| leitan | TheJulia: can be security groups an issue ? | 00:02 |
|---|---|---|
| *** baoli has quit IRC | 00:03 | |
| *** rama_y has quit IRC | 00:06 | |
| *** penick has joined #openstack-ironic | 00:07 | |
| *** etingof has quit IRC | 00:13 | |
| *** links has joined #openstack-ironic | 00:20 | |
| *** charcol has joined #openstack-ironic | 00:43 | |
| *** penick has quit IRC | 00:43 | |
| *** etingof has joined #openstack-ironic | 00:51 | |
| *** zhangfei has joined #openstack-ironic | 00:57 | |
| *** baoli has joined #openstack-ironic | 01:02 | |
| *** baoli has quit IRC | 01:06 | |
| *** phuongnh has joined #openstack-ironic | 01:13 | |
| *** fragatina has quit IRC | 01:14 | |
| openstackgerrit | Ruby Loo proposed openstack/python-ironic-inspector-client master: TEST. Please. Ignore. https://review.openstack.org/516805 | 01:21 |
| *** leitan has quit IRC | 01:24 | |
| *** mjturek has quit IRC | 01:24 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 01:26 | |
| *** AlexeyAbashkin has quit IRC | 01:31 | |
| *** baoli has joined #openstack-ironic | 01:44 | |
| zhangfei | morning | 01:45 |
| zhangfei | Could you help review diskimage-builder on aarch64 | 01:45 |
| zhangfei | https://review.openstack.org/512174 | 01:45 |
| zhangfei | https://review.openstack.org/512175 | 01:45 |
| patchbot | patch 512174 - diskimage-builder - Fix grub2 dependency on arm64 | 01:45 |
| patchbot | patch 512175 - diskimage-builder - Add debian minimal requirement for arm64 | 01:45 |
| *** penick has joined #openstack-ironic | 01:49 | |
| *** namnh has joined #openstack-ironic | 01:53 | |
| *** hw_wutianwei has joined #openstack-ironic | 01:58 | |
| *** hshiina has joined #openstack-ironic | 02:00 | |
| *** baoli has quit IRC | 02:06 | |
| *** zshi has joined #openstack-ironic | 02:14 | |
| *** zhangfei has quit IRC | 02:18 | |
| *** zhangfei has joined #openstack-ironic | 02:19 | |
| *** zhangfei has quit IRC | 02:20 | |
| *** zhangfei has joined #openstack-ironic | 02:20 | |
| openstackgerrit | Merged openstack/ironic master: Update ironic.sample.conf https://review.openstack.org/516245 | 02:24 |
| openstackgerrit | Merged openstack/ironic master: Small fixes in the common reference architecture docs https://review.openstack.org/516376 | 02:24 |
| *** penick has quit IRC | 02:26 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 02:26 | |
| *** AlexeyAbashkin has quit IRC | 02:30 | |
| *** yangyapeng has quit IRC | 02:42 | |
| *** yangyapeng has joined #openstack-ironic | 02:47 | |
| *** ijw has quit IRC | 02:50 | |
| *** ijw has joined #openstack-ironic | 02:50 | |
| *** rbudden has quit IRC | 02:50 | |
| *** ijw has quit IRC | 02:51 | |
| *** etingof has quit IRC | 02:57 | |
| *** hshiina has quit IRC | 02:58 | |
| *** dtantsur|afk has quit IRC | 03:01 | |
| *** etingof has joined #openstack-ironic | 03:10 | |
| *** rama_y has joined #openstack-ironic | 03:22 | |
| *** ijw has joined #openstack-ironic | 03:24 | |
| *** fragatina has joined #openstack-ironic | 03:26 | |
| openstackgerrit | Nguyen Van Trung proposed openstack/ironic master: Add additional capabilities discovery for iRMC driver https://review.openstack.org/417347 | 03:26 |
| *** fragatina has quit IRC | 03:28 | |
| *** fragatina has joined #openstack-ironic | 03:28 | |
| *** etingof has quit IRC | 03:31 | |
| *** fragatina has quit IRC | 03:33 | |
| *** etingof has joined #openstack-ironic | 03:43 | |
| *** etingof has quit IRC | 03:48 | |
| *** zhangfei has quit IRC | 03:54 | |
| *** etingof has joined #openstack-ironic | 04:01 | |
| *** gcb_ has quit IRC | 04:04 | |
| *** gcb_ has joined #openstack-ironic | 04:05 | |
| *** hshiina has joined #openstack-ironic | 04:12 | |
| *** zhangfei has joined #openstack-ironic | 04:14 | |
| openstackgerrit | Nguyen Van Trung proposed openstack/ironic master: Add additional capabilities discovery for iRMC driver https://review.openstack.org/417347 | 04:14 |
| *** ijw has quit IRC | 04:25 | |
| *** sasha2 has quit IRC | 04:30 | |
| *** zhangjl has quit IRC | 04:30 | |
| *** etingof has quit IRC | 04:30 | |
| *** zhangjl has joined #openstack-ironic | 04:31 | |
| *** udesale has joined #openstack-ironic | 04:39 | |
| *** etingof has joined #openstack-ironic | 04:44 | |
| *** crushil has joined #openstack-ironic | 05:10 | |
| *** ijw has joined #openstack-ironic | 05:14 | |
| *** ijw has quit IRC | 05:16 | |
| *** ijw has joined #openstack-ironic | 05:22 | |
| *** ijw has quit IRC | 05:27 | |
| *** ijw has joined #openstack-ironic | 05:36 | |
| *** marios has joined #openstack-ironic | 05:58 | |
| *** ijw has quit IRC | 06:04 | |
| *** udesale__ has joined #openstack-ironic | 06:12 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector master: Imported Translations from Zanata https://review.openstack.org/516528 | 06:13 |
| *** udesale has quit IRC | 06:15 | |
| *** udesale__ has quit IRC | 06:18 | |
| *** udesale__ has joined #openstack-ironic | 06:19 | |
| *** udesale has joined #openstack-ironic | 06:21 | |
| *** udesale__ has quit IRC | 06:23 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic master: Follow up Secure boot support for irmc-virtual-media driver https://review.openstack.org/516861 | 06:44 |
| *** d0ugal_ has quit IRC | 06:45 | |
| *** links has quit IRC | 06:53 | |
| *** d0ugal_ has joined #openstack-ironic | 06:55 | |
| *** jtomasek has joined #openstack-ironic | 06:56 | |
| *** mjura has joined #openstack-ironic | 07:08 | |
| *** mjura has quit IRC | 07:09 | |
| *** mjura has joined #openstack-ironic | 07:11 | |
| *** spectr-RH has joined #openstack-ironic | 07:12 | |
| *** spectr has quit IRC | 07:13 | |
| *** links has joined #openstack-ironic | 07:20 | |
| *** Dinesh_Bhor has joined #openstack-ironic | 07:23 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic master: Follow up Secure boot support for irmc-virtual-media driver https://review.openstack.org/516861 | 07:25 |
| *** charcol has quit IRC | 07:25 | |
| *** zhangjl has quit IRC | 07:29 | |
| *** zhangjl has joined #openstack-ironic | 07:30 | |
| *** hoonetorg has quit IRC | 07:39 | |
| *** rcernin has quit IRC | 07:42 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic master: Follow up Secure boot support for irmc-virtual-media driver https://review.openstack.org/516861 | 07:42 |
| *** tesseract has joined #openstack-ironic | 07:48 | |
| *** hoonetorg has joined #openstack-ironic | 07:55 | |
| pas-ha | morning ironic :) | 07:57 |
| *** AlexeyAbashkin has joined #openstack-ironic | 08:07 | |
| *** rbartal has joined #openstack-ironic | 08:11 | |
| *** rbartal__ has joined #openstack-ironic | 08:11 | |
| *** rbartal_ has joined #openstack-ironic | 08:12 | |
| *** e0ne has joined #openstack-ironic | 08:17 | |
| *** pcaruana has joined #openstack-ironic | 08:17 | |
| *** Goneri has joined #openstack-ironic | 08:18 | |
| *** dtantsur has joined #openstack-ironic | 08:23 | |
| dtantsur | morning ironic | 08:23 |
| dtantsur | lost part of scrollback, please re-ping if needed :) | 08:23 |
| *** spectr-RH has quit IRC | 08:32 | |
| *** spectr has joined #openstack-ironic | 08:33 | |
| *** zshi has quit IRC | 08:39 | |
| *** zshi has joined #openstack-ironic | 08:39 | |
| *** d0ugal_ has quit IRC | 08:42 | |
| *** rcernin has joined #openstack-ironic | 08:42 | |
| *** d0ugal has joined #openstack-ironic | 08:44 | |
| *** d0ugal has quit IRC | 08:44 | |
| *** d0ugal has joined #openstack-ironic | 08:44 | |
| *** amarao has joined #openstack-ironic | 08:47 | |
| *** racedo_ has joined #openstack-ironic | 08:50 | |
| *** amarao has quit IRC | 08:59 | |
| *** ccamacho has quit IRC | 09:00 | |
| openstackgerrit | Nguyen Van Trung proposed openstack/ironic master: Add additional capabilities discovery for iRMC driver https://review.openstack.org/417347 | 09:02 |
| *** priteau has joined #openstack-ironic | 09:03 | |
| *** jistr_ is now known as jistr | 09:07 | |
| *** mgoddard_ has joined #openstack-ironic | 09:08 | |
| *** serlex has joined #openstack-ironic | 09:12 | |
| *** derekh has joined #openstack-ironic | 09:18 | |
| gcb_ | dtantsur: wondering if ironic need method from ironic-lib like https://github.com/openstack/ironic-lib/blob/master/ironic_lib/utils.py#L62 | 09:20 |
| gcb_ | https://github.com/openstack/ironic/blob/master/ironic/common/utils.py#L54 | 09:20 |
| *** ralonsoh has joined #openstack-ironic | 09:20 | |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/ironic master: Fix outdated docstring for _get_root_helper() https://review.openstack.org/516897 | 09:24 |
| dtantsur | gcb_: I think ironic should use one from ironic-lib, if that's your question | 09:26 |
| *** sulliwan has joined #openstack-ironic | 09:27 | |
| gcb_ | dtantsur: yes, thanks | 09:27 |
| gcb_ | will try to post a patch for that | 09:28 |
