:58
Tomatosoup-etingof, created the file as you suggested, although, after running ipmitool commands it remains empty11:03
etingofTomatosoup-, vbmc restarted?11:03
Tomatosoup-etingof, that must be it, but how to restart vbmc if it isn't managed by systemd?11:05
etingofTomatosoup-, just kill all vbmc processes11:05
etingofTomatosoup-, and then `vbmc start <domain>`11:06
Tomatosoup-etingof, Error: Cannot get interface MTU on 'ironic-pxe': No such device11:07
etingofTomatosoup-, seems this has smth to do with libvirt domain network configuration...11:12
Tomatosoup-etingof, might be, are there any special requirements on the domain network except from the obvious fact that it must be reachable from the node that we use ipmitool?11:13
etingofTomatosoup-, this domain can't get up and running. can you start it with virsh? I'd probably look into its network configuration to make sure it refers to existing host network interfaces11:16
Tomatosoup-etingof, ah, i think i know the origin of the error11:18
Tomatosoup-etingof, used default interface name config from openstack-helm11:18
Tomatosoup-etingof, thanks for all the help :)11:18
etingof\o/11:18
*** fultonj has quit IRC11:21
*** fultonj has joined #openstack-ironic11:23
*** phuongnh has quit IRC11:28
*** sambetts|afk has quit IRC11:29
*** sambetts_ has joined #openstack-ironic11:31
*** sambetts_ is now known as sambetts11:35
*** sambetts is now known as sambetts|afk11:35
*** bfournie has quit IRC11:38
*** sambetts|afk has quit IRC11:46
*** sambetts_ has joined #openstack-ironic11:47
TheJuliagood morning11:52
*** rh-jelabarre has joined #openstack-ironic11:52
etingofgood morning TheJulia o/11:53
dtantsurmorning TheJulia! do we have releases to do today? /me suspects ALL OF THEM :)11:54
TheJuliadtantsur: I created them all, -1'ed the ironic one for a single patch, and they also asked for cycle hilights... so I'll be doing that as soon as I have coffee11:54
dtantsurnice!11:55
TheJuliadtantsur: well, all minus ironic-staging-drivers11:55
dtantsuryep, I can do that11:55
TheJuliaI think we need to go ahead and pull that into our governance this next cycle, just very explicitly place the context around that11:55
TheJuliapas-ha: thoughts on ^^^11:55
pas-hano objections from me11:55
dtantsurI have mixed feelings. We need to make sure people don't treat it as official support from us.11:56
pas-haCAPS BOLD LETTERS11:56
pas-haDISCLAIMER11:56
dtantsuryeah11:56
TheJuliadtantsur: that would be awesome, since I think my key for signing commits is on one of the hard drives someplace in the back of the bus....11:56
pas-haand LOG.warning on import :-)11:56
dtantsurwe can set supported=False on them11:57
TheJuliapas-ha: and somehow make those wrapped in <h1></h1>11:57
pas-hathat's definitely11:57
pas-hare supported = False11:57
TheJuliathis sounds like we have a consensus and a plan \o/11:57
TheJuliaAnyway, Coffee time11:58
*** sambetts_ has quit IRC12:04
*** sambetts_ has joined #openstack-ironic12:04
* dtantsur releases staging-drivers12:12
*** bfournie has joined #openstack-ironic12:12
*** MattMan has quit IRC12:12
*** MattMan has joined #openstack-ironic12:12
TheJuliadtantsur: thanks12:13
TheJuliammmm coffeeeeee12:13
TheJuliaa couple quick reviews https://review.openstack.org/#/c/589702/ https://review.openstack.org/59020312:16
patchbotpatch 589702 - ironic - Fix not exist deploy image within irmc-virtual-med...12:16
patchbotpatch 590203 - ironic - A minor update to documentation of `ilo` hardware ...12:16
dtantsurTheJulia: I'd prefer the 3rdpary CI passing on the former12:22
*** trown|outtypewww is now known as trown12:22
TheJuliaack, I'm thinking we just release and backport12:22
TheJuliaHate to do that, but c'est la vie12:22
dtantsurTheJulia: yes, both changes are fine to backport.12:23
openstackgerritDmitry Tantsur proposed openstack/ironic-specs master: Rework the inspector boot management spec  https://review.openstack.org/58923012:27
dtantsuranother take at making ^^^ saner12:27
TheJuliadtantsur: yeah, just wanting to avoid backports at this point but it is unavoidable :)12:32
TheJuliadtantsur: please review https://review.openstack.org/590289   I left out removing classic drivers as I didn't want anything to be perceived as a negative change.12:34
