| openstackgerrit | Jarrod Johnson proposed a change to stackforge/pyghmi: Detect and notify Console users of disconnect https://review.openstack.org/70259 | 00:00 |
|---|---|---|
| openstackgerrit | A change was merged to stackforge/pyghmi: Detect and notify Console users of disconnect https://review.openstack.org/70259 | 00:06 |
| *** matsuhashi has joined #openstack-ironic | 00:24 | |
| *** epim has joined #openstack-ironic | 00:26 | |
| *** thedodd has quit IRC | 00:29 | |
| *** epim_ has joined #openstack-ironic | 00:30 | |
| *** epim has quit IRC | 00:30 | |
| *** epim_ is now known as epim | 00:30 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Improve handling of invalid input in HashRing class https://review.openstack.org/70267 | 00:31 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: API validates driver name for both POST and PATCH https://review.openstack.org/68018 | 00:31 |
| *** epim_ has joined #openstack-ironic | 00:33 | |
| *** epim has quit IRC | 00:35 | |
| *** epim_ is now known as epim | 00:35 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Minor update for _check_clear_reservation https://review.openstack.org/70273 | 00:43 |
| devananda | ok, and with that, i think i need to call it a day :-D | 00:45 |
| NobodyCam | :) enjoy your night devananda :) | 00:45 |
| * NobodyCam gets his garlic and other vampire warding equment ready | 00:47 | |
| NobodyCam | equipment even :-p | 00:47 |
| *** epim has quit IRC | 00:53 | |
| *** harlowja is now known as harlowja_away | 01:04 | |
| *** epim has joined #openstack-ironic | 01:06 | |
| *** epim has quit IRC | 01:11 | |
| *** harlowja_away is now known as harlowja | 01:26 | |
| *** nosnos has joined #openstack-ironic | 01:41 | |
| *** rloo has quit IRC | 01:54 | |
| *** thedodd has joined #openstack-ironic | 02:18 | |
| *** matsuhashi has quit IRC | 02:34 | |
| *** jbjohnso has quit IRC | 02:35 | |
| *** matsuhas_ has joined #openstack-ironic | 02:37 | |
| *** aignatov_ is now known as aignatov | 02:42 | |
| *** thedodd has quit IRC | 02:44 | |
| *** matsuhashi has joined #openstack-ironic | 03:07 | |
| *** matsuhas_ has quit IRC | 03:08 | |
| *** matsuhashi has quit IRC | 03:08 | |
| *** matsuhashi has joined #openstack-ironic | 03:20 | |
| *** matsuhashi has quit IRC | 03:34 | |
| *** matsuhashi has joined #openstack-ironic | 03:35 | |
| *** thedodd has joined #openstack-ironic | 03:38 | |
| *** matsuhashi has quit IRC | 03:43 | |
| *** matsuhashi has joined #openstack-ironic | 03:50 | |
| *** thedodd has quit IRC | 04:18 | |
| *** ko__ has joined #openstack-ironic | 04:24 | |
| *** harlowja is now known as harlowja_away | 04:26 | |
| *** ko__ has quit IRC | 04:31 | |
| *** blamar has quit IRC | 04:40 | |
| *** blamar has joined #openstack-ironic | 04:45 | |
| *** coolsvap_away has quit IRC | 05:08 | |
| *** aignatov is now known as aignatov_ | 05:09 | |
| *** blamar has quit IRC | 05:13 | |
| *** aignatov_ is now known as aignatov | 05:17 | |
| *** vkozhukalov has joined #openstack-ironic | 05:26 | |
| *** ko__lazy has joined #openstack-ironic | 05:29 | |
| *** pradipta has joined #openstack-ironic | 05:47 | |
| *** rwsu has quit IRC | 05:59 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/68024 | 06:06 |
| *** coolsvap has joined #openstack-ironic | 06:12 | |
| *** vkozhukalov has quit IRC | 06:51 | |
| *** aignatov is now known as aignatov_ | 07:28 | |
| *** mdurnosvistov has joined #openstack-ironic | 08:21 | |
| *** Guest10145 has quit IRC | 08:26 | |
| *** vkozhukalov has joined #openstack-ironic | 08:29 | |
| openstackgerrit | Ghe Rivero proposed a change to openstack/ironic: Rename and update ironic-deploy-helper rootwrap https://review.openstack.org/68340 | 08:30 |
| *** matsuhashi has quit IRC | 08:37 | |
| *** ndipanov has joined #openstack-ironic | 08:40 | |
| *** aignatov_ is now known as aignatov | 08:52 | |
| *** jistr has joined #openstack-ironic | 08:58 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Add inactive timeout to dd https://review.openstack.org/48198 | 09:34 |
| *** athomas has joined #openstack-ironic | 09:35 | |
| *** matsuhashi has joined #openstack-ironic | 09:44 | |
| *** matsuhashi has quit IRC | 10:05 | |
| *** athomas has quit IRC | 10:05 | |
| *** athomas has joined #openstack-ironic | 10:12 | |
| *** martyntaylor has joined #openstack-ironic | 10:16 | |
| *** matsuhashi has joined #openstack-ironic | 10:16 | |
| *** martyntaylor has quit IRC | 10:20 | |
| *** aignatov is now known as aignatov_ | 10:27 | |
| *** tatyana has joined #openstack-ironic | 10:34 | |
| *** martyntaylor has joined #openstack-ironic | 10:38 | |