| *** zhangfei has quit IRC | 09:29 | |
| *** lucas-afk is now known as lucasagomes | 09:31 | |
| *** milan has joined #openstack-ironic | 09:32 | |
| milan | morning Ironic! :) | 09:32 |
| *** lifeless has quit IRC | 09:32 | |
| *** hw_wutianwei has quit IRC | 09:36 | |
| *** etingof has quit IRC | 09:37 | |
| *** ccamacho has joined #openstack-ironic | 09:37 | |
| *** rbartal_ has quit IRC | 09:37 | |
| *** rbartal_ has joined #openstack-ironic | 09:38 | |
| *** ralonsoh has joined #openstack-ironic | 09:39 | |
| *** zhangfei has joined #openstack-ironic | 09:43 | |
| *** rama_y has quit IRC | 09:43 | |
| *** etingof has joined #openstack-ironic | 09:55 | |
| *** jkilpatr has joined #openstack-ironic | 09:56 | |
| *** mat128 has joined #openstack-ironic | 09:56 | |
| *** links has quit IRC | 09:57 | |
| *** jkilpatr has quit IRC | 10:11 | |
| openstackgerrit | Merged openstack/ironic-inspector master: Imported Translations from Zanata https://review.openstack.org/516528 | 10:12 |
| *** namnh has quit IRC | 10:12 | |
| dtantsur | hey milan | 10:20 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic master: Remove python-oneviewclient from Ironic OneView drivers https://review.openstack.org/480284 | 10:22 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic master: Remove python-oneviewclient from Ironic OneView drivers https://review.openstack.org/480284 | 10:24 |
| *** parlos has joined #openstack-ironic | 10:34 | |
| *** Goneri has quit IRC | 10:38 | |
| *** udesale has quit IRC | 10:50 | |
| milan | morning dtantsur! :) | 10:51 |
| *** Goneri has joined #openstack-ironic | 10:53 | |
| milan | aarefiev, morning, mind "finishing" review on https://review.openstack.org/#/c/510928/3..4/ironic_inspector/introspect.py? I've implemented your suggestion there | 10:53 |
| patchbot | patch 510928 - ironic-inspector - Use db transactions when error-finish is called | 10:53 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/ironic master: Simplify the logic of validate_network_port https://review.openstack.org/516928 | 10:54 |
| *** makowals has quit IRC | 10:54 | |
| *** etingof has quit IRC | 10:54 | |
| *** zhangfei has quit IRC | 10:55 | |
| milan | dtantsur, btw would <3 your opinion on this one: https://review.openstack.org/#/c/516754/1 | 10:55 |
| patchbot | patch 516754 - ironic-inspector - Relax last introspection lock | 10:55 |
| aarefiev | milan: morning, sure, I've missed it was updated | 10:56 |
| * aarefiev is checking | 10:56 | |
| milan | aarefiev, no worries, for some reason zuul removed it's results from it, I've rechecked it last night | 10:56 |
| dtantsur | milan: to be honest, I feel like -2'ing things that remove locking from inspector.. we've already had enough pain in all parts of body from using a complex db-based schema instead of distributed lock... | 10:56 |
| milan | dtantsur, hmmm | 10:57 |
| *** rcernin has quit IRC | 10:57 | |
| milan | dtantsur, so you'd be more in favour of using tooz locks there | 10:58 |
| dtantsur | anyway, this patch is plainly incorrect at first glance ;) | 10:58 |
| milan | dtantsur, how comes? :) | 10:58 |
| dtantsur | oh wow, it's even worse at the 2nd glance | 10:59 |
| *** sambetts|afk is now known as sambetts | 10:59 | |
| dtantsur | initially I thougt you're misunderstanding the goal of this lock. however, now I see you're replacing the wait with an error. that's a no-go. | 10:59 |
| dtantsur | and it seems racy | 10:59 |
| dtantsur | well, yes, I've been advocating for not using database for locking since day 0, but I was in minority :-/ | 11:00 |
| milan | sort of a best-effort thing: 409 try later | 11:00 |
| dtantsur | no. no. no. no. no | 11:00 |
| dtantsur | no. | 11:00 |
| dtantsur | no. | 11:00 |
| dtantsur | please | 11:00 |
| dtantsur | this is a breaking change, and we have enough of this 409 in ironic | 11:01 |
| milan | that would have been my argument: ironic does that O:-) | 11:01 |
| * aarefiev hehm gerrit hangs and post msg few times | 11:02 | |
| sambetts | yeah I'm seeing the same thing | 11:02 |
| sambetts | aarefiev: ^ | 11:02 |
| dtantsur | milan: search our bugzilla over "node is locked by host" | 11:02 |
| *** lifeless has joined #openstack-ironic | 11:02 | |
| milan | morning sambetts! :) | 11:03 |
| aarefiev | morning sambetts, dtantsur | 11:03 |
| milan | dtantsur, hmm, right but a 409 seems more fair towards the user than a 202 with an eventual failure ;) | 11:03 |
| milan | one can always retry in the client, server's more tricky | 11:04 |
| dtantsur | why eventual failure? it's working right now, e.g. in case of tripleo | 11:04 |
| milan | if fails, it's not reported | 11:04 |
| milan | * to the caller | 11:04 |
| dtantsur | milan: please submit an update to tripleo workflows first then. that'll teach you how easy it is ;) | 11:04 |
| milan | :D | 11:04 |
| dtantsur | sorry, what fails? we merely wait for several seconds, then proceed as normal. that make it work in 99% of cases. | 11:05 |
| dtantsur | if we do this, then we should just remove this option completely, and let the caller ensure this delay | 11:05 |
| dtantsur | which is something that caused users a lot of pain around Kilo or Liberty, I think | 11:05 |
| milan | dtantsur, I mean if you have concurrent calls for same node , it is being retried, when ironic says: 409 node locked by .... what I meant... iirc it's some ~30 retries before giving up | 11:06 |
| milan | but no worries | 11:06 |
| * milan reads on :) | 11:06 | |
| milan | dtantsur, you mean like removing the lock altogether? | 11:07 |
| *** etingof has joined #openstack-ironic | 11:07 | |
| dtantsur | we used to not have this lock, and it was painful | 11:08 |
| dtantsur | now your change will break anyone who runs more than 4-5 introspections in parallel | 11:08 |
| dtantsur | wait, it will break anyone who runs even 2 in parallel | 11:08 |
| milan | actually I've just got a crazy idea: if dnsmasq and inspector are (in tripleo, eventually) packed together in a single container, then the lock actually protects the dnsmasq and can stay as it's still functional (mind the disjoint DHCP pools) | 11:09 |
| dtantsur | but if you simply remove the lock and the feature, it will break people who run 4-5 introspections | 11:09 |
| milan | dtantsur, why you think? I't would tell them: retry on one of them (it's inside a transaction) | 11:09 |
| sambetts | in kolla dnsmasq and inspector are separate containers, and IMO should be separate containers | 11:09 |
| dtantsur | milan: yes, and their scripts will fail, because they don't handle 409 now | 11:10 |
| dtantsur | this is what a breaking change means - what worked starts giving HTTP error | 11:10 |
| milan | dtantsur, ahh....hmmmm | 11:10 |
| dtantsur | sambetts: will our iptables magic work this way? | 11:10 |
| milan | sambetts, why do you think it's better in separate containers? | 11:11 |
| *** phuongnh has quit IRC | 11:11 | |
| milan | dtantsur, so if I implemented it with waiting, but with the select, would it be OK? | 11:12 |
| sambetts | dtantsur: yes, the inspecotr and dnsmasq containers have to run has net=host anyway so they share a net namespace | 11:12 |
| sambetts | milan: because a dnsmasq container is reusable | 11:12 |
| sambetts | milan: and pretty standard | 11:12 |
| dtantsur | milan: I suspect the database code is racy. what if 2 processes do the same select? | 11:12 |
| dtantsur | and then the same check succeeds for them | 11:13 |
| nicodemos | good morning, ironic. | 11:13 |
| milan | dtantsur, it's in a transaction, one of them suppose to pass | 11:13 |
| dtantsur | milan: selects? they're not affected by transactions in some transaction modes | 11:13 |
| dtantsur | hey nicodemos | 11:13 |
| milan | dtantsur, it's inside of the start introspection | 11:13 |
| milan | morning nicodemos! :) | 11:13 |
| milan | dtantsur, which in turn is in write session so one of them should pass | 11:14 |
| nicodemos | thanks for the reviews and hey milan. :) | 11:15 |
| nicodemos | dtantsur :P | 11:15 |
| dtantsur | milan: wait, what are we talking about? the "select" statement on line 659 can return the same result for 2 nodes, and these 2 nodes will run introspection simultaneously | 11:15 |
| dtantsur | for 1 node - yes, transaction will prevent it, I think | 11:15 |
| milan | dtantsur, but that's the point of a transaction, no? | 11:16 |
| * milan might be missing some info though | 11:16 | |
| dtantsur | sorry, I don't get it. transaction does not lock tables completely IIRC | 11:16 |
| dtantsur | you have 2 processes updating two different rows, why should any of these transactiosn fail? | 11:17 |
| milan | dtantsur, yeah, fair point, that seems the bit I missed indeed: the transaction locks a row, not a table | 11:19 |