patchbotpatch 590289 - releases - Ironic cycle highlights for rocky12:34
*** markguz_ has joined #openstack-ironic12:34
*** sambetts_ has quit IRC12:36
*** sambetts_ has joined #openstack-ironic12:37
dtantsurTheJulia: reviewed12:37
TheJuliadtantsur: gah, ironic-staging-drivers releaese failed12:38
*** markguz_ has quit IRC12:38
dtantsurugh12:39
dtantsurany link to logs?12:39
TheJuliathe instructions have been posted to the mailing list, it appeared in the last couple days12:39
dtantsurlucasagomes: hey, do you still own readthedocs for ironic-staging-drivers?12:40
lucasagomesdtantsur, lemme check but I think so12:41
lucasagomesdtantsur, yup I do12:41
dtantsurlucasagomes: mind adding Julia and me? I'm dtantsur there12:42
lucasagomesdtantsur, not at all!12:42
lucasagomesdtantsur, added. Do you know Julia's username ?12:42
lucasagomesTheJulia, ^12:42
*** oanson has joined #openstack-ironic12:42
TheJuliaI don't have a username there12:43
lucasagomesah right12:43
dtantsurTheJulia: okay, I'll handle it for now.12:43
TheJuliadtantsur: ack12:43
dtantsurTheJulia: https://review.openstack.org/59029512:45
patchbotpatch 590295 - ironic-staging-drivers - Configure webhook for publishing docs12:45
dtantsurfeel free to approve, there's nothing much to review there..12:45
*** ianychoi has quit IRC12:49
dtantsurTheJulia: re the highlights patch: are deploy steps really exposed to operators?12:53
dtantsurI thought it's all refactoring so far.. well, modulo the deploy_step node's field, which is not overly useful yet.12:54
TheJuliadtantsur: I thought we enabled posting it api for deployment12:54
TheJuliawhich to me seems exposed12:54
*** sambetts_ has quit IRC12:54
Tomatosoup-etingof, if you are still eager to help, I've adjusted network configs but now other error occur `libvirt: QEMU Driver error : unsupported configuration: append not supported in this QEMU binary`12:54
TheJuliaerrr, maybe I have old python-ironicclient12:55
dtantsurTheJulia: posting requested deploy steps? I'm not sure.12:55
TheJuliadtantsur: checking12:56
Tomatosoup-might be connected with qemu image i'm creating, not sure tho12:56
*** sambetts_ has joined #openstack-ironic12:56
TheJuliaerrrrrr, we might not have :\12:56
TheJuliarutro, we didn't :(12:57
* TheJulia sighs12:57
dtantsurTheJulia: yeah, I think we never got to it12:57
TheJuliaWe must have gotten pretty far though :\12:58
dtantsurwe have good chances of finishing it in stein, I think12:58
TheJuliavery good I think12:59
TheJuliaupdated the patch12:59
TheJuliaand again... ;)13:03
etingofTomatosoup-, looking at the source, this error might be related to some chardev append property, but I have no idea what that is -- https://github.com/libvirt/libvirt/blob/master/src/qemu/qemu_command.c#L503213:08
openstackgerritOpenStack Release Bot proposed openstack/ironic stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59030913:08
openstackgerritOpenStack Release Bot proposed openstack/ironic stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59031013:08
openstackgerritOpenStack Release Bot proposed openstack/ironic master: Update reno for stable/rocky  https://review.openstack.org/59031113:08
dtantsur\o/13:09
dtantsurTheJulia: any plans for stable releases? I'm quite tempted to release queens13:09
TheJuliadtantsur: I think so, I think we need to take a quick look at richard's backport first13:10
dtantsurTheJulia: right. I'm a bit worried it did not pass one of their jobs..13:10
TheJuliaAhh! was it the job they added?13:10
dtantsurand thanks to the creative tls certificate, I cannot check why it failed13:10
dtantsurno, the non-UEFI one (IIRC)13:10
TheJuliaugh...13:10
etingofTomatosoup-, possibly something about this option? `-chardev file,append=on|off`13:11
*** markguz_ has joined #openstack-ironic13:11
TheJuliaHas anyone reached out to Arkady regarding the certificate?13:11
TheJuliabecause I think I've raised the issue 3 or more times before13:11
* dtantsur too13:12
dtantsurI recall talking to someone13:12
Tomatosoup-etingof, during qemu image creation, right?13:13