| *** max_lobur_afk is now known as max_lobur | 10:39 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Fix 'run_as_root' parameter check in execute() https://review.openstack.org/70324 | 10:40 |
| *** aignatov_ is now known as aignatov | 10:56 | |
| openstackgerrit | Mikhail Durnosvistov proposed a change to openstack/ironic: Removes use of timeutils.set_time_override https://review.openstack.org/67432 | 11:07 |
| *** lucasagomes has joined #openstack-ironic | 11:37 | |
| *** coolsvap has quit IRC | 12:24 | |
| *** ko__lazy is now known as ko__away | 12:29 | |
| openstackgerrit | Aleksandr Gordeev proposed a change to openstack/ironic: Fix ssh_port type in _parse_driver_info() from ssh.py https://review.openstack.org/70336 | 12:34 |
| openstackgerrit | Aleksandr Gordeev proposed a change to openstack/ironic: Fix ssh_port type in _parse_driver_info() from ssh.py https://review.openstack.org/70336 | 13:04 |
| agordeev | morning! and thanks for your helpful comments, guys :) | 13:13 |
| openstackgerrit | A change was merged to openstack/ironic: Update docstrings in ssh.py https://review.openstack.org/70033 | 13:22 |
| openstackgerrit | Mikhail Durnosvistov proposed a change to openstack/ironic: Removes use of timeutils.set_time_override https://review.openstack.org/67432 | 13:27 |
| *** romcheg has joined #openstack-ironic | 13:28 | |
| *** jdob has joined #openstack-ironic | 13:38 | |
| *** matsuhashi has quit IRC | 13:41 | |
| *** ko__away is now known as ko__ | 13:41 | |
| *** aignatov is now known as aignatov_ | 13:44 | |
| *** max_lobur is now known as max_lobur_afk | 13:49 | |
| *** matsuhashi has joined #openstack-ironic | 13:50 | |
| *** matsuhashi has quit IRC | 13:57 | |
| *** aignatov_ is now known as aignatov | 13:59 | |
| *** matsuhashi has joined #openstack-ironic | 14:04 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace nonexistent mock assert methods with real ones https://review.openstack.org/70349 | 14:10 |
| *** jbjohnso has joined #openstack-ironic | 14:13 | |
| ko__ | Hi everyone.. | 14:15 |
| ko__ | I am running a bm-vm deployment using ironic and when dd was being executed, i noticed something like /bin/dd if=/var/lib/ironic/images/e66dd98f-9b8e-40e0-a11d-3a9e3d43d666/disk of=/dev/disk/by-path/ip-120.0.0.41:3260-iscsi-iqn-e66dd98f-9b8e-40e0-a11d-3a9e3d43d666-lun-1-part1 bs=1M oflag=direct | 14:16 |
| ko__ | My disk image has some partitions.. but it seems like the disk image is getting dumped to the partition 1 on the bm-vm.. | 14:16 |
| ko__ | or may be my understanding is flawed.. | 14:17 |
| ko__ | can someone please help me.. | 14:17 |
| lucasagomes | ko__, yup, the images are expected to be partition images not disk images (otherwise you wouldn't be able to expand image fs when it's being deploy and things like that) | 14:26 |
| lucasagomes | so the pxe driver is going to format the disk for you and copy the partition image over | 14:26 |
| lucasagomes | that's why pxe needs you to input the pxe_root_gb option | 14:27 |
| *** matty_dubs|gone is now known as matty_dubs | 14:27 | |
| lucasagomes | so it knows which size you want the root partition to be | 14:27 |
| *** matsuhashi has quit IRC | 14:28 | |
| * lucasagomes have a call in 2 minutes so will be away for a while | 14:28 | |
| ko__ | oh.. k.. correct me if i am wrong, in real case secnarios, this pxe_root_gb will be picked up from flavor.. right..? | 14:29 |
| lucasagomes | ko__, afair yea, when you create the flavor in nova you can say things like swap_mb, ephemeral_gb, root_gb | 14:35 |
| *** yongli has quit IRC | 14:36 | |
| *** nosnos has quit IRC | 14:37 | |
| ko__ | lucasagomes: thanks... | 14:44 |
| lucasagomes | ko__, np :) | 14:44 |
| *** yongli has joined #openstack-ironic | 14:52 | |
| ko__ | I have anothe doubt. If my Linux disk Image has two partitions, say 1 for boot another for root, how am i supposed to create the partition image for it..? Or do I have to settle down with root only..? | 15:01 |
| ko__ | my question is can i include both partition in the image? | 15:01 |
| *** russellb is now known as rustlebee | 15:02 | |
| lucasagomes | ko__, the /boot should be part of the root... Right now, the idea is to deploy cloud images | 15:06 |
| lucasagomes | which are prepared images | 15:07 |
| *** yuriyz has quit IRC | 15:07 | |
| lucasagomes | by prepare I mean, these images are already configured to be immediately used by openstack | 15:08 |
| lucasagomes | and they are configured with things like cloud-init | 15:09 |
| lucasagomes | so it can use the metadata service to be configured etc | 15:09 |
| *** max_lobur_afk is now known as max_lobur | 15:09 | |
| ko__ | k.. | 15:09 |
| lucasagomes | if you look at the diskimage-builder tool for e.g | 15:09 |
| lucasagomes | https://github.com/openstack/diskimage-builder | 15:09 |