| dtantsur | right | 11:19 |
| dtantsur | if it locked a table, we would be in trouble with scaling :) | 11:19 |
| milan | dtantsur, correct | 11:19 |
| milan | though honestly the odds are quite low for a race at the same time ;) so it may be a good trade off (supposing I update docstrings and implement a waiting/retry logic), no? O:-) | 11:21 |
| *** serlex has quit IRC | 11:21 | |
| * milan still hopes he could get away with it :D | 11:21 | |
| milan | it's called relaxing after all ;) | 11:21 |
| * dtantsur takes a deep breath | 11:21 | |
| milan | lol :D | 11:21 |
| dtantsur | milan: first, if we bother with HA work, "odds are low" is not an argument. weren't it you who was talking about thousands of introspections? ;) | 11:22 |
| dtantsur | milan: second, the odds are not so low. it's a very common case to launch a lot of introspections at the same time | 11:22 |
| milan | dtantsur, fair enough, the argument should be complete ;) | 11:23 |
| milan | * now, from my pov | 11:24 |
| *** bfournie has quit IRC | 11:28 | |
| milan | sambetts, how about the container being a blackbox that ships both the parts of the service needed to complete the introspection | 11:28 |
| milan | * both/all including the config | 11:28 |
| *** bfournie has joined #openstack-ironic | 11:28 | |
| sambetts | depends on your outlook on containers, from my perspective I'm working with k8s so in that case the containers are self contained processes, and the pod and helm chart includes all the configuration and provides it to the container at deploy time | 11:30 |
| sambetts | so in this case container == process separation | 11:30 |
| sambetts | and trasportability | 11:30 |
| sambetts | but in your case a container == service separation and transportability | 11:31 |
| milan | hmmm, yeah, my point was more like service I think; esp. because both inspector and dnsmasq should be turned off together | 11:31 |
| sambetts | inspector can run without dnsmasq though | 11:32 |
| milan | to prevent unprotected dnsmasq interfering with e.g deploy | 11:32 |
| *** e0ne has quit IRC | 11:32 | |
| sambetts | and I could deploy a different DHCP servicer | 11:32 |
| *** asane has joined #openstack-ironic | 11:32 | |
| *** bfournie has quit IRC | 11:32 | |
| sambetts | in your case I'd have to completely rebuild and redeploy eveyrthing, in my case I'd just swap out the dnsmasq container for a differnet dhcp service | 11:33 |
| sambetts | and in cases when we're using vmedia for inspection (incomplete but on its way) you might not need dnsmasq/dhcp at all | 11:33 |
| milan | fair point, you'd just change a driver | 11:34 |
| sambetts | and when you are deploying bifrost, you just deploy one dnsmasq container | 11:34 |
| milan | how about the turning off? isn't the configuration/dependency of this couple "leaking"? | 11:34 |
| sambetts | to serve for both ironic and inspector | 11:34 |
| *** e0ne has joined #openstack-ironic | 11:35 | |
| milan | sambetts, because you may need the dnsmasq to serve ironic but the other dnsmasq to be shut off? | 11:35 |
| *** asane_ has joined #openstack-ironic | 11:35 | |
| sambetts | in the case of dnsmasq thats a configuration thing, not a service level thing, e.g. you should just change the config in dnsmasq not run two instnaces of DNSmasq | 11:37 |
| sambetts | there are some scenarios when you might run two dnsmasq but then the ironic one and the inspector one will using the same container images, and just run two instances of it | 11:38 |
| sambetts | with different configs | 11:38 |
| milan | I may be missing some information, so if inspector fails how would the dnsmasq be reconfigured as to not allow the inspection image to boot? How is it done now in your case? | 11:40 |
| sambetts | if inspector fails? the service is broken? | 11:41 |
| milan | yup | 11:41 |
| milan | if e.g oom kills inspector | 11:41 |
| *** asane has quit IRC | 11:43 | |
| sambetts | you'd just remove the default file from dnsmasq's config right? | 11:43 |
| sambetts | /from the tftpboot dir | 11:43 |
| milan | yeah, though I'd prefer it being removed automatically ;) | 11:44 |
| milan | I mean what process/controller/thing would disable the booting | 11:44 |
| milan | to prevent the collisions | 11:44 |
| sambetts | in the case of k8s, the containers are deployed as a pair, if one dieds then the pods is marked dead | 11:45 |
| milan | so the information about their mutual dependency "leaked" the container in favour of k8s configuration, no? | 11:46 |
| milan | but maybe it's cool, I dunno | 11:46 |
| *** ccamacho has quit IRC | 11:46 | |
| milan | also other thing, they always have to be scheduled on the same node, correct? | 11:46 |
| *** jcoufal has joined #openstack-ironic | 11:47 | |
| milan | s/they/ inspector and it's dnsmasq sidekick | 11:47 |
| sambetts | yeah, in k8s thats what a pod does | 11:47 |
| * dtantsur has to refill his coffee supply, back in an hour or two | 11:48 | |
| sambetts | np | 11:48 |
| *** dtantsur is now known as dtantsur|bbl | 11:48 | |
| milan | so to me it sounds the decoupling from other services it's better if they're part of a single container rather than the case of a shared dnsmasq | 11:48 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/ironic master: Leaverage oslo.log default log levels https://review.openstack.org/516970 | 11:48 |
| milan | dtantsur|bbl, have a good one! :) thx for the feedback! | 11:49 |
| milan | though I agree it sacrifices a bit of flexibility | 11:49 |
| * milan prefers decoupling | 11:50 | |
| *** jcoufal_ has joined #openstack-ironic | 11:50 | |
| *** dprince has joined #openstack-ironic | 11:52 | |
| milan | also, if they're packed together it means they're integrated (tested, configured, debugged) and one should not worry about issues integrating on their own; so UX should be better | 11:53 |
| sambetts | milan: yeah depends on your outlook on containers, like I said kolla focuses on microservice separation so you have ironic-api ironic-conductor ironic-inspector dnsmasq, and then OSA uses "fat" containers so you have an ironic container or a nova container | 11:53 |
| *** jcoufal has quit IRC | 11:54 | |
| milan | sambetts, right, gonna be tradeoffs I guess, not a single solution that suits them all | 11:54 |
| *** hw_wutianwei has joined #openstack-ironic | 11:54 | |
| *** jlabarre has joined #openstack-ironic | 11:54 | |
| milan | sambetts, do you think it would be possible to have both ways live together for the different usecases/outlooks? | 11:55 |
| * milan very keen on having them packed together if the dnsmasq-based filter is used for inspector | 11:56 | |
| *** bfournie has joined #openstack-ironic | 12:04 | |
| *** lucasagomes is now known as lucas-hungry | 12:04 | |
| *** zhangfei has joined #openstack-ironic | 12:04 | |
| *** litao__ has quit IRC | 12:06 | |
| *** weshay|PTO is now known as weshay | 12:09 | |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/ironic master: Simplify the logic of validate_network_port https://review.openstack.org/516928 | 12:09 |
| openstackgerrit | ChangBo Guo(gcb) proposed openstack/ironic master: Simplify the logic of validate_network_port https://review.openstack.org/516928 | 12:12 |
| *** udesale has joined #openstack-ironic | 12:18 | |
| *** jcoufal_ has quit IRC | 12:21 | |
| *** jcoufal has joined #openstack-ironic | 12:23 | |
| *** trown|outtypewww is now known as trown | 12:25 | |
| *** rbudden has joined #openstack-ironic | 12:26 | |
| *** leitan has joined #openstack-ironic | 12:26 | |
| leitan | Morning metalers | 12:27 |
| leitan | anyone with experience with UCS fiber channel connected servers ? | 12:28 |
| *** catintheroof has joined #openstack-ironic | 12:28 | |
| leitan | im about to make it but ... Error finding the disk or partition device to deploy the image onto: No suitable device was found for deployment using these hints | 12:28 |
| leitan | so im guessing that the coreOS image is not finding via FC my disk | 12:28 |
| leitan | i have hinted the wwn | 12:29 |
| leitan | but maybe its a driver issue | 12:29 |
| leitan | maybe TheJulia sambetts ? | 12:30 |
| * TheJulia reads | 12:32 | |
| TheJulia | Also o/ from somewhere over the pacific | 12:33 |
| sambetts | TheJulia: oooo nice views? | 12:33 |
| TheJulia | leitan: You likely need a different ipa image, maybe Debian or fedora. Something that ships the FC drivers | 12:33 |
| TheJulia | For the habit | 12:33 |
| TheJulia | Err HBA | 12:34 |
| Tengu | -.- why did I read "For the hobbit" ?! | 12:34 |
| sambetts | leitan: I'm not sure on that one, I've had little experience with our (/me is from Cisco) FC stuff, but I would expect that it is a driver thing | 12:34 |
