Friday, 2019-09-20

*** goldyfruit___ has joined #openstack-ironic00:01
*** igordc has quit IRC00:10
openstackgerritJulia Kreger proposed openstack/ironic master: Lower MTU override  https://review.opendev.org/68324900:25
*** mrda has quit IRC00:28
*** gyee has quit IRC00:28
*** jcoufal has joined #openstack-ironic00:31
*** mmethot_ has quit IRC00:40
*** mmethot has joined #openstack-ironic00:40
*** kaifeng has joined #openstack-ironic00:42
*** mmethot has quit IRC00:45
*** TxGirlGeek has quit IRC00:45
*** mmethot has joined #openstack-ironic00:45
*** mmethot has quit IRC00:46
*** mmethot has joined #openstack-ironic00:47
*** prometheanfire has left #openstack-ironic00:53
*** mmethot has quit IRC00:55
*** mmethot_ has joined #openstack-ironic00:55
*** rloo has quit IRC01:15
*** whoami-rajat has joined #openstack-ironic01:17
*** mrda has joined #openstack-ironic01:35
openstackgerritJulia Kreger proposed openstack/ironic master: wip: if wholedisk set boot dev to disk  https://review.opendev.org/68326201:52
openstackgerritGhanshyam Mann proposed openstack/ironic master: [train][goal] Define new 'ironic-tempest-ipv6-only' job  https://review.opendev.org/68269201:53
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Don't try to put a bootloader in place when bootable  https://review.opendev.org/67933301:54
openstackgerritJulia Kreger proposed openstack/ironic-python-agent master: Don't try to put a bootloader in place when bootable  https://review.opendev.org/67933301:55
*** ricolin has joined #openstack-ironic01:55
TheJuliadtantsur|afk: Ive decomposed two failures... and counted. MTU errors like... 70% of the time.01:56
TheJuliawell, two failed changes... maybe 301:56
TheJuliaAlso a boot_option insanity bug01:56
TheJuliadetails on the etherpad01:56
*** jcoufal has quit IRC01:56
* TheJulia goes and drinks now01:56
TheJuliadtantsur|afk: oh, https://review.opendev.org/683249 appears to resolve our MTU issues01:57
patchbotpatch 683249 - ironic - Lower MTU override - 2 patch sets01:57
*** dsneddon has quit IRC02:04
*** dsneddon has joined #openstack-ironic02:17
*** dsneddon has quit IRC02:22
kaifengmorning TheJulia, welcome back :)02:29
*** dsneddon has joined #openstack-ironic02:45
*** mkrai_ has joined #openstack-ironic02:45
*** mkrai_ has quit IRC02:50
*** mkrai has joined #openstack-ironic02:57
*** ricolin_ has joined #openstack-ironic03:04
*** bobmel has joined #openstack-ironic03:05
*** ricolin has quit IRC03:06
openstackgerritMerged openstack/ironic master: CI: move libvirt images to /opt for standalone and multinode jobs  https://review.opendev.org/68144603:06
*** bobmel has quit IRC03:10
*** TxGirlGeek has joined #openstack-ironic03:27
*** gkadam has joined #openstack-ironic03:58
*** gkadam has quit IRC03:58
*** dsneddon has quit IRC03:59
*** dsneddon has joined #openstack-ironic04:00
*** dsneddon has quit IRC04:05
*** rh-jelabarre has quit IRC04:32
*** _mmethot_ has joined #openstack-ironic04:41
*** mmethot_ has quit IRC04:41
*** TxGirlGeek has quit IRC05:07
*** tzumainn has quit IRC05:08
*** TxGirlGeek has joined #openstack-ironic05:11
*** pcaruana has joined #openstack-ironic05:28
*** dsneddon has joined #openstack-ironic05:34
*** dsneddon has quit IRC05:40
*** pcaruana has quit IRC05:53
*** ricolin_ is now known as ricolin05:54
*** jawad_axd has joined #openstack-ironic05:58
*** jawad_ax_ has joined #openstack-ironic06:02
*** jawad_axd has quit IRC06:03
*** TxGirlGeek has quit IRC06:08
arne_wiebalckGood morning, ironic!06:30
*** trident has quit IRC06:49
*** trident has joined #openstack-ironic07:01
*** trident has quit IRC07:07
*** bobmel has joined #openstack-ironic07:07
*** bobmel has quit IRC07:13
*** trident has joined #openstack-ironic07:17
*** yolanda has quit IRC07:21
*** yolanda has joined #openstack-ironic07:23
*** rcernin has quit IRC07:29
*** rpittau|afk is now known as rpittau07:31
rpittaugood morning ironic! o/07:31
rpittauTGIF!07:31
*** yolanda has quit IRC07:32
*** dsneddon has joined #openstack-ironic07:36
*** e0ne has joined #openstack-ironic07:46
*** yolanda has joined #openstack-ironic07:47
*** rachit7 has joined #openstack-ironic07:49
*** iurygregory has joined #openstack-ironic07:51
*** e0ne has quit IRC07:59
iurygregorymorning all08:00
iurygregoryits friday \o/08:00
rachit7iurygregory: morning08:02
*** dsneddon has quit IRC08:07
*** alexmcleod has joined #openstack-ironic08:14
rpittauhey iurygregory :)08:14
mgoddardmorning08:22
*** e0ne has joined #openstack-ironic08:24
rpittauhey mgoddard :)08:25