openstackgerritOpenStack Release Bot proposed openstack/networking-generic-switch stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59031213:13
openstackgerritOpenStack Release Bot proposed openstack/networking-generic-switch stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59031313:13
openstackgerritOpenStack Release Bot proposed openstack/networking-generic-switch master: Update reno for stable/rocky  https://review.openstack.org/59031413:14
openstackgerritOpenStack Release Bot proposed openstack/networking-baremetal stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59031513:14
openstackgerritOpenStack Release Bot proposed openstack/networking-baremetal stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59031613:14
openstackgerritOpenStack Release Bot proposed openstack/networking-baremetal master: Update reno for stable/rocky  https://review.openstack.org/59031713:14
openstackgerritOpenStack Release Bot proposed openstack/ironic-ui stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59031813:14
openstackgerritOpenStack Release Bot proposed openstack/ironic-ui stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59031913:14
openstackgerritOpenStack Release Bot proposed openstack/ironic-ui master: Update reno for stable/rocky  https://review.openstack.org/59032013:14
openstackgerritOpenStack Release Bot proposed openstack/ironic-python-agent stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59032113:15
openstackgerritOpenStack Release Bot proposed openstack/ironic-python-agent stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59032213:15
openstackgerritOpenStack Release Bot proposed openstack/ironic-python-agent master: Update reno for stable/rocky  https://review.openstack.org/59032313:15
TheJuliaall the changes!13:15
TheJuliaand there will be more13:15
*** sambetts_ has quit IRC13:15
dtantsurlol, yeah13:16
openstackgerritOpenStack Release Bot proposed openstack/ironic-inspector stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59032413:16
openstackgerritOpenStack Release Bot proposed openstack/ironic-inspector stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/59032513:16
TheJuliaAnd somehow my pxe/ipxe patch is already in merge conflict again....13:16
openstackgerritOpenStack Release Bot proposed openstack/ironic-inspector master: Update reno for stable/rocky  https://review.openstack.org/59032613:16
TheJulia:(13:16
TheJuliaAll, I've removed -2s blocking items from landing to master that were changes. Effectively we're no open for Stein features/items/changes to land to master.13:17
TheJulias/no/now/13:17
etingofTomatosoup-, well, that code snippet I stumbled upon is about the qemu-system command itself, not qemu-img I think...13:17
TheJuliadtantsur: I guess this means next week will be nothing but specs :)13:18
*** sambetts_ has joined #openstack-ironic13:18
dtantsurtrue13:18
openstackgerritJulia Kreger proposed openstack/virtualbmc master: Change jobs for Python3 effort  https://review.openstack.org/58929513:20
TheJuliadtantsur: ^^^ removed the test reno to ensure that the job worked13:20
openstackgerritMerged openstack/ironic master: Follow-up to always link MAC address files  https://review.openstack.org/58926513:21
*** markguz_ has quit IRC13:21
dtantsurTheJulia: I think we're supposed to run all jobs when zuul.d is modified anyway13:22
*** sambetts_ has quit IRC13:22
TheJuliaI think the reno stuff limits it to release notes changes. Regardless the destructions indicated that it may be necessary to self test the change.13:22
*** sambetts_ has joined #openstack-ironic13:23
dtantsurokie, we'll know it soon13:24
TheJuliait didn't...13:24
TheJuliabut yeah13:24
TheJuliaanyway13:24
TheJuliafive gazillion things to do next13:24
*** jcoufal has joined #openstack-ironic13:27
*** jcoufal_ has joined #openstack-ironic13:28
*** jcoufal has quit IRC13:31
*** sambetts_ has quit IRC13:37
*** sambetts has joined #openstack-ironic13:38
*** jistr is now known as jistr|call13:39
*** baha has joined #openstack-ironic13:52
*** sambetts has quit IRC13:55
*** sambetts_ has joined #openstack-ironic13:58
*** jistr|call is now known as jistr13:59
*** amotoki_ is now known as amotoki14:01
*** markguz_ has joined #openstack-ironic14:03
dtantsurTheJulia: do you plan on including https://review.openstack.org/#/c/590337/ into the release?14:04