| TheJulia | sambetts: no view, call for a doctor or paramedic a little while ago : ( | 12:34 |
| sambetts | :( | 12:34 |
| TheJulia | We didn’t turn, so I guess we’re not diverting to Hawaii | 12:36 |
| leitan | TheJulia: :( many hours to go ? | 12:36 |
| TheJulia | 9-ish, I think | 12:36 |
| leitan | not that bad, i still remember my 36 hours flight + stops to the Hong Kong summit | 12:37 |
| leitan | ok, ill try why another ipa image, and make sure that has the HBA driver | 12:37 |
| leitan | or maby try to use pxe_ucs and agent_ucs first | 12:39 |
| leitan | as driver for the node | 12:39 |
| sambetts | leitan: the ironic Cisco drivers won't make a difference to that, they only provide as much control as IPMItool | 12:39 |
| *** zhangfei has quit IRC | 12:39 | |
| sambetts | leitan: e.g. power, and boot device etc | 12:39 |
| *** crushil has quit IRC | 12:40 | |
| leitan | sambetts: i was actually reading the driver code now, and yes, besides some api calls to the ucs manager | 12:40 |
| leitan | nothing magical there | 12:40 |
| *** asane_ has quit IRC | 12:40 | |
| leitan | hmmm, i think ill go with the images shipped with the Red Hat director ... | 12:42 |
| leitan | if those guys doesnt have the FC driver on it | 12:42 |
| leitan | nobody does :P | 12:42 |
| sambetts | haha, if not you can build an image using DIB and include additional packages that way (although I don't know what those packages names are) | 12:44 |
| milan | morning leitan, Tengu, TheJulia! :) | 12:46 |
| leitan | sambetts: haha, me neither, so ill go with that first ... director deploys bareemtal on UCS so should be there waiting for me to succeed on this | 12:47 |
| * TheJulia should try and go back to sleep | 12:47 | |
| leitan | Morning milan ! | 12:47 |
| leitan | Did you take some sleeping pills TheJulia ? | 12:47 |
| Tengu | 'ternoon milan :) | 12:47 |
| milan | :) | 12:48 |
| leitan | 9:47 am here in Argentina :P | 12:48 |
| milan | TheJulia, what works for me it's 2xIPA (pun intended) ;) | 12:48 |
| TheJulia | leitan: I think I’ve gotten three or so hours, which is not horrible | 12:48 |
| leitan | not bad really | 12:49 |
| TheJulia | Yeah, Hopefully maybe ice cream will appear shortly | 12:49 |
| leitan | sugar and sleeping are not friends :P | 12:50 |
| TheJulia | Indeed | 12:50 |
| leitan | buy your best friend maybe haha | 12:50 |
| TheJulia | Hmm... the cats must be up. I keep getting alerts for video | 12:50 |
| leitan | buy -> but | 12:50 |
| leitan | i remote feed my cats with and IR servo powered arduino :D | 12:52 |
| openstackgerrit | Merged openstack/ironic master: Allow to set default ifaces in DevStack https://review.openstack.org/513152 | 12:52 |
| openstackgerrit | Merged openstack/ironic master: Remove unnesessary description for config parameters in cinder group https://review.openstack.org/516192 | 12:52 |
| TheJulia | I was in a store this week and one of the things they were selling was a cat camera with remote control laser pointer so you can “play” with your cats remotely | 12:53 |
| leitan | hahaha, got to love iot startups | 12:55 |
| *** gcb_ has quit IRC | 12:57 | |
| *** gcb_ has joined #openstack-ironic | 13:00 | |
| *** sasha21 has joined #openstack-ironic | 13:01 | |
| *** rloo has joined #openstack-ironic | 13:02 | |
| *** lucas-hungry is now known as lucasagomes | 13:08 | |
| *** fellypefca has joined #openstack-ironic | 13:10 | |
| *** baoli has joined #openstack-ironic | 13:16 | |
| fellypefca | Good morning everyone :) | 13:16 |
| fellypefca | Hey, is the zuul gate broken or did I just get a random error? | 13:22 |
| sambetts | fellypefca: there's nothing on the whiteboard https://etherpad.openstack.org/p/IronicWhiteBoard to indicate things are broken, link to patch? | 13:24 |
| fellypefca | sambetts https://review.openstack.org/#/c/468428/ | 13:24 |
| patchbot | patch 468428 - ironic - Migrate python-oneviewclient validations to Ironic... | 13:24 |
| fellypefca | it's on the multinode-multitenant | 13:25 |
| sambetts | fellypefca: that is a known race condition that can occur, and isn't fixed yet, so best bet is to recheck | 13:26 |
| fellypefca | sambetts oh, ok! thanks :) | 13:26 |
| rloo | hi and good morning sambetts, fellypefca | 13:31 |
| sambetts | o/ rloo | 13:31 |
| fellypefca | Good morning rloo :D | 13:33 |
| *** mat128 has quit IRC | 13:34 | |
| rloo | dtantsur|bbl: wrt https://review.openstack.org/#/c/516463/, because I authored it, and there are enuf +1 there, I think it is safe for you to +A. TheJulia is probably still on a flight... | 13:35 |
| patchbot | patch 516463 - ironic (stable/pike) - Add missing job to gate | 13:35 |
| *** zshi has quit IRC | 13:37 | |
| fellypefca | TheJulia Hi! If you have some time, could you take a look? https://review.openstack.org/#/c/480284/ | 13:37 |
| patchbot | patch 480284 - ironic - Remove python-oneviewclient from Ironic OneView dr... | 13:37 |
| milan | morning fellypefca, rloo! :) | 13:48 |
| rloo | hiya milan! | 13:49 |
| milan | aarefiev, just left you a comment on your comment so you can comment while commenting ;) Thx for the review! | 13:49 |
| *** MattMan has quit IRC | 13:49 | |
| *** MattMan has joined #openstack-ironic | 13:49 | |
| milan | aarefiev, so how about I squeeze the patches with a refactor of the finished? | 13:49 |
| milan | for the record, milan is talking about https://review.openstack.org/#/c/510928/4 | 13:50 |
| patchbot | patch 510928 - ironic-inspector - Use db transactions when error-finish is called | 13:50 |
| *** Goneri has quit IRC | 13:51 | |
| fellypefca | morning milan :) | 13:52 |
| *** amarao has joined #openstack-ironic | 13:53 | |
| *** Goneri has joined #openstack-ironic | 13:53 | |
| *** ralonsoh has quit IRC | 13:59 | |
| aarefiev | milan: do we need one more method? Could we handle all the stuff in fsm_event? | 14:00 |
| milan | aarefiev, I failed to figure out how | 14:00 |
| milan | i thought the same | 14:00 |
| milan | aarefiev, it wouldn't be new method but rather dropping the process_error in favour of a refactored node_info.finished | 14:01 |
| *** hshiina has quit IRC | 14:04 | |
| *** crushil has joined #openstack-ironic | 14:04 | |
| *** rama_y has joined #openstack-ironic | 14:05 | |
| *** tesseract has quit IRC | 14:06 | |
| aarefiev | yeah, I thought about dropping process_error as well, but then fsm_event will "smell" I guess | 14:06 |
| *** mat128 has joined #openstack-ironic | 14:06 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-inspector master: Zuul: add file extension to playbook path https://review.openstack.org/516995 | 14:06 |
| *** mat128 has quit IRC | 14:08 | |
| milan | aarefiev, maybe, it's still good for the transitions that don't change more state fields; wait and process and maybe enrol (or maybe not) | 14:10 |
| milan | aarefiev, but believe me, I feel worse than you about it cause it's my code :DDD | 14:10 |
| milan | * that sucks | 14:11 |
| aarefiev | milan: it isn't so bad, I don't have better ideas in my mind | 14:12 |
| aarefiev | so it works for me | 14:13 |
| milan | aarefiev, txh, you mean to refactor the finished? | 14:13 |
| aarefiev | yep | 14:13 |
| milan | and to drop the process_error | 14:13 |
| milan | ack | 14:13 |
| milan | cool, I'll hack that | 14:13 |
| aarefiev | thx | 14:14 |
| milan | aarefiev, btw if you could also revisit the dnsmasq thing that'd be awesome! :) sambetts too :P https://review.openstack.org/#/c/466448/ | 14:16 |
| patchbot | patch 466448 - ironic-inspector - Introducing a dnsmasq PXE filter driver | 14:16 |
| *** nicodemos has quit IRC | 14:16 | |
| *** rama_y has quit IRC | 14:16 | |
| * milan updated aarefiev 's input | 14:16 | |
| leitan | sorry guys quick question, im trying to update the deploy_kernel and ramdisk via openstack cli | 14:17 |
| leitan | im using node set , but can get it right | 14:18 |
| aarefiev | milan, ahh so what the plan for dnsmasq driver testing? | 14:18 |
| leitan | somebody that remember the syntax ? | 14:18 |
| leitan | forget it got it working --driver-info | 14:19 |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic master: Get a new OneView client when needed https://review.openstack.org/510644 | 14:24 |
| *** nicodemos has joined #openstack-ironic | 14:25 | |
| milan | aarefiev, so there's the patch that updates the devstack to deploy with that driver | 14:25 |
| milan | https://review.openstack.org/#/c/468712/55 | 14:25 |
| patchbot | patch 468712 - ironic-inspector - TEST check dnsmasq PXE filter in devstack | 14:25 |
| milan | that's been passing for quite a while | 14:26 |
| milan | I've done some manual checks | 14:26 |
| milan | also anupn did some testing if I'm not mistaken | 14:26 |