iurygregorymgoddard, o/08:25
mgoddardhi iurygregory & rpittau08:25
*** dsneddon has joined #openstack-ironic08:34
*** e0ne has quit IRC08:42
*** mkrai has quit IRC08:51
*** ociuhandu has joined #openstack-ironic08:52
*** dtruong has quit IRC08:54
*** problem_v has quit IRC08:54
*** rcernin has joined #openstack-ironic08:55
kaifengmorning all, hooray friday!08:58
*** dsneddon has quit IRC09:01
*** mkrai has joined #openstack-ironic09:05
*** dsneddon has joined #openstack-ironic09:06
mkrairpittau, mgoddard iurygregory rachit7 good morning!09:10
*** dsneddon has quit IRC09:10
iurygregorymorning kaifeng and mkrai09:10
mkraiDo we have any patch to fix ironic gate failure?09:11
mkraiit seems bad, almost all request failed :(09:11
iurygregoryaww the mtu one failed also =(09:12
*** pcaruana has joined #openstack-ironic09:12
mgoddardmorning mkrai09:15
*** igordc has joined #openstack-ironic09:15
mkraiiurygregory, Yes I just checked09:15
*** igordc has quit IRC09:20
*** derekh has joined #openstack-ironic09:27
*** dtantsur|afk is now known as dtantsur09:27
dtantsurmorning ironic09:27
dtantsurTheJulia: thanks!09:28
*** ociuhandu has quit IRC09:30
*** ociuhandu_ has joined #openstack-ironic09:30
iurygregorymorning dtantsur09:33
rpittauhey dtantsur :)09:33
openstackgerritpengyuesheng proposed openstack/networking-baremetal master: Update the constraints url  https://review.opendev.org/68333809:35
*** igordc has joined #openstack-ironic09:39
dtantsurTheJulia: could you point me to the place where you see local boot issues with RAID?09:41
arne_wiebalckTheJulia: Thanks for submitting the Forum sessions. I put basically all input into the PTG Planning etherpad ... not sure if some of the stuff there should better be a session.09:48
*** rcernin has quit IRC09:49
*** e0ne has joined #openstack-ironic09:54
etingofrpioso, o/ by way of addressing bdodd's comments, I've rebased my floppy & idrac.boot patches. however I am not pushing them yet because I thought may be you want to push your version instead?09:56
openstackgerritMark Goddard proposed openstack/networking-generic-switch master: Allow network name format to be configured  https://review.opendev.org/68318709:59
*** e0ne has quit IRC10:02
*** ociuhandu_ has quit IRC10:07
*** ociuhandu has joined #openstack-ironic10:07
*** ociuhandu has quit IRC10:14
*** ociuhandu has joined #openstack-ironic10:18
*** ociuhandu has quit IRC10:22
*** rachit7 has quit IRC10:26
*** ociuhandu has joined #openstack-ironic10:32
*** ociuhandu has quit IRC10:33
*** ociuhandu has joined #openstack-ironic10:34
*** ociuhandu has quit IRC10:39
*** ociuhandu has joined #openstack-ironic10:40
*** mkrai has quit IRC10:52
*** mkrai_ has joined #openstack-ironic10:52
*** pcaruana has quit IRC10:56
*** e0ne has joined #openstack-ironic11:03
openstackgerritMerged openstack/ironic-python-agent-builder master: CI: run a DIB job and make the inspector job non-voting  https://review.opendev.org/68264811:06
openstackgerritDmitry Tantsur proposed openstack/ironic master: [WIP] Allow retrying PXE boot if it takes too long  https://review.opendev.org/68312711:06
*** dsneddon has joined #openstack-ironic11:06
*** bobmel has joined #openstack-ironic11:10
dtantsurfolks, another small improvement for the CI: https://review.opendev.org/#/c/681166/11:11
patchbotpatch 681166 - ironic-tempest-plugin - Skip the manual cleaning test if software RAID tes... - 1 patch set11:11
openstackgerritDmitry Tantsur proposed openstack/ironic master: DNM try breaking PXE boot  https://review.opendev.org/68313111:11
*** rachit7 has joined #openstack-ironic11:11
openstackgerritMerged openstack/networking-baremetal master: Update the constraints url  https://review.opendev.org/68333811:13
*** bobmel has quit IRC11:15
*** lucasagomes has joined #openstack-ironic11:19
*** e0ne has quit IRC11:24
*** igordc has quit IRC11:26
*** ociuhandu has quit IRC11:27
*** pcaruana has joined #openstack-ironic11:29
*** dsneddon has quit IRC11:31
*** rh-jelabarre has joined #openstack-ironic11:32
*** pcaruana has quit IRC11:42
*** pcaruana has joined #openstack-ironic11:42
*** dsneddon has joined #openstack-ironic11:42
*** dmellado has quit IRC11:45
*** dsneddon has quit IRC11:47
*** dmellado has joined #openstack-ironic11:48
*** dsneddon has joined #openstack-ironic12:01
*** rachit7 has quit IRC12:02
*** dsneddon has quit IRC12:06
*** ociuhandu has joined #openstack-ironic12:06
*** ociuhandu has quit IRC12:06
*** ociuhandu has joined #openstack-ironic12:07
*** goldyfruit___ has quit IRC12:12
openstackgerritDmitry Tantsur proposed openstack/ironic master: Allow retrying PXE boot if it takes too long  https://review.opendev.org/68312712:14