| *** pradipta has quit IRC | 15:09 | |
| lucasagomes | it's a tool to customize cloud images | 15:10 |
| *** pradipta has joined #openstack-ironic | 15:10 | |
| lucasagomes | it's not going to install the OS from the beggining etc | 15:10 |
| *** rwsu has joined #openstack-ironic | 15:13 | |
| *** lucasagomes is now known as lucas-hungry | 15:26 | |
| ko__ | lucas-hungry: k.. but i see that the diskimage-builder labels the root file system with cloudimg-rootfs when creating disk images... what is this label used for. My understanding was that ironic will use this label to find which partition has the root fs and then dump it to the partition 1 of the baremetal... | 15:27 |
| ko__ | looks like my understanding is flawed.. | 15:28 |
| *** coolsvap has joined #openstack-ironic | 15:38 | |
| *** thedodd has joined #openstack-ironic | 15:48 | |
| NobodyCam | good morning Ironic | 15:50 |
| max_lobur | morning NobodyCam ! and everyone else :) | 15:51 |
| NobodyCam | morning max_lobur | 15:52 |
| mdurnosvistov | Morning all =) | 15:53 |
| NobodyCam | morning mdurnosvistov :) | 15:54 |
| *** doddstack has joined #openstack-ironic | 16:02 | |
| *** vkozhukalov has quit IRC | 16:04 | |
| *** thedodd has quit IRC | 16:05 | |
| NobodyCam | brb | 16:08 |
| *** mdurnosvistov has quit IRC | 16:19 | |
| NobodyCam | lucas-hungry: ping me when your back | 16:20 |
| *** lucas-hungry is now known as lucasagomes | 16:28 | |
| lucasagomes | NobodyCam, ping | 16:28 |
| NobodyCam | hey hey lucasagomes | 16:29 |
| lucasagomes | morning max_lobur NobodyCam | 16:29 |
| NobodyCam | lucasagomes: have a second for https://review.openstack.org/#/c/55549/ | 16:29 |
| lucasagomes | NobodyCam, sure | 16:29 |
| NobodyCam | the comments on patch set 20 are worth reading | 16:30 |
| *** jistr has quit IRC | 16:30 | |
| *** jbjohnso has quit IRC | 16:31 | |
| *** aignatov is now known as aignatov_ | 16:32 | |
| *** jbjohnso has joined #openstack-ironic | 16:32 | |
| * NobodyCam makes another cup of Joe. | 16:38 | |
| devananda | morning, all | 16:45 |
| NobodyCam | Good morning devananda :) | 16:45 |
| lucasagomes | devananda, morning | 16:47 |
| lucasagomes | I'm reviewing 55549 | 16:47 |
| lucasagomes | there's some things | 16:47 |
| NobodyCam | :) | 16:48 |
| lucasagomes | why _check_clear_reservation is static? | 16:48 |
| lucasagomes | I mean because "_<name>" means private | 16:48 |
| lucasagomes | and then doing NodeController._check... looks odd | 16:48 |
| lucasagomes | to access a private method from outside | 16:48 |
| lucasagomes | _check_clear_reservation could just be a private method in that class, I don't see the point in making it static | 16:49 |
| *** ndipanov is now known as ndipanoff | 16:52 | |
| devananda | lucasagomes: you're very right. | 16:53 |
| lucasagomes | devananda, there's no tests on that method | 16:53 |
| NobodyCam | I totaly missed that lastnight | 16:54 |
| lucasagomes | and there's also one else kinda useless there (investigating it) | 16:54 |
| devananda | lucasagomes: i was in a hurry to land all the things yesterday, since the gate was unblocked at last | 16:54 |
| lucasagomes | I will put some comments | 16:54 |
| lucasagomes | devananda, I see, it's my fault as well. I slowed down my reviews in order to test the deploy etc | 16:54 |
| devananda | lucasagomes: i feel like, this patch has been needed for >2 months, we should just land it and iteratively improve it | 16:54 |
| lucasagomes | devananda, indeed | 16:54 |
| devananda | ^ applies to several patches, actually | 16:54 |
| lucasagomes | right | 16:54 |
| lucasagomes | do you think we should just land it and fix it in some following patches? | 16:55 |
| NobodyCam | brb | 16:59 |
| devananda | lucasagomes: or i can fix it and post a new patch set | 17:00 |
| devananda | on top of tha treview | 17:00 |
| lucasagomes | devananda, that sounds better :) | 17:01 |
| lucasagomes | devananda, 1 sec I will add the comments | 17:01 |
| *** martyntaylor has quit IRC | 17:04 | |
| max_lobur | brb | 17:09 |
| lucasagomes | devananda, commented | 17:11 |
| lucasagomes | devananda, so I'm a bit scared because that method doesn't seems to be right tested | 17:11 |
| lucasagomes | if you remove a piece of the code there (one else that I don't understand the logic) all the tests still pass | 17:12 |
| lucasagomes | so it needs some more coverage, and we need to add some tests to verify the logic of that method directly (not only functional tests around the patch()) | 17:12 |
| devananda | lucasagomes: yea, so taht else is bogus. i fixed it in the following patch | 17:12 |
| devananda | lucasagomes: but it def needs better tests | 17:12 |
| devananda | lemme just fix this... | 17:13 |