| TheJulia | So I found out that the current gerrit web-ui doesn’t let iPad users leave in-line comments anymore :( | 14:26 |
| milan | ouch | 14:27 |
| aarefiev | milan: so are you planning to add new job? | 14:27 |
| *** rfolco has quit IRC | 14:27 | |
| milan | aarefiev, aah | 14:27 |
| milan | I didn't get your question | 14:27 |
| aarefiev | yeah, I mean ci | 14:28 |
| milan | dunno, do you think it's a blocker for the merge? | 14:28 |
| milan | :) | 14:28 |
| milan | I mean if you think we need one, let's have one | 14:28 |
| milan | but so far it seems the patch isn't breaking anyone | 14:28 |
| TheJulia | fellypefca: w/r/t your requested review, I have a nit on the Reno so we don’t feel the need to edit it later. Commented on the review, but also +2’ed. I’d like to see oneview CI reply before wf+1 | 14:29 |
| aarefiev | milan: no it isn't, I guess, but for sure we need to test it on ci | 14:29 |
| milan | aarefiev, regularly you mean, so later on we'd deprecate the default iptables, correct? | 14:29 |
| aarefiev | yeah, if we say that we support dnsmasq driver we need to test it regularly | 14:31 |
| *** rama_y has joined #openstack-ironic | 14:31 | |
| aarefiev | or say it's experimental feature | 14:31 |
| *** udesale has quit IRC | 14:31 | |
| milan | TheJulia, how you solved the in-line comments? Firefox works? I'd say it's worth poking the infra about it :-/ | 14:31 |
| milan | aarefiev, makes sense | 14:32 |
| *** windsn has quit IRC | 14:32 | |
| *** fellypefca has quit IRC | 14:32 | |
| TheJulia | milan: you used to be able to go into single page diff view and be able to leave a comment | 14:32 |
| TheJulia | milan: at least with safari on iOS. | 14:32 |
| * milan 's OS X safari sometimes freaks out too btw | 14:33 | |
| milan | TheJulia, I see, I guess it should be fixed | 14:33 |
| TheJulia | I’ve never had it freak anymore than chrome on OS X. | 14:33 |
| *** fellypefca has joined #openstack-ironic | 14:34 | |
| TheJulia | milan: likely, although I’ve been trying to encourage one of the iOS git client authors to add gerrit support.’ | 14:34 |
| *** windsn has joined #openstack-ironic | 14:34 | |
| milan | TheJulia, that'd be nice | 14:35 |
| fellypefca | TheJulia Thank you! We will fix the Reno and investigate the CI issue. :) | 14:36 |
| milan | TheJulia, mine sometimes fails to show the whole page so the control elements aren't accessible so no work unless restart browser :) | 14:36 |
| *** fellypefca has quit IRC | 14:37 | |
| *** fellypefca has joined #openstack-ironic | 14:37 | |
| milan | dtantsur|bbl, would have blamed me for using the OS X :P | 14:37 |
| TheJulia | milan: fun! I need to go back to Linux native, although hopefully I’ll have a home lab again in a month or two and can just do dev on that. | 14:39 |
| *** zhangfei has joined #openstack-ironic | 14:39 | |
| milan | TheJulia, reminds me of the Dexter's Laboratory cartoon :D | 14:40 |
| * TheJulia wonders if she could find a low cost cabinet and just sink bolts into the slab.... | 14:40 | |
| *** markguz has quit IRC | 14:42 | |
| milan | frankenstein style ;) | 14:42 |
| *** tesseract has joined #openstack-ironic | 14:43 | |
| rloo | hi (and good morning even if it isn't) TheJulia, where are you now? :) | 14:47 |
| rloo | milan and inspector folks, up to you to review zuul v3-related patches, eg: https://review.openstack.org/#/c/516784/ | 14:48 |
| patchbot | patch 516784 - python-ironic-inspector-client - Move legacy python-ironic-inspector-client jobs | 14:48 |
| milan | rloo, thx | 14:49 |
| * milan looks | 14:49 | |
| rloo | milan: this one too: https://review.openstack.org/#/c/516995/ | 14:49 |
| patchbot | patch 516995 - ironic-inspector - Zuul: add file extension to playbook path | 14:49 |
| rloo | milan: sorry, the latter one hasn't passed zuul yet | 14:50 |
| milan | rloo, i like the latter one more :DD | 14:50 |
| milan | ouch | 14:50 |
| rloo | milan: heh. be glad you didn't submit the patches | 14:50 |
| milan | rloo, thanks for submitting actually | 14:50 |
| * milan still fights HA these days | 14:50 | |
| rloo | milan: we have too many ironic-related projects :-( https://etherpad.openstack.org/p/ironic-zuulv3-intree-tracking | 14:51 |
| * rloo volunteers to backport inspector zuul-related patches to stable branches. sigh. | 14:52 | |
| milan | omg we're the biggest project | 14:52 |
| rloo | milan: really? find that hard to believe | 14:52 |
| rloo | milan: define 'big' ! | 14:52 |
| milan | well based on the number of subprojects ;) | 14:53 |
| rloo | milan: cuz 'big in spirit', yeah, i believe it! :D | 14:53 |
| milan | :D | 14:53 |
| TheJulia | rloo: about 6 hours out from Sydney over the pacific | 14:53 |
| rloo | milan: oh. would have thought neutron would beat us; don't they have a lot of plugins | 14:53 |
| rloo | TheJulia: wow | 14:53 |
| TheJulia | And Good morning | 14:53 |
| * TheJulia needs to take a nap soon-ish | 14:53 | |
| * milan just saying from the impression of that etherpad ;) | 14:53 | |
| rloo | milan: ha ha, good thing pas-ha started that! | 14:54 |
| milan | TheJulia, sleep well! | 14:54 |
| milan | pas-ha ++ :) | 14:54 |
| milan | TheJulia, btw when do you present? | 14:54 |
| * milan would like to fingers-crossed that time, or what do you say ;) | 14:55 | |
| TheJulia | milan: no, likely what we call tapcon in the states, basically hex head concrete screw. | 14:55 |
| sambetts | neutron only doesn't have more sub-projects because they kicked us all out | 14:55 |
| TheJulia | milan: uhh, which time :) | 14:55 |
| milan | we say drzet palce (modulo the punctuation) | 14:56 |
| milan | sambetts, there-there :D | 14:56 |
| * rloo sees that sambetts is still bitter er happy about that neutron decision | 14:56 | |
| *** fragatina has joined #openstack-ironic | 14:56 | |
| milan | TheJulia, no issues rounding it up to days :D | 14:56 |
| milan | most likely gonna be night my time | 14:56 |
| *** fragatina has quit IRC | 14:57 | |
| sambetts | rloo: haha, /me has a business lunch with the new neutron PTL to discuss it in Syndey | 14:57 |
| TheJulia | milan: Monday and Tuesday, I think. | 14:57 |
| rloo | sambetts: good luck. you're paying, right? ;) | 14:57 |
| milan | sambetts, that sounds like a serious business lunch :) | 14:57 |
| *** fragatina has joined #openstack-ironic | 14:57 | |
| leitan | guys quick question, i have switches de deploy image from CoreOS to redhat, it boots, the python ironic agent heartbeats ok to /v1/commands 200 OK, and the node is on clean_wait status | 14:57 |
| leitan | but stays there ... | 14:57 |
| milan | TheJulia, good luck then! :) | 14:58 |
| leitan | can be possible that is actually cleaning ? :O | 14:58 |
| leitan | or just stuck in there doing nothing | 14:58 |
| * milan -> downstream -> mtg | 14:58 | |
| TheJulia | leitan: possibly, console output would be good to check | 14:58 |
| *** mgoddard_ has quit IRC | 14:59 | |
| sambetts | leitan: you could also check the node internal_driver_info field and see if it has clean steps lists | 14:59 |
| sambetts | listed | 14:59 |
| * milan googled pics of tapcons, ack, should hold better than the frankenstein ones ;) | 15:00 | |
| leitan | TheJulia: console output says http://i67.tinypic.com/jpk6f6.png | 15:01 |
| *** Nisha_Agarwal has joined #openstack-ironic | 15:01 | |
| sambetts | leitan: maintenance is false? | 15:01 |
| rloo | TheJulia: I'm good with the changes, but are the non-voting failures ok? https://review.openstack.org/#/c/516389/ | 15:01 |
| patchbot | patch 516389 - bifrost (stable/pike) - Move legacy bifrost jobs in-tree | 15:01 |
| leitan | sambetts: is false yes | 15:01 |
| sambetts | cool, just making sure :) | 15:02 |
| leitan | sambetts TheJulia http://paste.openstack.org/show/625218/ | 15:02 |
| leitan | node show with the internal info steps | 15:02 |
| sambetts | leitan: in the clean_step field is shows you the running step | 15:04 |
| *** fragatina has quit IRC | 15:04 | |
| sambetts | erase_devices can take a long time it runs a full shred of the disk | 15:04 |
| *** rama_y has quit IRC | 15:04 | |
| TheJulia | A very long time, depending on the amount of storage present :( | 15:05 |
| leitan | sambetts: maybe finally saw the HD via the fiber channel | 15:05 |
| leitan | and its erasing the 44gb LUN | 15:05 |
| TheJulia | Interesting that it took such a long time to boot, at least by the second count on the screenshot unless you hit enter a few times and the login screen refreshed | 15:05 |
| sambetts | leitan: \o/ | 15:06 |
| leitan | TheJulia: yes i hit enter several times, boot time was normal | 15:06 |
| leitan | sambetts TheJulia the agent should be outputting on stdout if is actually erasing something ? | 15:07 |