*** rloo has joined #openstack-ironic12:34
*** ociuhandu has quit IRC12:40
*** ociuhandu has joined #openstack-ironic12:45
*** ociuhandu has quit IRC12:45
*** ociuhandu has joined #openstack-ironic12:46
*** mkrai_ has quit IRC12:46
*** mkrai has joined #openstack-ironic12:47
*** rcernin has joined #openstack-ironic12:56
*** ociuhandu has quit IRC12:57
*** ociuhandu has joined #openstack-ironic12:57
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Software raid: default partition table type change  https://review.opendev.org/67480313:07
*** dsneddon has joined #openstack-ironic13:08
*** ociuhandu has quit IRC13:09
*** jtwill98 has quit IRC13:12
openstackgerritDmitry Tantsur proposed openstack/ironic master: DNM try breaking PXE boot  https://review.opendev.org/68313113:12
*** ociuhandu has joined #openstack-ironic13:13
*** dsneddon has quit IRC13:13
*** ociuhandu has quit IRC13:16
*** ociuhandu has joined #openstack-ironic13:16
*** jawad_ax_ has quit IRC13:16
*** jawad_axd has joined #openstack-ironic13:17
*** jawad_axd has quit IRC13:22
TheJuliadtantsur:  the change default boot option to local change. It keeps getting rechecked when the underlying defect in the selection logic is still broken13:22
dtantsurTheJulia: sorry, maybe I'm too sleepy.. how does it cause random failures in the CI?13:22
TheJuliaIt doesn't, I was literally going down the list of most recently touched patches with failures13:23
TheJuliaFocus on MTU and fortnebula13:23
* TheJulia needs coffee and more coffee13:23
dtantsurah, I see13:24
openstackgerritDmitry Tantsur proposed openstack/ironic master: Allow retrying PXE boot if it takes too long  https://review.opendev.org/68312713:26
openstackgerritDmitry Tantsur proposed openstack/ironic master: DNM try breaking PXE boot  https://review.opendev.org/68313113:26
*** goldyfruit___ has joined #openstack-ironic13:28
*** tzumainn has joined #openstack-ironic13:32
rlooif anyone has any opinions on how to proceed with https://review.opendev.org/#/c/674803. Which seems to be basically undoing an earlier patch that landed in August.13:33
patchbotpatch 674803 - ironic-python-agent - Software raid: default partition table type change - 10 patch sets13:33
*** goldyfruit_ has joined #openstack-ironic13:35
*** goldyfruit___ has quit IRC13:35
TheJuliarloo: I feel like if arne +2'ed it, he is comfortable with it merging. Do you know the patch it undoes in August, there was some back and forth on fixes13:36
*** goldyfruit___ has joined #openstack-ironic13:36
rlooTheJulia: should we delete the release note from the original patch in August -- since this one basically undoes that feature, so the release note doesn't make sense any more.13:37
TheJuliarloo: your going to need to link the original patch from august13:37
rlooTheJulia: i suspect there is one line from that original patch, that is still in this new patch. the rest was deleted.13:37
rlooTheJulia: https://review.opendev.org/#/c/675275/. sorry, was in the review.13:38
patchbotpatch 675275 - ironic-python-agent - Software raid: mbr/gpt partition table alternative (MERGED) - 2 patch sets13:38
*** goldyfruit_ has quit IRC13:39
dtantsurmorning rloo13:41
rloohappy friday morning dtantsur :)13:41
dtantsurI seem to remember that the older patch was a less invasive part of the patch in question13:41
dtantsurmaybe things have changed since then13:41
rloodtantsur: yup, the code is fairly simple -- no way for user to override. i'm ok with that, but we can't have a release note saying we have a feature, and then remove the feature w/o ... ?13:46
dtantsurfair13:46
dtantsurit must have changed since I last saw it13:46
andreinhi guys, not sure if this is the place to ask questions about diskimage builder, but I'm hoping someone can help me out. I've built my own image based on the centos7 image, however, after my server boots from it the network service is reporting as failed (although everything works correctly). this happens because /etc/sysconfig/network-scripts/ifcfg-eth0 exists, but my interfaces are actually named eno1-4.13:47
*** mkrai has quit IRC13:48
andreinis there any way to convince DIB to remove that file from the image?13:48
*** mkrai_ has joined #openstack-ironic13:48
TheJuliadtantsur: If you look at https://storage.gra1.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_778/683262/1/check/ironic-tempest-ipa-wholedisk-direct-tinyipa-multinode/77869fb/testr_results.html.gz do you get the same feeling that the job was almost done?13:49