patchbotpatch 590337 - ironic-staging-drivers - Clarify status for ironic-staging-drivers14:04
TheJuliano, just wanted to go ahead and get that out of the way since I had the repo in front of me. I think it can merge anytime14:04
TheJuliaI really need to stop touching the same lines of code across multiple streams of patches. :\14:05
*** sambetts_ has quit IRC14:05
*** jaypipes has quit IRC14:06
*** sambetts_ has joined #openstack-ironic14:07
*** jaypipes has joined #openstack-ironic14:10
dtantsurTheJulia: ugh, we somehow ended up using the official theme: https://ironic-staging-drivers.readthedocs.io/en/latest/14:14
dtantsurtime to another fix!14:14
TheJuliadtantsur: propose it to governance ;)14:14
TheJulialol14:14
dtantsura bit too late for rocky :)14:14
* TheJulia feels slightly lazy, but yeah we need to fix that :(14:14
TheJuliayeah14:14
jrollmorning everyone14:15
* jroll still doesn't love the idea of bringing staging drivers into governance, that means we're responsible for them, at that point just bring them into the tree14:15
dtantsurmorning jroll14:15
*** ajya has quit IRC14:17
*** ajya has joined #openstack-ironic14:17
jrollTheJulia: let me know if you need any help with releasing things :)14:18
* jroll double checks the guide14:18
TheJuliawe'll have to do the release notes follow-ups once we merge the branch changes14:19
TheJuliasome of the steps we have, no longer apply or get done for us of course14:19
jrollTheJulia: did we do the release_mapping patch?14:20
TheJuliajroll: like two weeks ago :)14:20
jrolloh wow, ahead of the game :)14:20
*** sambetts_ has quit IRC14:21
openstackgerritJulia Kreger proposed openstack/ironic master: ipxe boot interface - begin the process  https://review.openstack.org/58340714:22
openstackgerritJulia Kreger proposed openstack/ironic master: Change BFV tempest job to explicitly use ipxe  https://review.openstack.org/58341014:22
TheJuliajroll: yeah, I just tried to get our ducks in a row so we didn't really have to wait or worry at the last minute14:23
TheJuliaIt is too much stress to have to deal with that at the point when we have to release14:23
TheJuliaNobodyCam: for your python3-ish reviewing pleaseure https://review.openstack.org/#/c/588325/14:24
patchbotpatch 588325 - bifrost - Begin change over to use python314:24
*** sambetts_ has joined #openstack-ironic14:24
*** gkadam_ has joined #openstack-ironic14:26
jrollTheJulia: I misread this one yesterday: https://review.openstack.org/#/c/589702/314:27
patchbotpatch 589702 - ironic - Fix not exist deploy image within irmc-virtual-med...14:27
jrollseems landable, I rechecked14:27
TheJuliaif you misread it, on my second patch I misread it too :\14:27
TheJuliapass, not patch14:27
jrollheh14:27
jrollyou had me guiding you incorrectly14:27
jrolloh, I skimmed over openstackgerrit's messages14:28
jrollcongrats on the rocky branch cut \o/14:28
TheJuliaThanks! We should all be really proud. We've done some amazing work this cycle!14:29
jroll++14:30
*** gkadam has quit IRC14:30
openstackgerritJulia Kreger proposed openstack/networking-baremetal master: Changing CI job templates for python3-first  https://review.openstack.org/59034914:35
*** sambetts_ has quit IRC14:39
openstackgerritJulia Kreger proposed openstack/virtualbmc master: Update name for virtualbmc CI job  https://review.openstack.org/58929414:39
dtantsurbroken lower-constraints in ironic-staging-drivers \o/14:39
* dtantsur cries14:39
* TheJulia cries too14:40
*** sambetts_ has joined #openstack-ironic14:40
*** hongbin has joined #openstack-ironic14:41
dtantsurokay, proposed fixes, will cool down in a park while zuul is working14:42
*** rpioso|afk is now known as rpioso14:42
rpiosoGood morning14:42
dtantsurmorning rpioso14:43
*** dtantsur is now known as dtantsur|brb14:43
rpiosodtantsur|brb: o/14:43
openstackgerritMerged openstack/virtualbmc master: Enable debug logging in tests  https://review.openstack.org/58834514:45
*** priteau has quit IRC14:52
mgoddardhi TheJulia: do you know if bifrost supports UEFI? Looks like it was not finished: https://review.openstack.org/#/c/268808/314:53