| TheJulia | rloo: the non voting jobs have a high failure rate due to any transitory issue impacting them. I need to look at the logs, but that is not a in-flight satellite wifi thing. | 15:07 |
| rloo | TheJulia: ok, i looked at the master patch and it was the same there, so I +2'd it :) | 15:07 |
| *** mat128 has joined #openstack-ironic | 15:08 | |
| TheJulia | leitan: makes sense, so it is shredding then since we have no real support to detect if an attached San lun supports magical features, like to zero out the disk virtually | 15:08 |
| * TheJulia wonders if we should write that, it should only be a few hundred lines of code. | 15:08 | |
| TheJulia | Then again... testing.... would be very difficult | 15:09 |
| sambetts | yeah its basically right now it basically doing dd from /dev/zero | 15:09 |
| sambetts | for 44gb | 15:09 |
| leitan | sambetts TheJulia it would be great if its actually executing it, that the agent log it | 15:09 |
| leitan | that means that it found the HD | 15:10 |
| leitan | and its doing something and not stuck | 15:10 |
| *** mjura has quit IRC | 15:10 | |
| *** zhangfei has quit IRC | 15:10 | |
| TheJulia | leitan: It should be reporting it out to the console, maybe it was before the screen refreshed with the login prompt :( | 15:11 |
| rloo | pas-ha: networking-generic-switch has stable branches? Were you going to backport the zuul v3-related stuff to them? | 15:11 |
| leitan | actually i found out about the FC drivers due to the ouput that ironic wasnt able to write to a disk that didnt found on the server | 15:11 |
| leitan | maybe something similar for clenaing | 15:11 |
| leitan | TheJulia: ooh ok | 15:13 |
| leitan | maybe i missed it | 15:13 |
| leitan | but the peridiocal report ... doesnt say anything | 15:13 |
| leitan | maybe there too ? | 15:13 |
| leitan | i know its a heartbeat, but something like that there | 15:14 |
| sambetts | leitan: ironic itself actually checks if its running / not running and will keep the clean_step field up to date on every heartbeat from the agent | 15:15 |
| TheJulia | I don’t remember exactly the mechanism used with the heartbeat to report status. We call an external utility to handle the wipe, and sadly it, as far as I’m aware, does not report anything while it is running. IPA is just waiting for the pid to exit 0 basically | 15:15 |
| *** ijw has joined #openstack-ironic | 15:15 | |
| sambetts | yeah there isn't a percentage progress or anything, only running or finished | 15:16 |
| leitan | TheJulia: ok, so if i can log in into the deploy image i can look up for a dd running command | 15:16 |
| TheJulia | sambetts: I wonder if it supports something like some of the versions of dd do with SIG_USR1/2 to get status/progress data... | 15:16 |
| TheJulia | leitan: it is actually using shred on the device, not dd. Thus we support patterns like write random data, then zero out. We also support things like ATA secure erase via different logic, but with FC we don’t have concept or ability to look up the capabilities over scsi and act accordingly | 15:18 |
| sambetts | I expect it does on things like coreos and other OSs but tinycore has the busybox version of all those utilities so it might not | 15:18 |
| TheJulia | Lack of hardware access is likely a big part of it | 15:18 |
| leitan | TheJulia: ill look for that then | 15:18 |
| leitan | is there a redhatter here ? the default password is not root/ROOTPW anymore :( | 15:20 |
| sambetts | dtantsur|bbl, milan ^ | 15:20 |
| * milan scrolls back | 15:20 | |
| *** catintheroof has quit IRC | 15:21 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-inspector stable/pike: Add zuul3 jobs in-tree https://review.openstack.org/517010 | 15:24 |
| milan | leitan, here but dunno about passwd | 15:24 |
| milan | btw etingof has a spec up for reporting the progress in these cases | 15:24 |
| milan | leitan, actually the user might be cloud-user but I dunno for the IPA, what the password might be | 15:25 |
| milan | that's for the regular image (and I guess we make sure root can't log-in) | 15:25 |
| *** rpioso|afk is now known as rpioso | 15:26 | |
| milan | s/IPA/IPA based on RHEL/ | 15:26 |
| * milan risen a downstream question | 15:27 | |
| *** AlexeyAbashkin has quit IRC | 15:29 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-inspector stable/ocata: Add zuul3 jobs in-tree https://review.openstack.org/517011 | 15:30 |
| *** rbartal_ has quit IRC | 15:32 | |
| *** rbartal__ has quit IRC | 15:32 | |
| *** rbartal has quit IRC | 15:32 | |
| milan | btw I guess a conductor doesn't execute anything on IPA over SSH... or? | 15:34 |
| sambetts | milan: conducotr uses IPA rest API to control it | 15:35 |
| milan | yeah so that way no ssh keys one could piggy-back with | 15:35 |
| sambetts | ansible driver uses ansible to do stuff on the baremetal which requires SSH, but only that the image has an authorized_key built into it | 15:36 |
| sambetts | so no private keys will be able to be stolen from the image | 15:36 |
| openstackgerrit | Ruby Loo proposed openstack/ironic-python-agent stable/pike: Move legacy jobs into tree https://review.openstack.org/517013 | 15:37 |
| *** spectr has quit IRC | 15:37 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-python-agent stable/ocata: Move legacy jobs into tree https://review.openstack.org/517014 | 15:38 |
| rpioso | Good morning, ironicers | 15:38 |
| milan | sambetts, ++ thought the same | 15:38 |
| milan | morning rpioso | 15:38 |
| milan | :) | 15:38 |
| *** Sukhdev_ has joined #openstack-ironic | 15:39 | |
| *** dprince has quit IRC | 15:43 | |
| milan | rloo, how do I tell from the zuul log that your patch was "executed"? :) | 15:44 |
| rpioso | milan: :) | 15:44 |
| rloo | milan: good question. dunno. | 15:44 |
| milan | lol :D | 15:44 |
| rloo | milan: i just look to see if zuul runs the same tests it ran before. | 15:44 |
| rloo | milan: i guess maybe i should look through all the test logs to see if they worked, but... that would be too much for me... | 15:45 |
| * milan guess somewhere in the logs should be mentioned that it's picking the settings from the zuul.yaml | 15:45 | |
| *** hw_wutianwei has quit IRC | 15:45 | |
| rloo | milan: the other way to look at it is, when the stuff is removed from project-config an openstack-zuul-whatever, the tests should still run :) | 15:45 |
| milan | rloo, basically I'm looking for a line somewhere that would just prove the ./zuul.yaml was picked up | 15:46 |
| milan | aha | 15:46 |
| milan | so if the other patch fails, this one didn't work? | 15:46 |
| milan | :) | 15:46 |
| rloo | milan: i was just assuming it was picked up, cuz that's the pattern used by all the other projects :) | 15:46 |
| rloo | milan: yeah, there are 3 patches. but i don't know, until after the other two land, whether you will see failures... | 15:47 |
| milan | rloo, you mean assuming? ;) | 15:47 |
| milan | * that's the pattern...* | 15:47 |
| sambetts | it will be picked up, and loaded by zuulv3 (if not thats a bug in zuulv3) also the job names should be different from the legacy ones so you should see both jobs running while there is an overlap | 15:47 |
| rloo | milan: i mean, i don't know that we'll know if things fail, until *after* the other two patches land. | 15:47 |
| *** dtantsur|bbl is now known as dtantsur | 15:47 | |
| milan | rloo, cool, I'm just joking :) | 15:48 |
| rloo | milan: maybe I should have said 'I trust...' :D | 15:48 |
| milan | dtantsur appeared. magic. | 15:48 |
| milan | :D | 15:48 |
| dtantsur | leitan: not sure what the question is about, but I doubt we have default passwords anywhere | 15:48 |
| dtantsur | maaagical owlet! | 15:48 |
| dtantsur | morning rloo, rpioso, TheJulia | 15:49 |
| * milan would love to reddit dtantsur 's post about passwords :DDDD | 15:49 | |
| *** ralonsoh has joined #openstack-ironic | 15:50 | |
| *** rama_y has joined #openstack-ironic | 15:53 | |
| rpioso | dtantsur: :) | 15:56 |
| *** Sukhdev has joined #openstack-ironic | 15:57 | |
| milan | rloo, +2 on that I'm not familiar with the details but I believe since it's running OK should be cool to merge | 16:00 |
| rloo | milan: thx. as sambetts mentioned, if it didn't work, you wouldn't see the additional two (same, but diff name) tests. | 16:01 |
| * milan believes in what rloo trusts :) | 16:01 | |
| * rloo has trust that milan is a team player! | 16:02 | |
| milan | :D | 16:02 |
| *** ijw has quit IRC | 16:02 | |
| windsn | Hi, can anyone kindly tell me if a baremetal node is deployed via nova-compute, does the compute-manager always tried to bind the newly created neutron port to the node that is being used to spawn a baremetal node ? Thanks | 16:06 |
| *** amarao1 has quit IRC | 16:06 | |
| sambetts | windsn: that depends what version of OpenStack you are using | 16:06 |