*** sthussey has joined #openstack-ironic13:49
dtantsurandrein: you may want to ask on #openstack-dib13:50
andreinthanks!13:50
dtantsurTheJulia: I've seen such errors. Indeed, everything is seemingly okay.13:51
dtantsurTheJulia: oh, btw, https://review.opendev.org/#/c/682609/ modifies something you've modified before13:51
patchbotpatch 682609 - ironic - Do not log an error on heartbeat in deploying/clea... - 3 patch sets13:51
TheJuliaxXraphXx: I've left a comment https://review.opendev.org/#/c/674803/10 you might want to take a look. I agree with rloo the prior release note needs to go. It also likely didn't render because it looks like it was created without the hash value for the docs build to be able to construct history.13:53
patchbotpatch 674803 - ironic-python-agent - Software raid: default partition table type change - 10 patch sets13:53
dtantsuraha, TheJulia: ERROR ironic.conductor.utils [None req-ec9d3a98-f2be-4c81-84f3-fcf70815ba90 None None] Node d3df3249-5e03-4410-8409-aad904dfcf0b failed deploy step {u'priority': 100, u'interface': u'deploy', u'step': u'deploy', u'argsinfo': None}. Error: IPMI call failed: power on.: IPMIFailure: IPMI call failed: power on.13:53
TheJuliawoot!13:53
dtantsurError: no response from RAKP 3 message\nError: Received an Unexpected RAKP 4 message\nError in open session response message : Unknown (0x6C)\n\nError: Unable to establish IPMI v2 / RMCP+ session\n'.: ProcessExecutionError: Unexpected error while running command.13:53
dtantsurI feel like adding retries for literally everything ironic does >_<13:53
TheJuliadtantsur: we already have retries for ipmi exec operations13:54
* TheJulia wrote the selective retry code13:54
dtantsurmmm, probably not for this one13:54
TheJuliamaybe instead, we burn ipmi with fire13:54
TheJuliano, but perhaps add it to the search list?13:54
* dtantsur pokes etingof13:54
* TheJulia gets out a pencil sharpener13:55
*** dsneddon has joined #openstack-ironic13:55
dtantsurI don't see anything abnormal in the virtualbmc logs, hmm13:55
* etingof hides in the cloud13:56
*** belmoreira has quit IRC13:56
dtantsurTheJulia: a very wild guess: can we have issues with IPMI when crossing between testing nodes?13:57
dtantsurto confirm that we need to check if this error ever happens on non-multinode jobs13:57
TheJuliathat is stupidly highly likely13:58
rpittauhey TheJulia welcome back :)14:00
*** belmoreira has joined #openstack-ironic14:00
*** dsneddon has quit IRC14:00
dtantsurTheJulia: do you know how to use http://logstash.openstack.org?14:01
TheJuliahttps://ef6fea84da73ed48af61-f850a1c88b63080f1e34c13fe4924008.ssl.cf2.rackcdn.com/683262/1/check/ironic-tempest-ipa-partition-redfish-tinyipa/782fb5b/controller/logs/screen-ir-api.txt.gz <-- ouch, the conductor became ready 4 seconds AFTER the devstack script blew up trying to enroll nodes :(14:01
dtantsurouch14:02
dtantsurdo we need a loop waiting for non-empty driver list?14:02
TheJuliabasically14:02
TheJuliait only tries six times seemingly in rapid succession14:03
TheJuliaquick fix14:03
*** mkrai_ has quit IRC14:04
TheJuliaoh  no, we're relying upon the client integrated retries14:04
TheJuliahmmm14:04
dtantsurhow long did it take to start?14:05
TheJuliastarted 2:38:35, ready 2:38:5414:05
TheJuliastill kind of a long time, but it looks like the db interactions were part of that. It could have just been slow on the back-end disk.14:06
dtantsur20 seconds. a lot, but not too much.14:06
dtantsuryep14:06
*** openstackgerrit has quit IRC14:06
TheJuliaI've typically seen ~10 seconds, so if we put a small wait, it would be appropriate to give the conductor time to actually fire up14:07
dtantsurlooping on `driver list` is what I often recommend people to do14:07
TheJuliaand then the odds of failure on enrolling nodes should be greatly reduced14:07
TheJuliaif you have some example code for that, we coudl do that, I was thinking just putting sleep 10 before enroll_nodes14:07
* TheJulia has no running ironic at the moment14:08
dtantsurI have devstack here, lemme try something14:08
TheJuliak14:08
TheJuliamgoddard: re: https://review.opendev.org/#/c/683104/2 we don't do beta releases which seems to make your change confusing14:10
patchbotpatch 683104 - bifrost - Switch git_branch to stable/train - 2 patch sets14:10
TheJulia*gasp* https://review.opendev.org/#/c/683262/ has a +1 from zuul14:11