patchbotpatch 268808 - bifrost - Initial support for EFI booting14:53
TheJuliamgoddard: Really trying to head down that path in my sparish time right now, there is a newer where I've been working towards getting UEFI booting VMs for testing, and there are some other dhcp settings. It is not hard to add, it is just hard to test14:54
mgoddardTheJulia: cool, would be a nice addition. Asking for a friend (verdurin)14:57
TheJuliamgoddard: yeah, I actually want to try and get some things like direct http boot running, but I'm unsure14:57
TheJuliait all started with an email from arm folks saying they wanted to use ironic, but in standalone scenarios, and were having some issues.14:58
TheJulia(This has since resulted in a few minor tweaks in the pxe code to make lives easier/better for standalone deployments)14:59
verdurinTheJulia: it worked perfectly with TripleO, once I had a new enough ipxe.efi for my hardware.15:01
mgoddardTheJulia: I have an experiment I've been wanting to try for a while - rip out the service deployment/config from bifrost and replace with kolla-ansible15:02
openstackgerritJulia Kreger proposed openstack/ironic master: Add conductor_group docs  https://review.openstack.org/58927715:02
mgoddardcould probably do that with about 10-20 lines of YAML config15:02
mgoddardkeep the node provisioning automation in bifrost15:02
TheJuliaverdurin: Yeah, for arm I had to build it :\  But I can't cross-arch build easily with existing tools to build ramdisks and deployment images :(15:03
TheJuliathat could be really interesting15:03
mgoddardthat's another story, I guess my thinking was it would be one less place to get this working and maintain15:03
TheJuliayeah15:05
TheJuliathe goal of bifrost is and has always been to facilitate easy standalone deployment and help ensure ironic doesn't horribly break it15:06
mgoddardit's a noble goal15:06
openstackgerritJulia Kreger proposed openstack/virtualbmc master: Update name for virtualbmc CI job  https://review.openstack.org/58929415:08
TheJuliaindeed15:09
TheJuliaalso kind of hard to maintain ;)15:09
mgoddardyeah15:09
NobodyCamGood Morning Ioronic'ers15:10
mgoddardwouldn't it be nice if adding neutron support was as simple as putting enable_neutron: true in a config file15:10
mgoddard(this is openstack, perhaps not quite that simple, but kolla gets close)15:10
mgoddardoh also, we could scale up to a multinode control plane15:12
*** sambetts_ has quit IRC15:13
NobodyCamTheJulia: Im confused ... it could be that I've not yet had coffee... in 588325 the run.yaml files use_python3 is exported as false but in install-deps.sh we say # Lets use python3 by default / USE_PYTHON3=${USE_PYTHON3:-true} ??? it very well could be my lack of coffee affecting my ability to think15:14
rpiosoNobodyCam: Good morning15:16
NobodyCamGood Morning rpioso :)15:16
*** sambetts_ has joined #openstack-ironic15:16
TheJuliaNobodyCam: some of the distros we have jobs for don't support python35 or 36. Or install 3.415:22
NobodyCamits the install-deps.sh file that caused me pause. I left it as a comment15:23
NobodyCamI need to actually wakeup with more coffee15:23
*** jcoufal_ has quit IRC15:24
TheJuliacoffee +++++15:24
NobodyCam:)15:24
NobodyCamit got a +1 at least until I can loop back and look at it again :)15:25
TheJuliamgoddard: that is true, would that bring value in your case? I think what really prevents me from adding new ironic feature support directly into the ansible modules is the overhead time of keeping things kind of just in a running state... so if it could be close to that easy... maybe it makes some sense15:27
TheJuliaNobodyCam: ack15:27
*** dtantsur|brb is now known as dtantsur15:29
*** jcoufal has joined #openstack-ironic15:30
dtantsurTheJulia: our unsupported message talks about deprecation, it may not be entirely correct: http://logs.openstack.org/37/590337/1/check/ironic-staging-drivers-dsvm-all-drivers/2207ff8/logs/screen-ir-cond.txt.gz?level=WARNING#_Aug_09_14_37_57_60559715:31
TheJuliaoooh! ahh!15:32
TheJuliaI guess that makes sense to change15:32
dtantsurTheJulia, pas-ha, need review on https://review.openstack.org/#/c/590352/ and https://review.openstack.org/#/c/590344/ for the release15:33