| sambetts | I think we fixed that bug in the last release | 16:07 |
| windsn | sambetts: I am using Newton. According to the documentation, https://docs.openstack.org/ironic/latest/user/index.html , Nova compute manager creates (unbound) tenant virtual interfaces (VIFs) in the Networking service according to the network interfaces requested in the nova boot request | 16:07 |
| leitan | dtantsur: about the ipa image user and password RH 7.4 | 16:07 |
| *** catintheroof has joined #openstack-ironic | 16:07 | |
| openstackgerrit | Merged openstack/ironic stable/ocata: Add job definitions yaml to zuul configuration https://review.openstack.org/516438 | 16:07 |
| rloo | dtantsur (and others): I've updated zuul v3 status on etherpad: zhttps://etherpad.openstack.org/p/ironic-zuulv3-intree-tracking | 16:08 |
| leitan | thanks milan | 16:08 |
| rloo | dtantsur: I need to poke you to get sushy & virtualbmd ones done :) | 16:08 |
| rloo | dtantsur: poke poke poke :D | 16:08 |
| milan | lol :D | 16:08 |
| milan | dtantsur is talking right now on a meeting I should pay attention to :D | 16:09 |
| leitan | hey, node is available now sambetts TheJulia :) | 16:09 |
| sambetts | windsn: yeah, so there are two things, if you are using the "neutron" network interface in Ironic for dynamic tenent network programming, then nova creates the port in neutron leaves it unbound and sends it to ironic and then ironic will bind it with the information it has about the baremetal and where its physically connected | 16:09 |
| sambetts | leitan: awesome!!!! | 16:09 |
| leitan | lets retry the nova boot | 16:09 |
| windsn | sambetts: Is this a bug in Newton ? nova-compute manager for ironic trying to bind to ports, because i keep getting the error "ERROR nova.compute.manager PortBindingFailed: Binding failed for port 3d21808c-fb7e-46a2-8631-68bb08d604ba, please check neutron logs for more information." ? Thanks | 16:10 |
| *** trown is now known as trown|lunch | 16:10 | |
| sambetts | windsn: yes | 16:10 |
| *** mgoddard has joined #openstack-ironic | 16:10 | |
| sambetts | windsn: it was fixed in Ocata | 16:11 |
| dtantsur | leitan: I'm pretty sure there is not any passwords by default | 16:11 |
| dtantsur | rloo: thanks :) ETOOMANYTHINGSTODO | 16:11 |
| dtantsur | if anybody wants to take over sushy and vbmc - go ahead | 16:11 |
| dtantsur | otherwise I'll get there by EOW | 16:11 |
| leitan | dtantsur: thanks | 16:12 |
| rloo | dtantsur: thx! heh | 16:12 |
| sambetts | windsn: that bug will occur if there is no neutron agent running on the Nova compute that is running the ironic agent, if there is a newton agent running in that location then it'll "fake" bind it to that host | 16:13 |
| windsn | sambetts: Can you kindly point me to the commit where this was fixed? I can try cherry-picking that commit to resolve this issue? Another thing, my enabled_network_interfaces = flat,noop, so when this is the case, only ironic binds, not nova compute-manager ? | 16:13 |
| *** amarao has quit IRC | 16:13 | |
| windsn | sambetts: Yeah i dont have an all-in-one system, instead i have a separate controller + compute node, and the compute node is running neutron dhcp/L2 agent etc... whereas the ironic conductor is running on my controller node, so binding will fail for me | 16:14 |
| sambetts | windsn: where is your ironic-nova-compute running? on the controller? | 16:15 |
| sambetts | windsn: it was quite a significant change that fixed the bug, new APIs in ironic and everything so not an easy backport :/ | 16:15 |
| windsn | sambetts: btw, in nova.conf when used for baremetal, should this be set "compute_manager=ironic.nova.compute.manager.ClusteredComputeManager" ? | 16:16 |
| sambetts | I don't think so, but dtantsur might be able to confirm ^ | 16:17 |
| windsn | windsn: Yeah the nova-compute for ironic is running on that controller | 16:17 |
| windsn | sambetts: Yeah the nova-compute for ironic is running on that controller | 16:17 |
| *** windsn has quit IRC | 16:17 | |
| dtantsur | windsn: this is not required since IIRC Newton | 16:17 |
| dtantsur | required before that | 16:18 |
| sambetts | :/ windsn's DC'd... | 16:18 |
| dtantsur | :( | 16:18 |
| *** windsn has joined #openstack-ironic | 16:19 | |
| leitan | hmmm ... same results when nova boot | 16:20 |
| leitan | ill remove the wwn attribute from the node | 16:20 |
| *** Sukhdev has quit IRC | 16:20 | |
| windsn | sambetts: With the bug fixed, would i still need to have neutron agents running on the controller node, which is running nova-compute ? No right ? | 16:22 |
| *** efried is now known as efried_rollin | 16:22 | |
| dtantsur | windsn: in case you missed my comment, compute_manager was required before Newton | 16:23 |
| sambetts | windsn: correct, with it fixed there isn't a need to have a neutron agent running on the controller node | 16:23 |
| *** markguz has joined #openstack-ironic | 16:24 | |
| *** ijw has joined #openstack-ironic | 16:24 | |
| sambetts | windsn: you might be able to avoid this error being fatal by setting vif_plugging_is_fatal = false in your nova config | 16:25 |
| sambetts | windsn: the error will still occur by the deployment should continue | 16:26 |
| windsn | dtantsur: Thanks for pointing that out, I missed your comment earlier | 16:26 |
| sambetts | and because its a port for baremetal thats pre-configured at actually should work as expected even though its "unbound" | 16:27 |
| windsn | sambetts: I can try setting "vif_plugging_is_fatal" and see what happens .. Unfortunately i dont have a choice here, since I am using Newton, I would have to manually port this bug in my system running Newton, would be really grateful if you can kindly point me to the commit -id for this | 16:28 |
| *** serlex has joined #openstack-ironic | 16:28 | |
| sambetts | windsn: if you missed the message above, I said it was quite a significant change (many commits in nova and ironic) that fixed the bug, new APIs in ironic etc so not such an easy backport :/ | 16:29 |
| leitan | anyone kwons if its extremely necesary specify the wwn to do a FC installation ? | 16:29 |
| *** markguz has quit IRC | 16:30 | |
| leitan | cleaning seems to happen just fine | 16:30 |
| sambetts | windsn: https://github.com/openstack/nova/blame/stable/ocata/nova/virt/ironic/driver.py is the commit in nova, but that requires other things in ironic | 16:30 |
| leitan | ill try to run the nova boot removing the wwn parameter | 16:30 |
| sambetts | leitan: depends if there are any other disks connected to the system | 16:30 |
| leitan | sambetts: just one | 16:31 |
| sambetts | leitan: without any root device hints ironic will pick the first HDD it finds >4gb | 16:31 |
| sambetts | or the smallest HDD it finds > 4gb | 16:31 |
| sambetts | (i can't remeber which it is) | 16:31 |
| leitan | great so should work on my sceneario without the wwn | 16:31 |
| Nisha_Agarwal | dtantsur, hi....is there any update from customer on the issue https://bugzilla.redhat.com/show_bug.cgi?id=1460915 | 16:31 |
| openstack | bugzilla.redhat.com bug 1460915 in python-proliantutils "The pxe_ilo driver ignores power requests under certain conditions with HP BL460" [High,Assigned] - Assigned to dtantsur | 16:31 |
| leitan | BTW i was able to catch the the cleaning running on the STDOUT of the server kvm now | 16:32 |
| sambetts | leitan: nice!! | 16:32 |
| dtantsur | Nisha_Agarwal: hi! no private comments there, so I guess nothing.. | 16:32 |
| Nisha_Agarwal | dtantsur, ok. | 16:32 |
| openstackgerrit | Merged openstack/ironic master: Reword interface information in multitenancy docs https://review.openstack.org/516687 | 16:33 |
| leitan | ok ill try doing that, removing the wwn | 16:33 |
| leitan | sorry to be so "asky" | 16:33 |
| *** markguz has joined #openstack-ironic | 16:34 | |
| sambetts | leitan: no problem, its nice to talk to someone trying to run the software and hear their experiences because it can lead to us improving our docs (btw that patch about the interfaces merged) or discover issues we can improve | 16:35 |
| leitan | certaintly, i guess im used to other core services IRC hostility hahaha | 16:36 |
| vdrok | good morning everyone! | 16:36 |
| * vdrok was having fun debugging icehouse nova for the whole day | 16:36 | |
| dtantsur | morning vdrok, happy icehouse | 16:36 |
| dtantsur | leitan: ugh, I'm sorry for your past experience :( | 16:37 |
| vdrok | hey dtantsur :) | 16:37 |
| *** tzumainn has joined #openstack-ironic | 16:37 | |
| leitan | vdrok: thats fun, like neutron database deadlocking | 16:38 |
| leitan | dtantsur: used to it, been here since B release | 16:38 |
| dtantsur | wow | 16:38 |
| * dtantsur is here since I | 16:39 | |