patchbotpatch 683262 - ironic - wip: if wholedisk set boot dev to disk - 1 patch set14:11
mgoddardTheJulia: hmm, I see your point14:12
mgoddardthis is why deployment projects normally cycle-trailing14:13
TheJuliayeah, I'm fine if you want to cut a release and then a minor rev afterwards though14:13
mgoddardI don't mind too much if it doesn't make the first release14:13
mgoddardkolla won't need it for a while14:13
TheJuliaAnd bifrost is reporting ""msg": "Node 4e41df61-84b1-5856-bfb6-6b5f2cd3dd11 reached failure state \"inspect failed\"; the last error is Failed to start inspection: Version requested but version discovery document was not found and allow_version_hack was False"" now anyway14:14
TheJuliaI suspect a new version of ansible may be expecting different arguments.14:15
*** openstackgerrit has joined #openstack-ironic14:16
openstackgerritDmitry Tantsur proposed openstack/ironic master: devstack: wait for conductor to start and register itself  https://review.opendev.org/68340014:16
dtantsurTheJulia: like this ^^?14:16
TheJuliaor a library has broken something14:16
*** whoami-rajat has quit IRC14:17
TheJuliadtantsur: +214:17
*** dsneddon has joined #openstack-ironic14:17
TheJuliaLooks like our mtu fix is in the final stages to merge. If everyone would please hold off on rechecks so we can get some additional fixes in, it would be greatly appreciated14:19
*** iurygregory_ has joined #openstack-ironic14:21
TheJuliarloo: mgoddard: if you could review/approve https://review.opendev.org/#/c/683400/ it would be appreicated14:21
patchbotpatch 683400 - ironic - devstack: wait for conductor to start and register... - 1 patch set14:21
*** iurygregory has quit IRC14:21
* dtantsur hopes he did not make a silly mistake there :)14:22
TheJuliait looked good to me14:22
TheJuliaI'm a little worried about the conditional logic, but it will fail most likely if it is not good14:22
*** rcernin has quit IRC14:23
TheJuliathe theme and overall idea is what I +2'ed :)14:23
mgoddardlooking14:24
dtantsurthe conditional logic is cargo-culted from another place and tested locally14:24
TheJuliak14:24
TheJuliadtantsur: you indicated there was yet another gate fix earlier?14:25
rlooTheJulia: dtantsur: +A. thx!14:25
dtantsurhttps://review.opendev.org/#/c/683127/ may help with random PXE failures14:25
patchbotpatch 683127 - ironic - Allow retrying PXE boot if it takes too long - 6 patch sets14:25
mgoddarddtantsur: we don't have set -e in devstack, right?14:26
dtantsurI don't think we do14:26
dtantsurnot entirely sure though, why?14:26
mgoddardjust wondering if the driver list exits non-zero if there are no drivers14:27
mgoddardchecking14:27
dtantsurI don't think so14:27
rlooit shouldn't. it isn't an error if there are no entries.14:28
rloobut worth checking :)14:28
dtantsurhmmm, I remember something.. lemme try14:28
mgoddardconfirmed14:28
mgoddard014:28
TheJuliaheh14:28
dtantsuryep, it's okay14:29
dtantsurwe had a request to make it a failure, but we refused14:29
mgoddarddoes it matter that we don't check the hostname?14:30
TheJuliait doesn't matter as long as the conductor is up14:30
mgoddardi.e. is it enough in multinode to just have one conductor up?14:30
mgoddardI guess so14:30
dtantsurthey'll get created on one conductor and will rebalance later?14:30
TheJuliathey will, I've actually seen it happen14:31
dtantsurironic magic!14:32
mgoddardTheJulia: -2 for MTU :(14:32
dtantsurokay folks, I'm going to another flat visit, then straight for beers14:32
TheJulia#(#@!*@*#!(14:32
dtantsur:*14:32
dtantsur:(14:32
rpittaudtantsur: enjoy both :)14:32
dtantsurflat visits are rarely fun :)14:32
dtantsurhave a great weekend14:32
iurygregory_same error from before on mtu?14:33
TheJuliaany objection if we make the the multinodes non-voting for a few days?14:33
dtantsurwe probably have to :(14:33
dtantsurit becomes a pre-release tradition, should we add it to the check-list?14:33
TheJuliaThats kind o fhw hat I'm thinking, they bring their own unique races14:33
* dtantsur stops kidding and actually goes14:33
TheJuliao/14:33
TheJuliaAny other cores have feelings on doing so?!?14:33
*** dtantsur is now known as dtantsur|afk14:33
* TheJulia kind of agrees we should just doc it as a normal thing because I think I've had to do it for 3 out of the last six releases?!?14:35
*** rcernin has joined #openstack-ironic14:38
*** ociuhandu has quit IRC14:39
rpittauTheJulia: it's ok for me about the temp non-voting for multinodes14:41