| lucasagomes | devananda, right, thanks | 17:13 |
| devananda | lucasagomes: do we agree that this is a suitable approach? compare the rpc_node to the patched_node | 17:13 |
| devananda | lucasagomes: max and I talked yesterday about the RFC for PATCH objects, and I feel that what we're doing is within the RFC scope | 17:14 |
| lucasagomes | devananda, hmm maybe it's not needed to test it after the node is patched | 17:14 |
| lucasagomes | we could just take a look at the patch object and see if there's some operation being done to the /reservation path | 17:14 |
| devananda | so that was max_lobur's suggestion on rev19 | 17:15 |
| devananda | and he outlined what looked to me like a good approach | 17:15 |
| lucasagomes | (would be better if we had some helper functions to work with dictonaries, because the patch structure is a list of dictionaries) | 17:15 |
| lucasagomes | lemme take a look | 17:15 |
| devananda | right - max_lobur proposed exactly that | 17:16 |
| devananda | however | 17:16 |
| devananda | in further discussion, i feel like that would actually be very inconsistent | 17:16 |
| devananda | a PATCH document may contain N discrete operations | 17:16 |
| devananda | each is evaluated semantically and sequentially by jsonpatch | 17:16 |
| devananda | then we pass the net change (the resulting Node object) to the conductor to evaluate | 17:16 |
| lucasagomes | hmm makes sense, we could have 2 operations on the same path in the same patch | 17:17 |
| devananda | so a PATCH document could, in principle, contain one or more operations that -- if submitted lone -- would fail, but collectively ,the net result is accepted | 17:17 |
| devananda | if we are applying some rules to each op individually, bu tnot all rules, it's inconsistent | 17:18 |
| lucasagomes | yea devananda indeed I think ur correct here | 17:18 |
| devananda | k k | 17:18 |
| lucasagomes | devananda, just wondering, would make sense to make the break of the reservation out of the patch() | 17:19 |
| lucasagomes | PUT /nodes/<uuid>/break | 17:19 |
| lucasagomes | idk just a hypothesis | 17:19 |
| *** matty_dubs is now known as matty_dubs|lunch | 17:19 | |
| lucasagomes | threat the reservation as an internal attribute, but expose a way to break it, that would remove all this logic of tests patch operations etc | 17:20 |
| devananda | hmm | 17:20 |
| devananda | like we separate out states | 17:20 |
| lucasagomes | yea | 17:21 |
| NobodyCam | that might be more consistent for Nobodycam's thought process | 17:21 |
| lucasagomes | so the break of the node is a complete different action than just setting attributes | 17:21 |
| devananda | :) | 17:21 |
| lucasagomes | there's a whole logic around setting it | 17:21 |
| devananda | yes | 17:21 |
| max_lobur | back | 17:21 |
| NobodyCam | +1 | 17:21 |
| lucasagomes | I mean not setting (because setting is internal) | 17:21 |
| lucasagomes | breaking it | 17:21 |
| devananda | also, i suspect taht'll make a lot more sense if we ever want to require separate privileges for that action | 17:21 |
| lucasagomes | indeed | 17:22 |
| lucasagomes | encapsulating it, in it's own call | 17:22 |
| lucasagomes | would make the code more sane | 17:22 |
| lucasagomes | to add more logic around it | 17:22 |
| lucasagomes | instead of squashing everything into patch() | 17:22 |
| devananda | lucasagomes: so that's basically a whole different code change to do | 17:23 |
| lucasagomes | (but that would need a complete refactor of that patch, ofc the functions there could be reused) | 17:23 |
| devananda | not just a little fix of the logic in the "else" line | 17:24 |
| devananda | ye | 17:24 |
| devananda | yea | 17:24 |
| lucasagomes | yea | 17:24 |
| max_lobur | I like the subresource approach too | 17:24 |
| * devananda wonders if this patch's long history might be in some part a result of the author not being in channel much | 17:25 | |
| * lucasagomes feels bad cause he took a long time to review that patch | 17:26 | |
| * max_lobur makes tea :-P | 17:27 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/70382 | 17:31 |
| *** mdurnosvistov has joined #openstack-ironic | 17:33 | |
| *** hstimer has joined #openstack-ironic | 17:53 | |
| *** hstimer has quit IRC | 17:55 | |
| *** marty- has joined #openstack-ironic | 17:55 | |
| *** harlowja_away is now known as harlowja | 17:55 | |
| *** hstimer has joined #openstack-ironic | 17:56 | |
| *** marty- has quit IRC | 17:56 | |
| *** matty_dubs|lunch is now known as matty_dubs | 18:13 | |
| *** doddstack has quit IRC | 18:20 | |
| *** athomas has quit IRC | 18:20 | |
| * NobodyCam makes more coffee | 18:21 | |
| lucasagomes | devananda, NobodyCam, re nova driver | 18:22 |