patchbotpatch 590352 - ironic-staging-drivers - Fix lower constraints15:33
patchbotpatch 590344 - ironic-staging-drivers - Do not use the official theme from oslosphinx15:33
dtantsurI'm also going to provide a fix for the "async" deprecation, stay tuned15:35
*** ajya has quit IRC15:35
*** ajya has joined #openstack-ironic15:35
dtantsurTheJulia, pas-ha, https://review.openstack.org/59040115:36
patchbotpatch 590401 - ironic-staging-drivers - Do not use deprecated async parameter for passthru15:36
dtantsurvdrok: ^^^15:36
TheJuliawheeeeee15:37
*** priteau has joined #openstack-ironic15:37
mgoddardTheJulia: I think I need to try it to find out where it falls short, but it could be promising. Limiting the scope of bifrost to tools that build images and interact with the ironic API might help to reduce the maintenance overhead, and allow you to focus on improving that experience15:41
TheJuliamgoddard: worth a try at least15:43
prometheanfireglad lower-constraints are being worked on15:45
dtantsurmgoddard: neutron support is not easy, IIRC ansible modules do not support vif attach/detach15:45
dtantsurone of the reasons I started writing metalsmith (in python)15:45
TheJuliadtantsur: and that is only because I've not had time to get back to them15:47
dtantsurTheJulia: this is also because of proper microversions support in openstacksdk, which only appeared weeks ago :)15:47
TheJuliadtantsur: yeah, it has also become broken a few times silently15:47
TheJuliac'est la vie15:48
*** tssurya has quit IRC15:48
dtantsuryep15:48
mgoddardah, I see15:51
vdrokdtantsur: heya, yup, ill take a look at the patches to staging today15:51
*** ajya has quit IRC16:01
TheJuliadtantsur: I hvae a question for you on the conductor group patch16:01
dtantsursure16:02
rpiosoWhat's expected when the BIOS apply_configuration() clean step successfully completes? Should the requested settings been applied to the baremetal server or just recorded and available for application during the next system reset/reboot?16:02
dtantsurrpioso: my gut feeling it should apply the changes and set reboot_required16:03
TheJuliawhat dtantsur said16:03
rpiosodtantsur, TheJulia: How would an error during the application be captured by the user of the API?16:04
TheJuliaI _think_ one of the cases there is a hard need to reboot before exiting...  but ideally apply and set reboot required16:04
TheJuliarpioso: CLEAN_ERROR and last_error field16:04
TheJuliaI'm going to take a break, bbiab16:08
rpiosodtantsur, TheJulia: I'm not familiar with setting reboot_required.16:09
rpiosoWhere's that done?16:10
dtantsurrpioso: ah, it's called cleaning_reboot. grep for it, you even have it in iDRAC RAID.16:10
*** derekh has quit IRC16:14
jrollto be clear, reboot_required is a flag in the clean_step decorator16:15
jrollI think16:15
rpiosojroll: I looked there, but didn't see it. More coffee?16:16
jrollmaybe that's an IPA thing16:16
* rpioso is not as familiar with in-band cleaning16:16
jrollrpioso: looks like dtantsur is right about cleaning_reboot flag16:18
rpiosodtantsur: I see that in iDRAC RAID. However, I don't understand why it's there. The iDRAC configuration job had a reboot job associated with it. When its _resume_cleaning() is called, the jobs already completed.16:18
rpiosojroll: ^^^16:18
dtantsurwell, it's not a question for me, I guess. I don't know why this RAID support is written this way.16:18
jrollhrm16:19
rpiosodtantsur: I, too. I suppose some more investigation is needed.16:19
jrollrpioso: you mean the raid configuration clean step handles the reboot?16:20
rpiosodtantsur: Anyways, how can the general cleaning mechanism detect an error that occurs when the reboot is performed and provide useful info to the operator about the failure?16:20
dtantsurrpioso: I'm not sure from the top of my head16:21
dtantsuryou may have to trigger reboot right from the clean step16:21
jrollyeah, I've no idea where reboot_required went. it looks like now steps trigger the reboot themselves, then set cleaning_reboot to tell the conductor there was a reboot and the agent may not be back yet16:22
rpiosodtantsur: That's what I'm thinking. And not resume cleaning until the reboot is done and the result available to the clean step to report.16:22