mgoddardTheJulia: if we are confident these are CI issues and not ironic issues, then I'm +114:42
xXraphXxHello ironic o/. TheJulia, about https://review.opendev.org/#/c/674803/ -> answered, I'm actually a bit lost about the way it should be fixed now (sth like patchset 8 ?)... So, I don't want to block anything: should we just wait for it ? If so we may as well revert https://review.opendev.org/#/c/675275/14:43
patchbotpatch 674803 - ironic-python-agent - Software raid: default partition table type change - 10 patch sets14:43
patchbotpatch 675275 - ironic-python-agent - Software raid: mbr/gpt partition table alternative (MERGED) - 2 patch sets14:43
xXraphXx*wait for after Train I mean14:45
openstackgerritJulia Kreger proposed openstack/ironic master: Make multinode jobs non-voting  https://review.opendev.org/68341214:47
NobodyCamGood Morning Folks, and OFC...14:47
NobodyCamTGIF!14:47
TheJuliamgoddard: likely a mixed bag of things around race conditions :(14:47
*** ociuhandu has joined #openstack-ironic14:49
*** ociuhandu has quit IRC14:49
*** ociuhandu has joined #openstack-ironic14:50
iurygregory_something in common with the 2 failures: the host have the disk problem and we are using LIBVIRT_STORAGE_POOL_PATH, when we add this option the storyboard also mentioned that we should lower the concurrency for tempest https://storyboard.openstack.org/#!/story/200652014:52
*** iurygregory_ is now known as iurygregory14:53
*** TxGirlGeek has joined #openstack-ironic14:55
*** TxGirlGeek has quit IRC14:56
*** TxGirlGeek has joined #openstack-ironic14:57
*** dsneddon has quit IRC15:00
* iurygregory "but maybe not related since it worked on the check (2nd run) " CI why you do this =( 15:01
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Software raid: default partition table type change  https://review.opendev.org/67480315:02
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Fix doc rendering  https://review.opendev.org/68341915:03
*** _mmethot_ has quit IRC15:04
*** _mmethot_ has joined #openstack-ironic15:05
*** bnemec is now known as beekneemech15:06
*** belmoreira has quit IRC15:07
*** gyee has joined #openstack-ironic15:07
*** belmoreira has joined #openstack-ironic15:11
*** tzumainn has quit IRC15:11
rlooTheJulia: xXraphXx: wrt https://review.opendev.org/#/c/674803/. I don't think/know if we should try to get it into Train. Having said that, if you/we feel that we don't want (or are not sure if we want) https://review.opendev.org/#/c/675275/ -- I would prefer if we reverted that before doing a Train release.15:12
patchbotpatch 674803 - ironic-python-agent - Software raid: default partition table type change - 11 patch sets15:12
patchbotpatch 675275 - ironic-python-agent - Software raid: mbr/gpt partition table alternative (MERGED) - 2 patch sets15:12
openstackgerritWill Szumski proposed openstack/tenks master: Add debugging statements to wait for resources loop  https://review.opendev.org/65902515:14
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Revert "Software raid: mbr/gpt partition table alternative"  https://review.opendev.org/68342115:14
*** trident has quit IRC15:14
xXraphXxrloo, sure, done. Thanks for your feedback15:16
arne_wiebalckrloo: I agree. xXraphXx: thanks!15:18
*** goldyfruit___ has quit IRC15:21
*** trident has joined #openstack-ironic15:24
*** goldyfruit has joined #openstack-ironic15:26
*** trident has quit IRC15:29
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Revert "Software raid: mbr/gpt partition table alternative"  https://review.opendev.org/68342115:32
arne_wiebalckrloo: xXraphXx: This however means we won't have UEFI/gpt support for s/w RAID in Train?15:33
xXraphXxyep15:33
xXraphXxbut anyway, these pr would not have been enough by themselves15:33
arne_wiebalckah, no?15:33
arne_wiebalcknot having this support for the first release supporting s/w RAID is a pity though15:34
TheJuliaI actually would like to see it15:35
TheJuliaI get why, and if we merge the second patch that while I have a slight issue with it, ultimately thats just a bug we could fix after the initial release since bios boot or not uefi firmware loading on netboot is just an edge case of what _should_ be happening15:36
arne_wiebalckI think we have a pretty good idea how we'd like to do it ... xXraphXx: the thing missing is the other patch you've put up?15:36
arne_wiebalckhttps://review.opendev.org/#/c/674879/ ?15:37
patchbotpatch 674879 - ironic-python-agent - image extension, install_bootloader improvements - 10 patch sets15:37