| lucasagomes | I added a retry mechanism, to retry the request if the API returns 409 | 18:22 |
| lucasagomes | (Conflict) | 18:22 |
| NobodyCam | :) pushed up? | 18:22 |
| lucasagomes | would make sense to put the number of attempts to retry and the delay between the attemps in the config options? | 18:23 |
| lucasagomes | NobodyCam, just finishing one more thing, related to ^ | 18:23 |
| lucasagomes | NobodyCam, we need another attempt/wait appraoch | 18:23 |
| lucasagomes | when destroying the instance | 18:23 |
| NobodyCam | ahh :) | 18:24 |
| lucasagomes | because we call the node tear_down first and then we have to wait to it to finish in order to start removing the pxe_ attributes and instance uuid from the node | 18:24 |
| lucasagomes | API doesn't allow you to remove the instance_uuid if the machine is active | 18:24 |
| lucasagomes | and that makes sense | 18:24 |
| *** aignatov_ is now known as aignatov | 18:24 | |
| devananda | lucasagomes: right | 18:34 |
| devananda | lucasagomes: so yes, config seems fine. I would have it be a time (in seconds), then use LoopingCall to poll | 18:34 |
| * lucasagomes gotta take a look at LoopingCall | 18:35 | |
| lucasagomes | thanks | 18:35 |
| devananda | eg, ironic.api_retry_interval = 1 ; ironic.api_max_retires = 5 | 18:36 |
| devananda | or something | 18:36 |
| devananda | lucasagomes: look at the baremetal code's IPMI driver for some examples of loopingcall | 18:36 |
| lucasagomes | devananda, cheers, yea baremetal/PXE seems to be using it | 18:37 |
| lucasagomes | will take a look | 18:37 |
| NobodyCam | morning walkies..brb.. | 18:37 |
| devananda | NobodyCam: think you could take a minute or two to update https://wiki.openstack.org/wiki/Ironic#Using_Disk_Image_Builder with the current status? | 18:43 |
| NobodyCam | is on my list for today :) | 18:44 |
| NobodyCam | ya we had a bunch of stuff land | 18:44 |
| NobodyCam | and I need to pull the reff to my repo | 18:44 |
| devananda | NobodyCam: awesome, ty | 18:45 |
| NobodyCam | :) | 18:45 |
| devananda | davidlenwell: hi! so - yes, pecan/wsme. using the sphinx extension for our api doc generation. | 18:46 |
| davidlenwell | devananda: awesome .. I'lm gonna look at the bug list and try to land a patch or two over the next week while I get familure with the project. | 18:46 |
| *** mdurnosvistov has quit IRC | 18:46 | |
| davidlenwell | I'll pose stupid questions here ;) | 18:47 |
| devananda | great :) | 18:47 |
| devananda | davidlenwell: take a look at the open reviews too. that's a good way to get familiar with what's changing // needs to be worked on | 18:48 |
| davidlenwell | will do | 18:48 |
| devananda | davidlenwell: lucasagomes is a good resource to direct API questions to, if you think you might be leaning towards taht side of the code base | 18:48 |
| lucasagomes | davidlenwell, welcome :) | 18:49 |
| davidlenwell | yeah .. I think that is the lower hanging fruit as far as what I can contribute to right out of the gate | 18:50 |
| davidlenwell | hi lucasagomes | 18:50 |
| davidlenwell | It will be next week before I really dig in.. | 18:50 |
| *** aignatov is now known as aignatov_ | 18:50 | |
| *** ko__ has quit IRC | 18:51 | |
| *** med_ has joined #openstack-ironic | 18:55 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace CONF.set_default with self.config https://review.openstack.org/70397 | 18:56 |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace CONF.set_default with self.config https://review.openstack.org/70397 | 18:58 |
| *** aignatov_ is now known as aignatov | 19:01 | |
| lucasagomes | NobodyCam, testing before pushing | 19:03 |
| *** tatyana has quit IRC | 19:05 | |
| *** thedodd has joined #openstack-ironic | 19:06 | |
| NobodyCam | morninf davidlenwell :) | 19:07 |
| NobodyCam | lucasagomes: w00 hoo | 19:08 |
| NobodyCam | ** morning even | 19:08 |
| lucasagomes | | 08ee9454-7775-4bb4-b609-746d4e7ff192 | ironictest | ACTIVE | None | Shutdown | public=172.24.4.3 | | 19:10 |
| lucasagomes | w00t! | 19:10 |
| lucasagomes | the power state takes a lot of time to be sync'd with ironic | 19:11 |
| * lucasagomes gotta take a look at it later | 19:11 | |
| NobodyCam | :) | 19:11 |
| davidlenwell | hi NobodyCam | 19:11 |
| NobodyCam | :) | 19:12 |
| NobodyCam | nice to see you back in channel :) | 19:12 |
| lucasagomes | NobodyCam, new patchset pushed and tested :) | 19:13 |
| lucasagomes | spawn() more reliable because it now retries requests | 19:13 |
| lucasagomes | destroy() working | 19:13 |
| * NobodyCam looks | 19:14 | |
| lucasagomes | some performance improvements when updating resources | 19:14 |
| lucasagomes | you can make a patch with a list of patches | 19:14 |