dtantsurright16:23
rpiosojroll: Ahhh ... Right! It's to inform the conductor that the agent may not be back, yet.16:23
rpiosojroll: Then it makes sense in iDRAC RAID.16:24
jroll:)16:24
dtantsuraha, that's what it was16:25
rpiosodtantsur, jroll, TheJulia: I'll add my comment to https://review.openstack.org/#/c/582314/16:25
patchbotpatch 582314 - ironic - Add BIOS interface to Redfish hardware type16:25
rpiosoAnd thank you :)16:26
dtantsurnp16:26
jrollwelcome16:26
jrollI still want to know what happened to reboot_required16:26
TheJuliai think it was always ipa based in returned data16:30
TheJuliabut... long long long time ago16:31
TheJuliaanyway, break16:31
*** gkadam_ has quit IRC16:32
bdoddrpioso, dtantsur, jroll, TheJulia: Thanks for those comments/discussion. I don't fully understand them all yet, but I recorded them for further study. :-)16:34
*** hongbin has quit IRC16:37
rpiosobdodd: AFAIU, apply_configuration() should reset the server to apply the requested BIOS settings. Seems to me that it should also determine whether or not the application was successful, so it can report the result to the conductor. That would probs require a periodic function to check when the BIOS settings were last updated, if they're consistent with what was requested, etc. to determine whether or not the clean16:40
rpiosostep was successful.16:40
rpiosodtantsur: rajinir and I are working on the Dell EMC 3d party testing of the UEFI bug fix backports.16:49
etingofrpioso, bdodd iirc sushy has the `Bios.pending_attributes` property that effectively polls BMC to see if BIOS attrs are applied16:50
*** dougsz_ has quit IRC16:50
rpiosoetingof: polls?16:50
etingofrpioso, https://github.com/openstack/sushy/blob/master/sushy/resources/system/bios.py#L7416:51
*** skazi has joined #openstack-ironic16:53
etingofrpioso, it's all tied up with the Redfish `Settings` object16:55
rpiosoetingof: Doesn't that method sample the BMC pending attributes and refresh the cached copy?16:56
rpiosoetingof: Check out that method's docstring -- https://github.com/openstack/sushy/blob/master/sushy/resources/system/bios.py#L65-L6616:58
etingofrpioso, yes, AFAIU, that's the way for the client to figure out when requested BIOS update went through on the BMC side16:58
etingofe.g. that's what the prospective periodic function should probably call and compare with what ironic expects...16:59
rpiosoetingof: Exactly!16:59
rpiosoAt least one of the things. I believe there a timestamp that could be examined.17:00
etingofrpioso, yes, and etag17:00
rpiosoetingof: Cool :)17:02
*** mgoddard has quit IRC17:02
etingofrpioso, iirc using etag and/or time is racy e.g. when you have 1+ client tackling BIOS settings more or less at the same time...17:03
etingofrpioso, on the other hand, BMC is not requires to show the whole set of attributes, still it can allow some attributes to be modified17:04
etingofrpioso, in the latter case attributes comparison approach may not work...17:04
* etingof feels this part of Redfish is still somewhat fishy17:05
rpiosoetingof: From an ironic conductor perspective, wouldn't multiple clients changing BIOS at the same time be kind of broken? It's risky enough when one client changes BIOS settings :)17:06
rpiosoetingof: In other words, practically speaking, aren't those things good enough for our purposes?17:07
rpiosodtantsur: I'm ~100% certain those test failures are false negatives. The backports were soooooo simple. There was a one line conflict at the top of the unit test module. The production module was clean.17:09
etingofrpioso, it is hard to say...17:10
etingofrpioso, for example, could it be racy even with a single client? what if the client times out and retries faster than BMC applies updates...17:11
dtantsurrpioso: is it worth rechecking?17:12
rpiosodtantsur: Not, yet. We're still investigating the root cause. It doesn't appear to be related to the change.17:13
dtantsurTheJulia: wdyt ^^?17:14
rpiosoFrom screen-ir-cond.txt: Error finding the disk or partition device to deploy the image onto: No suitable device was found for deployment - root device hints were not provided and all found block devices are smaller than 4294967296B.17:14