arne_wiebalckor is there still sth else?15:37
xXraphXxthat one, and there would also be missing some other small changes like the following wip: https://review.opendev.org/#/c/675672/ (or any esp relocation mechanism, that would allow the softraid to be bootable with uefi)15:37
patchbotpatch 675672 - ironic-python-agent - Additional command to build efi partitions - 2 patch sets15:37
*** belmoreira has quit IRC15:37
arne_wiebalckok, I see15:38
xXraphXx(and remove the raid 1 restriction constraint on the root device)15:38
arne_wiebalckI don't know how you see it, but I think that is too ambitious for Train (given the time left)15:39
*** trident has joined #openstack-ironic15:40
arne_wiebalckI commit myself to help with testing all this, early on in Ussuri.15:41
* arne_wiebalck is less selfless than this sounds since we will get some UEFI only hardware in the next hardware delivery 15:42
*** rcernin has quit IRC15:44
xXraphXxarne_wiebalck, yep agree, too ambitious -> softraid still works great for BIOS, and we are close for UEFI. On our side, we use it daily (with the small few amends above), combined with deploy templates, to spawn dynamically built raids based on nova flavors. Works great! So thanks to Ironicers for softraid and deploy steps/deploy templates !15:47
arne_wiebalckxXraphXx: wait ... you build the RAID via deploy steps?15:48
*** iurygregory has quit IRC15:52
xXraphXxarne_wiebalck, yep. I'm using a mixed driver between ansible + agent deploy driver. It's "mixed" (understand: sometimes so ugly it could hurt eyes :) ), so I'm having troubles extracting/cleaning it to share upstream15:54
*** jawad_axd has joined #openstack-ironic15:55
*** tzumainn has joined #openstack-ironic15:56
*** TxGirlGeek has quit IRC15:57
rpittaubye all, have a great weekend! o/15:57
arne_wiebalckxXraphXx: that sounds really great (apart from the "hurts eyes" part ;-)! Leveraging deploy steps for s/w was actually one of the next steps I had in mind. Will you be at the summit (to share some details)?15:57
*** rpittau is now known as rpittau|afk15:57
*** TxGirlGeek has joined #openstack-ironic15:58
*** jawad_axd has quit IRC15:59
xXraphXxarne_wiebalck, unfortunately not :/, may be next one. But definitely need to find some time to push it (even as a wip).16:02
mgoddardnight all, have a good weekend16:03
arne_wiebalckxXraphXx: that'd be great ... anyway: once I get round to have a look at deploy steps for s/w RAID, I'll bother you :)16:04
TheJuliaxXraphXx: arne_wiebalck: so, I've been in a meeting for a while... well a couple meetings. What do we feel the consensus is?16:06
xXraphXxarne_wiebalck, sure, happy to share16:06
arne_wiebalckTheJulia: As getting UEFI/gpt work will require multiple patches, including a wip, and due to lack of time for proper testing, we agreed that it is maybe better to not push this into Train.16:09
TheJuliaokay16:09
TheJuliaI just wanted to make sure things had not changed from start to end of the entire discussion16:09
arne_wiebalckTheJulia: xXraphXx is almost there, though, and I am happy to help with testing.16:10
arne_wiebalckTheJulia: So, I think this is not very far from adding support for UEFI/gpt in s/w RAID.16:10
arne_wiebalckTheJulia: And since peo16:11
arne_wiebalckTheJulia: And since people picked up s/w RAID before the release, they will also pick this feature up before Ussuri.16:11
TheJuliapeo?16:11
TheJuliaoh, heh16:11
TheJuliavery very very likely16:11
* arne_wiebalck is working with a new keyboard16:12
arne_wiebalckTheJulia: As a bonus xXraphXx uses s/w RAID with deploy steps!16:12
TheJuliaI picked up on that :)16:12
* TheJulia senses this will be a long weekend of recheck/patch approval16:17
*** pcaruana has quit IRC16:19
arne_wiebalckbye, everyone, have a good weekend!16:20
TheJuliayou too arne_wiebalck16:21
openstackgerritMerged openstack/ironic-tempest-plugin master: Skip the manual cleaning test if software RAID testing is enabled  https://review.opendev.org/68116616:21
*** derekh has quit IRC16:42
*** ociuhandu has quit IRC16:43
rpiosoTheJulia: I'm considering rebasing https://review.opendev.org/68202916:53
patchbotpatch 682029 - ironic - Add `FLOPPY` boot device constant - 1 patch set16:53
rpiosoTheJulia: It keeps failing in CI. Would rebasing prospectively help?16:53
TheJuliarpioso: you can rebase I guess... But CI has a high failure rate until a MTU setting fix meges16:54
TheJuliaMerges16:54
*** igordc has joined #openstack-ironic16:55
TheJuliarpioso:  wait on rebase16:55