| lucasagomes | instead of calling update to update each attribute | 19:14 |
| NobodyCam | hehehe | 19:14 |
| lucasagomes | :) | 19:15 |
| lucasagomes | and exception handlers fixed | 19:15 |
| lucasagomes | (maybe not all of them) | 19:15 |
| NobodyCam | like the _wait_for_provision_state :) | 19:16 |
| lucasagomes | :) | 19:16 |
| lucasagomes | retry_on_conflict() use a similar approach | 19:17 |
| NobodyCam | ya just reading :) | 19:17 |
| NobodyCam | nice :) | 19:17 |
| lucasagomes | forgot some small nits | 19:18 |
| lucasagomes | but it's alright | 19:18 |
| NobodyCam | :) | 19:19 |
| NobodyCam | gah another office lunch | 19:19 |
| NobodyCam | :-p | 19:19 |
| NobodyCam | but the office is 1200 minles away | 19:19 |
| *** vkozhukalov has joined #openstack-ironic | 19:19 | |
| *** ndipanoff is now known as ndipan_off | 19:20 | |
| lucasagomes | :p | 19:20 |
| lucasagomes | NobodyCam, where r u located now? | 19:21 |
| lucasagomes | still in california? | 19:21 |
| NobodyCam | yep :) | 19:22 |
| *** pradipta has quit IRC | 19:22 | |
| lucasagomes | ^^ nice | 19:26 |
| lucasagomes | I'm getting ready to the cold again | 19:26 |
| lucasagomes | returning to Ireland tomorrow | 19:26 |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Remove unnecessary json dumps/loads from tests https://review.openstack.org/70413 | 19:27 |
| * lucasagomes don't like long flights x.x | 19:27 | |
| *** aignatov is now known as aignatov_ | 19:28 | |
| NobodyCam | oh man but hten your comming out forhte sprint? | 19:29 |
| max_lobur | have a good weekend guys! | 19:29 |
| max_lobur | btw I will be out of office on Monday | 19:30 |
| NobodyCam | you too max_lobur :) | 19:30 |
| max_lobur | but I'll try to join the meeting | 19:30 |
| NobodyCam | teh meetings are logged:) | 19:30 |
| NobodyCam | oh cool-e-o | 19:30 |
| *** aignatov_ is now known as aignatov | 19:31 | |
| max_lobur | :) | 19:31 |
| max_lobur | anyway it's good when all attend | 19:31 |
| max_lobur | I will be able to ask question in proper moment :-P | 19:31 |
| NobodyCam | yes :) | 19:31 |
| NobodyCam | habe a good weekend | 19:31 |
| max_lobur | thx! | 19:32 |
| max_lobur | bye All! | 19:32 |
| lucasagomes | max_lobur, g'night! | 19:32 |
| *** max_lobur is now known as max_lobur_afk | 19:32 | |
| lucasagomes | NobodyCam, for the spring?! | 19:33 |
| NobodyCam | sprint :-p | 19:33 |
| lucasagomes | sprint* | 19:33 |
| lucasagomes | yea | 19:33 |
| lucasagomes | heh | 19:33 |
| NobodyCam | :-p | 19:33 |
| lucasagomes | dunno what you mean by that :P | 19:34 |
| NobodyCam | the ironic /OoO sprint in march | 19:34 |
| lucasagomes | ahh | 19:34 |
| lucasagomes | NobodyCam, I dunno, I didn't get a answer from my manager | 19:35 |
| lucasagomes | I hope I will be able to make it | 19:35 |
| * lucasagomes really wants to go | 19:35 | |
| * NobodyCam hopes so too | 19:35 | |
| *** ndipan_off has quit IRC | 19:37 | |
| lucasagomes | my gmail is horrible slow! Seems there's some problems with firefox and some hangout updates | 19:38 |
| lucasagomes | Script: https://talkgadget.google.com/_/scs/talk-static/_/js/k=wcs.wbl.en.QX9o2g257d8.O/m=b,r/am=CpTALiAuPAGMoQEBDg/rt=j/d=1/rs=AItRSTPIJS-Jnb5oiufGxbTJ2mWZuBTaqQ:806 | 19:38 |
| lucasagomes | >.< | 19:38 |
| NobodyCam | nice | 19:39 |
| devananda | dkehn: for the neutron update, should the api request contain the port_id in both the URI and body? | 19:45 |
| devananda | that seems really redundant | 19:45 |
| dkehn | I believe it should | 19:46 |
| devananda | dkehn: http://docs.openstack.org/api/openstack-network/2.0/content/PUT_os-ports-v2.0_updatePort_v2.0_ports__port_id__ports.html suggests it should not | 19:46 |
| devananda | the body does not contain "id" field | 19:47 |
| devananda | because that is in the URI | 19:47 |
| dkehn | devananda: yes your correct, http://docs.openstack.org/api/openstack-network/2.0/content/extra-dhc-opt-ext-update.html | 19:49 |
| devananda | just found that page too. great, thanks! | 19:50 |
| *** mdurnosvistov has joined #openstack-ironic | 19:59 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Adds Neutron support to Ironic https://review.openstack.org/66071 | 19:59 |
| devananda | dkehn: ^ | 19:59 |
| devananda | woops, flake8 ... one sec | 20:00 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Adds Neutron support to Ironic https://review.openstack.org/66071 | 20:01 |
| * NobodyCam takes lunch and steps afk for a bit. I will be back! | 20:11 | |
| lucasagomes | right I'm done for today | 20:17 |
| lucasagomes | devananda, NobodyCam dkehn have a good night | 20:17 |
| lucasagomes | enjoy the weekend | 20:17 |
| devananda | g'night lucasagomes ! | 20:17 |
| *** lucasagomes has quit IRC | 20:18 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Adds Neutron support to Ironic https://review.openstack.org/66071 | 20:19 |