dtantsurmm, does not see disks? curious17:14
TheJuliathat flavor is a partition image deploy with what? 40gb requested?17:15
TheJuliathat does seem odd...17:15
TheJuliaAnyway, rpioso if you guys feel 100% confident it is unrelated, I'm okay with it, but I'd hate for us to have to revert later or find out that something was really broken while you guys are still investigating17:16
TheJuliaalthough... that seems like a fairly distinct error :\17:16
rpiosodtantsur, TheJulia: And looking at the production code, management.py, its imports and interactions with other, external python code haven't changed.17:16
TheJuliacould a task still be running in the bmc from the boot mode being changed?17:17
rpiosoTheJulia: The boot mode isn't changed. It's configured to be UEFI for a looooong time.17:17
TheJuliarpioso: ack17:17
rpiosoThe idrac h/w type doesn't yet support the new management methods for configuring boot mode.17:18
rpioso:(17:18
TheJulia:(17:18
openstackgerritMerged openstack/ironic stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/59030917:18
rpiosoWhat's the deadline +/- for making it into the first release of Rocky?17:18
TheJuliarpioso: we cut 11.1.0 today.... as today is the RC1 deadline.17:19
dtantsurrpioso: we have stable/rocky already, so anything needs backporting.17:19
rpiosoOoops! Please disregard that silly question.17:19
* rpioso needs brain food17:20
TheJuliafood++17:20
* TheJulia gets some lunch17:21
rpiosoWe need it in a downstream Queens-based distro. The deadline for that is fast approaching.17:22
rpiosoThe prerequisite for that is the bug fix being available in stable/queens.17:23
dtantsurrpioso: how fast is fast? I think making the CI green is also quite a priority..17:23
rpiosodtantsur: We're actively working that :)17:24
rpiosoI don't know the lead time requirement.17:24
*** serlex has quit IRC17:24
rpiosoAnd the delivery date is an estimate.17:24
rpiosoSoon-ish, though.17:24
rpiosodtantsur, TheJulia: We'll dig in more and see how far we get.17:27
dtantsurgreat17:27
TheJuliarpioso: excellent17:27
*** gyee has joined #openstack-ironic17:47
verdurinTheJulia: I've managed to boot and (sort-of) enrol a node using EFI with Bifrost17:57
verdurinThere's something wrong with the ramdisk, so it's stuck in a loop, but Ironic has picked it up17:59
*** markguz_ has quit IRC18:01
*** priteau has quit IRC18:15
*** sambetts_ has quit IRC18:18
TheJuliaverdurin: fun :\18:21
TheJuliaverdurin: but also awesome18:21
*** dtantsur is now known as dtantsur|afk18:24
dtantsur|afko/18:24
*** skazi has quit IRC18:26
*** trown is now known as trown|lunch18:26
TheJuliagoodnight18:27
TheJuliajroll: I put an item on the meeting agenda for in the 21st to discuss ironic-staging-drivers, in the mean time I've -2'ed the patch19:06
jrollTheJulia: thanks19:06
TheJuliano problem :)19:08
*** skazi has joined #openstack-ironic19:12
openstackgerritJulia Kreger proposed openstack/ironic master: Add conductor_group docs  https://review.openstack.org/58927719:13
TheJuliaNobodyCam: replied to your questions on the py3 patch19:19
NobodyCamThank you, I hadn't had coffee when I looked at that :>19:20
*** jcoufal_ has joined #openstack-ironic19:26
TheJuliaI've not had any since... and I feel the need for coffee..19:27
*** ptx0 has quit IRC19:33
*** pcaruana has quit IRC19:33
*** jcoufal_ has quit IRC19:41
*** trown|lunch is now known as trown19:43
*** jcoufal_ has joined #openstack-ironic19:54
*** jcoufal_ has quit IRC19:55
*** ptx0_ has joined #openstack-ironic20:15
*** ptx0_ is now known as ptx020:20
*** trown is now known as trown|outtypewww21:04
*** jcoufal has quit IRC21:17
*** baha has quit IRC21:19
*** bfournie has quit IRC21:26
*** fultonj has quit IRC21:51
*** rcernin has joined #openstack-ironic22:09
*** jaganathan has quit IRC22:35
*** skazi has quit IRC22:44
*** gyee has quit IRC23:06
*** bfournie has joined #openstack-ironic23:13
*** dlbewley has quit IRC23:19
*** rpioso is now known as rpioso|afk23:24
*** dlbewley has joined #openstack-ironic23:45
*** dlbewley has quit IRC23:53
*** slagle has joined #openstack-ironic23:55

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