openstackgerritraphael.glon proposed openstack/ironic-python-agent master: Revert "Software raid: mbr/gpt partition table alternative"  https://review.opendev.org/68342116:56
rpiosoTheJulia: AFAIU, rebasing wipes out existing approvals. Right?16:56
rpiosoTheJulia: Ack on waiting.16:56
rpiosoetingof: ^^^16:56
rpiosoTheJulia: Please note idrac Redfish vmedia boot depends on that change.16:57
TheJuliarpioso: understood, so that patch is failing on the mtu issue16:58
TheJuliaso once that merges, we can recheck. But we need to wait.16:58
rpiosoTheJulia: Gotcha16:58
rpiosoTheJulia: Thank you!16:58
TheJuliaSep 20 16:38:16 ubuntu-bionic-fortnebula-regionone-0011601584 kernel: qr-3ff709f4-3b: dropped over-mtu packet: 1400 > 135016:58
TheJuliain https://6b681e9843df38a89642-8aefe49572c9ad91eb217e6fb5236bb6.ssl.cf5.rackcdn.com/682029/1/gate/ironic-tempest-ipa-partition-uefi-pxe_ipmitool-tinyipa/063c2e6/controller/logs/syslog.txt.gz16:58
*** ociuhandu has joined #openstack-ironic17:00
*** goldyfruit has quit IRC17:01
*** goldyfruit has joined #openstack-ironic17:01
openstackgerritMerged openstack/ironic-python-agent-builder master: Update tinycore from 8.x to 9.x  https://review.opendev.org/66659117:03
openstackgerritMerged openstack/ironic-python-agent-builder master: Remove dib/build.sh  https://review.opendev.org/68037617:03
*** ociuhandu has quit IRC17:04
TheJuliaugh, looks like it is going to be a couple hours before the ironic queue gets to the job in question17:06
TheJuliastill want an abort button....17:10
*** dougsz has joined #openstack-ironic17:20
*** e0ne has joined #openstack-ironic17:22
*** goldyfruit has quit IRC17:31
*** dougsz has quit IRC17:44
*** whoami-rajat has joined #openstack-ironic17:50
*** goldyfruit has joined #openstack-ironic18:07
*** henriqueof has joined #openstack-ironic18:09
*** dougsz has joined #openstack-ironic18:11
*** goldyfruit_ has joined #openstack-ironic18:14
*** goldyfruit has quit IRC18:17
*** igordc has quit IRC18:20
*** dsneddon has joined #openstack-ironic18:22
*** igordc has joined #openstack-ironic18:26
*** ricolin has quit IRC18:27
*** ricolin has joined #openstack-ironic18:27
*** dsneddon has quit IRC18:27
*** jistr has quit IRC18:28
*** jistr has joined #openstack-ironic18:28
*** dansmith has quit IRC18:29
*** dansmith has joined #openstack-ironic18:29
*** e0ne has quit IRC18:34
*** e0ne has joined #openstack-ironic18:36
*** e0ne has quit IRC18:36
*** dougsz has quit IRC18:36
*** e0ne has joined #openstack-ironic18:37
*** e0ne has quit IRC18:41
*** alexmcleod has quit IRC19:12
*** igordc has quit IRC19:13
*** ricolin has quit IRC19:16
*** igordc has joined #openstack-ironic19:19
*** lucasagomes has quit IRC19:25
*** dsneddon has joined #openstack-ironic19:31
*** ociuhandu has joined #openstack-ironic19:49
*** whoami-rajat has quit IRC20:00
*** ociuhandu has quit IRC20:03
openstackgerritJulia Kreger proposed openstack/ironic master: Fix biforst compatability  https://review.opendev.org/68372220:13
*** jtomasek has quit IRC20:32
openstackgerritJulia Kreger proposed openstack/bifrost master: DNM: testing ci  https://review.opendev.org/68372920:58
*** _mmethot_ has quit IRC21:05
*** mmethot_ has joined #openstack-ironic21:05
*** rloo has quit IRC21:25
*** rloo has joined #openstack-ironic21:26
*** rh-jelabarre has quit IRC21:43
*** rh-jelabarre has joined #openstack-ironic21:44
TheJuliadtantsur|afk: I have htis feeling something with with opesntack sdk has gone sideways with bifrost... specificially in the communication out to ironic-inspector. The conductor is throwing an error now, and yeah. :\22:15
TheJuliahttps://openstack.fortnebula.com:13808/v1/AUTH_e8fd161dc34c421a979a9e6421f823e9/zuul_opendev_logs_1b2/683729/1/check/bifrost-integration-tinyipa-ubuntu-xenial/1b2b68e/logs/ironic/ironic-conductor.log22:15
*** stevebaker has joined #openstack-ironic22:20
*** rh-jelabarre has quit IRC22:22
*** goldyfruit_ has quit IRC22:42
*** manjeets has quit IRC23:10
*** patchbot has quit IRC23:12
*** mgoddard has quit IRC23:17
*** patchbot has joined #openstack-ironic23:17
*** mgoddard has joined #openstack-ironic23:19
openstackgerritMerged openstack/ironic master: Lower MTU override  https://review.opendev.org/68324923:21
*** jawad_axd has joined #openstack-ironic23:24
TheJulia\o/23:27
TheJuliaand there was joy23:27
*** jawad_axd has quit IRC23:28
*** dsneddon has quit IRC23:29
*** sthussey has quit IRC23:47
rpiosoTheJulia: Indeed! I rechecked the change we discussed earlier.23:54

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