| openstackgerrit | Jarrod Johnson proposed a change to stackforge/pyghmi: Fix exception if login times out https://review.openstack.org/70434 | 20:22 |
| *** romcheg has quit IRC | 20:25 | |
| openstackgerrit | A change was merged to stackforge/pyghmi: Fix exception if login times out https://review.openstack.org/70434 | 20:25 |
| openstackgerrit | Jarrod Johnson proposed a change to stackforge/pyghmi: Allow special handling to preceed generic handler https://review.openstack.org/70437 | 20:31 |
| *** vkozhukalov has quit IRC | 20:34 | |
| openstackgerrit | A change was merged to stackforge/pyghmi: Allow special handling to preceed generic handler https://review.openstack.org/70437 | 20:36 |
| *** matty_dubs is now known as matty_dubs|gone | 20:47 | |
| * devananda makes food | 20:51 | |
| *** aignatov is now known as aignatov_ | 21:06 | |
| *** rloo has joined #openstack-ironic | 21:36 | |
| devananda | back | 21:38 |
| *** jdob has quit IRC | 21:40 | |
| * NobodyCam is back | 21:46 | |
| openstackgerrit | Jarrod Johnson proposed a change to stackforge/pyghmi: Reasonably relay IPMI errors to console handlers https://review.openstack.org/70448 | 21:48 |
| openstackgerrit | A change was merged to stackforge/pyghmi: Reasonably relay IPMI errors to console handlers https://review.openstack.org/70448 | 21:51 |
| *** openstackgerrit has quit IRC | 21:53 | |
| *** openstackgerrit has joined #openstack-ironic | 21:53 | |
| NobodyCam | devananda: question on lines 614-614 of https://review.openstack.org/#/c/66071/18/etc/ironic/ironic.conf.sample | 21:54 |
| NobodyCam | where is that comming from? | 21:54 |
| NobodyCam | and wont that conflict with https://review.openstack.org/#/c/70141/ | 21:55 |
| NobodyCam | s/614-614/614-615 | 21:59 |
| * devananda looks | 21:59 | |
| devananda | oh | 22:00 |
| devananda | no - it's fine | 22:00 |
| devananda | i'll fix this patch up to rebase on ghe's | 22:00 |
| NobodyCam | :) | 22:00 |
| NobodyCam | the #noqa on the doc string url ref is because it > 80 chr? | 22:05 |
| devananda | ya | 22:05 |
| NobodyCam | :) | 22:06 |
| NobodyCam | cheeter | 22:06 |
| NobodyCam | :-p | 22:06 |
| devananda | breakign the link onto 2 lines makes it non-clickable in the web docs | 22:06 |
| NobodyCam | ya | 22:06 |
| NobodyCam | just giving you a hard time , as its friday and i'm ready to +2 the patch | 22:07 |
| NobodyCam | and I think I'll steal the neutron client test and adapt for use in nova-driver | 22:08 |
| devananda | :) | 22:17 |
| devananda | i like the NeutronAPI class code, though i feel like i'm +2'ing my own work now ... | 22:18 |
| openstackgerrit | A change was merged to openstack/ironic: Fix ssh_port type in _parse_driver_info() from ssh.py https://review.openstack.org/70336 | 22:24 |
| openstackgerrit | A change was merged to openstack/ironic: Remove net_config_template options https://review.openstack.org/70141 | 22:28 |
| NobodyCam | devananda: ^^^ landed | 22:28 |
| devananda | sweet. rebasing | 22:28 |
| *** jbjohnso has quit IRC | 22:28 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Adds Neutron support to Ironic https://review.openstack.org/66071 | 22:30 |
| devananda | done | 22:30 |
| NobodyCam | just a dumb question on my part: | 22:33 |
| NobodyCam | why the acl.register_opts(CONF) | 22:33 |
| devananda | i wondered taht too | 22:34 |
| devananda | i beleive it's to get access to the keystone opts | 22:34 |
| NobodyCam | ahh | 22:34 |
| devananda | found the same thing (acl.register_opts(CONF) in our API service start-up code | 22:34 |
| NobodyCam | :) | 22:34 |
| NobodyCam | dkehn: around? | 22:37 |
| dkehn | yep | 22:37 |
| NobodyCam | wanta review your patch :) | 22:38 |
| NobodyCam | https://review.openstack.org/#/c/66071/ | 22:38 |
| dkehn | sure, thats a 1st | 22:38 |
| devananda | hehehe | 22:38 |
| NobodyCam | hehehehe | 22:38 |
| * devananda adds node-takeover hooks to pxe.py | 22:38 | |
| NobodyCam | nice :) | 22:38 |
| NobodyCam | http://www.youtube.com/watch?v=7GL6LH6ufhM | 22:39 |
| dkehn | NobodyCam: k | 22:40 |
| NobodyCam | devananda: +a'ing! | 22:40 |
| devananda | woot | 22:41 |
| dkehn | NobodyCam: this is what I sent to markmcclain http://www.youtube.com/watch?v=U0oOH2H56uc, who live in Atlanta | 22:41 |
| NobodyCam | ofc we that all befor mr J was done | 22:41 |
| dkehn | hedging bets | 22:42 |
| NobodyCam | brb | 22:42 |
| devananda | woops, switched the BP status too soon | 22:43 |
| devananda | i need to add one more patch | 22:43 |
| devananda | maybe 2 | 22:43 |
| NobodyCam | :) | 22:44 |
| devananda | dkehn: the neutron port_id -- this is the neutron port, not the MAC address, right? | 22:45 |
| devananda | *neutron port ID | 22:45 |
| devananda | NobodyCam: we need to know that value. it's not something we have today | 22:45 |
| dkehn | devananda: yes neutron port_id | 22:45 |
| devananda | dkehn: any thoughts on how we get that? | 22:46 |
| devananda | dkehn: ie, can I ask neutron "what is the port_id for this MAC" | 22:46 |
| dkehn | devananda: I was under the impression that your had that | 22:47 |
| devananda | dkehn: no inside Ironic | 22:47 |
| dkehn | ok | 22:47 |
| NobodyCam | devananda: lines 604 - 611 of nova-driver | 22:47 |
| devananda | NobodyCam: the noav drive is passing it up ... | 22:47 |
| NobodyCam | attaches to nodes port | 22:47 |
| devananda | ahh | 22:48 |
| NobodyCam | :) | 22:48 |
| devananda | NobodyCam: so we need that to pass the info DOWN to ironic, too | 22:48 |
| devananda | nvm! it is! | 22:49 |
| NobodyCam | :) | 22:49 |
| *** romcheg has joined #openstack-ironic | 22:49 | |
| NobodyCam | lol the patch is in both chack and gate queues on zuul at the same time | 22:51 |
| * NobodyCam wonuders if -infra ever thought of that edge case | 22:52 | |
| devananda | yea | 22:53 |
| devananda | it's fine, heh | 22:53 |
| * devananda prepares to hide incase Jenkins -1's it | 22:54 | |
| NobodyCam | lol passes all pep 8 and pys in tempest tests now | 22:54 |
| devananda | NobodyCam: if there are no VIFs registered on the ports belonging to the node | 22:56 |
| NobodyCam | check passed :) now just gate left :-p | 22:56 |
| devananda | NobodyCam: should PXE driver silently fail the neutron update, or raise? any thoughts? | 22:56 |
| NobodyCam | huh | 22:56 |
| NobodyCam | somehting happened | 22:56 |
| NobodyCam | oh wrong review | 22:57 |
| NobodyCam | :-p doh | 22:57 |
| NobodyCam | oh, | 22:57 |
| NobodyCam | log.error and continue I think is out best bet | 22:58 |
| NobodyCam | they could be using a single hub and not neutron | 22:59 |
| NobodyCam | waiting on tempest tests for gate... still queued | 23:07 |
| devananda | hmmm | 23:11 |
| devananda | call me paranoid but i've another thought | 23:11 |
| devananda | dkehn: if a node has N ports, Ironic will try to set the DHCP BOOT opts on all of them. what do you think is the likelyhood of 0 < x < N of those updates failing? | 23:12 |
| NobodyCam | ??? gulp ?? | 23:12 |
| devananda | NobodyCam: and if some but not all fail, is that exception-worthy? or just LOG.warning? | 23:12 |
| devananda | i think if at least 1 succeeds, we should continue | 23:13 |
| NobodyCam | warrning I would ... ya | 23:13 |
| dkehn | hmm, port_id specifies 1 port | 23:13 |
| NobodyCam | I think that is why we set all port opts | 23:13 |
| NobodyCam | we dont know which one will actually bee the boot interface | 23:14 |
| devananda | dkehn: yes, but a node may have >1 port | 23:14 |
| dkehn | true, I've gotta step out for a bit | 23:14 |
| devananda | dkehn: so the PXE driver will be calling NeutronAPI.update_port many times, potentially | 23:14 |
| devananda | NobodyCam: right | 23:15 |
| NobodyCam | n-driver will assign a unique id to each port | 23:15 |
| devananda | ye | 23:15 |
| devananda | yep | 23:15 |
| devananda | and ironic, during takeover,w ill attempt to update all of the ports | 23:15 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 23:21 |
| devananda | NobodyCam: posting ^ before i've written unit tests so you can see what i'm talking about | 23:21 |
| * NobodyCam clicks | 23:21 | |
| NobodyCam | can we include the port id's that failed in the log | 23:23 |
| NobodyCam | we have node, but which port | 23:24 |
| NobodyCam | as a admin I may look at the log as say of that a internal channel and not care | 23:24 |
| devananda | ah, yes | 23:24 |
| NobodyCam | not that as a admin I would know the my nodes port uuid by hart... but I might | 23:25 |
| NobodyCam | :-p | 23:25 |
| NobodyCam | yes I aggree with the logic | 23:25 |
| NobodyCam | oh tempest started on THE PATCH | 23:26 |
| NobodyCam | lol | 23:26 |
| NobodyCam | really: pip's wheel support requires setuptools >= 0.8 for dist-info support. | 23:27 |
| NobodyCam | does ubuntu even have wheel support | 23:27 |
| *** romcheg has quit IRC | 23:28 | |
| dkehn | b | 23:31 |
| NobodyCam | c? | 23:31 |
| *** thedodd has quit IRC | 23:37 | |
| *** mdurnosvistov has quit IRC | 23:38 | |
| dkehn | its friday | 23:48 |
| dkehn | so b = back | 23:48 |
| dkehn | NobodyCam: ^^^^ | 23:48 |
| NobodyCam | lol | 23:49 |
| dkehn | NobodyCam: hows the weather out there still warm? | 23:51 |
| *** jbjohnso has joined #openstack-ironic | 23:51 | |
| NobodyCam | its a bit chilly its like 68 today | 23:52 |
| NobodyCam | :-p | 23:52 |
| * NobodyCam ducks | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!