| leitan | first time trying Ironic though | 16:39 |
| vdrok | leitan: maybe you should have started with icehouse ironic? to better understand the project evolution :D | 16:40 |
| dtantsur | awesome :) I'm really glad to see more people interested in ironic | 16:40 |
| dtantsur | icehouse ironic LOL | 16:40 |
| dtantsur | ironic in icehouse was VERY ironic | 16:40 |
| windsn | sambetts: Just by setting "vif_plugging_is_fatal=False ", I still see this error which prevents deployment https://thepasteb.in/p/JZhpwqRLGRKcg | 16:40 |
| sambetts | windsn: see that error in nova? | 16:41 |
| sambetts | windsn: or in neutron? | 16:41 |
| * Nisha_Agarwal agree with dtantsur :) .... | 16:41 | |
| leitan | dtantsur: in icehouse most projects were ironic :P | 16:42 |
| dtantsur | true :) | 16:42 |
| windsn | sambetts: In nova its just a warning, i see this in openstack.log .. i think its coming from neutron | 16:42 |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic master: Remove python-oneviewclient from Ironic OneView drivers https://review.openstack.org/480284 | 16:42 |
| leitan | image we deployed an almost 1000 node cloud with CACTUS release ... | 16:43 |
| leitan | nova network ... | 16:43 |
| leitan | good times | 16:43 |
| windsn | sambetts: Actually nevermind, the error is still coming from nova https://thepasteb.in/p/Q1hBPk8rERKf8 | 16:43 |
| dtantsur | leitan: heh, cool :) I think nova-network is still there, if you want to give it a try :D | 16:44 |
| sambetts | windsn: did you restart the ironic nova compute and other nova services after setting that value? | 16:45 |
| sambetts | windsn: my understanding is that settting that value *should* mean port binding failing doesn't break the deployment the port will just remain "DOWN" in neutron | 16:45 |
| windsn | sambetts: In nova-compute logs, there is just a warning that binding fails, but the scheduler still fails with the error.. Yeah i restarted ironic nova-compute and nova-scheduler | 16:45 |
| *** catintheroof has quit IRC | 16:45 | |
| sambetts | :/ | 16:46 |
| sambetts | that seems wrong | 16:46 |
| *** markguz has quit IRC | 16:48 | |
| *** markguz has joined #openstack-ironic | 16:49 | |
| *** catintheroof has joined #openstack-ironic | 16:52 | |
| *** pcaruana has quit IRC | 16:52 | |
| *** marios has quit IRC | 16:54 | |
| *** ijw has quit IRC | 16:56 | |
| *** rama_y has quit IRC | 16:59 | |
| windsn | sambetts: I've tried again by restarting all nova services, nova-compute doesn't see this as an error anymore, but the nova-scheduler can't get past this, and flags this as an Error and blocks deployment.... Would you think i may have to set some other flag along with "vif_plugging_is_fatal=False " ? Thanks | 16:59 |
| openstackgerrit | Nguyen Van Trung proposed openstack/ironic master: Add additional capabilities discovery for iRMC driver https://review.openstack.org/417347 | 17:00 |
| *** markguz has quit IRC | 17:02 | |
| *** e0ne has quit IRC | 17:02 | |
| *** openstackgerrit has quit IRC | 17:03 | |
| sambetts | windsn: you might be able to repair it by setting both vif_plugging_is_fatal = False, and applying this diff to the ironic driver http://paste.openstack.org/show/625235/ which will prevent nova trying to bind the port | 17:04 |
| sambetts | ironic nova driver that is | 17:04 |
| windsn | sambetts: Thanks for the link... I'll try and repair it | 17:05 |
| *** markguz has joined #openstack-ironic | 17:05 | |
| sambetts | windsn: unfortunatly newton is now EOL so any bug fix can't be backported upstream any longer | 17:05 |
| *** jcoufal has quit IRC | 17:07 | |
| *** jcoufal has joined #openstack-ironic | 17:07 | |
| windsn | sambetts: I understand... will try your suggestion , hopefully it works | 17:09 |
| sambetts | windsn: i also ran into a similar issue a few months ago trying newton, luckly it was a new project and weren't too heavily tied to newton and were able to move to ocata instead | 17:11 |
| *** serlex has quit IRC | 17:11 | |
| *** ijw has joined #openstack-ironic | 17:12 | |
| sambetts | windsn: but I remeber that ^ was one of the possible solutions I came up with (but it wasn't accepted by upstream at the time because the port then remain forever "unbound" even though it actually works) | 17:13 |
| *** lucasagomes is now known as lucas-afk | 17:13 | |
| *** fragatina has joined #openstack-ironic | 17:16 | |
| windsn | sambetts: Yeah we are only moving to the next release in January , but need to demo ironic with our systems in a week or so and need to get past this using Newton | 17:17 |
| *** jcoufal has quit IRC | 17:21 | |
| *** jcoufal has joined #openstack-ironic | 17:22 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 17:22 | |
| *** Nisha_Agarwal has quit IRC | 17:30 | |
| *** trown|lunch is now known as trown | 17:34 | |
| *** AlexeyAbashkin has quit IRC | 17:36 | |
| leitan | sambetts: | provision_state | active | | 17:43 |
| leitan | i think i did it | 17:43 |
| sambetts | leitan: :O!!! | 17:43 |
| * leitan starts crying | 17:43 | |
| leitan | TIP: do not overconfigure stuff | 17:43 |
| *** openstackgerrit has joined #openstack-ironic | 17:44 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic master: devstack: Clean up some of the devstack code https://review.openstack.org/517063 | 17:44 |
| dtantsur | leitan: congrats! :) | 17:44 |
| dtantsur | and ++ to 'not overconfigure' :) | 17:44 |
| jlvillal | TheJulia: https://review.openstack.org/#/c/516463/ recheck finished and the jobs appear to be there now. | 17:46 |
| patchbot | patch 516463 - ironic (stable/pike) - Add missing job to gate | 17:46 |
| *** Sukhdev_ has quit IRC | 17:49 | |
| leitan | dtantsur: thanks, almost there, seems the booting process is not stuck on "raising network interfaces" | 17:50 |
| leitan | not = now | 17:51 |
| sambetts | leitan: which OS? and does it have cloud-init@ | 17:51 |
| sambetts | ? | 17:51 |
| leitan | sambetts: ubuntu, i built the image with the dib | 17:52 |
| leitan | it has cloudinit | 17:52 |
| leitan | now its failing to reach the metadata ip | 17:52 |
| leitan | cause seems to failed getting dhcp | 17:52 |
| openstackgerrit | Dmitry Tantsur proposed openstack/python-ironicclient master: Correct validation of RAID config in functional tests https://review.openstack.org/517065 | 17:53 |
| sambetts | hmmm, does your network have a neutron router attached? because if it doesn't then it'll require the isolated_metadata flag to be enabled in neutron to allow metadata via the DHCP namespace instead of via router namespace | 17:53 |
| openstackgerrit | Dmitry Tantsur proposed openstack/python-ironicclient master: Switch the deprecated "ironic" CLI to "latest" API version by default https://review.openstack.org/515064 | 17:53 |
| leitan | sambetts: yes but the faiulre is one step earlier | 17:53 |
| leitan | the baremetal doesnt get ip | 17:53 |
| leitan | on the interfaces | 17:53 |
| leitan | after deploying | 17:54 |
| sambetts | :/ oh... :/ weird... maybe a driver thing? | 17:54 |
| dtantsur | any reason not to use a configdrve? | 17:55 |
| sambetts | yeah ideally configdrive is the best option for metadata for baremetal | 17:56 |
| sambetts | much more reliable | 17:56 |
| leitan | sambetts dtantsur i created the image this way: DIB_CLOUD_INIT_DATASOURCES="Ec2, ConfigDrive, OpenStack" disk-image-create -o baremetal-$DISTRO_NAME-$DIB_RELEASE $DISTRO_NAME baremetal bootloader | 17:56 |
| sambetts | is config drive enable by deffault in nova or if not you should use --config-drive on the nova boot command to pass config drive info | 17:57 |
| sambetts | but if the baremetal isn't getting an IP thats a bigger concern | 17:57 |
| sambetts | leitan: on your DIB image create you will probably need to include the dhcp-all-interfaces element | 17:57 |
| sambetts | because otherwise ubuntu only brings up the first network interface it sees and if you have multiple interfaces that might not be the one that neutron has a port for | 17:58 |
| sambetts | and the mac won't match so then the machine won't get DHCP | 17:58 |
| sambetts | in newer ironic's we actually pass network interface configuration in via config drive to bring up the right interfaces | 17:59 |
| *** penick has joined #openstack-ironic | 17:59 | |
| leitan | sambetts: that will be it | 18:00 |
| *** ralonsoh has quit IRC | 18:01 | |
| leitan | ill recreate the image | 18:01 |
| leitan | disk-image-create dhcp-all-interfaces -o baremetal-$DISTRO_NAME-$DIB_RELEASE $DISTRO_NAME baremetal bootloader | 18:03 |
| *** penick has quit IRC | 18:04 | |
| *** mgoddard has quit IRC | 18:04 | |
| * leitan running the image builder sponsored by mycrappyinternetconnection.com | 18:05 | |
| *** windsn has quit IRC | 18:05 | |
| * milan be gone :P | 18:06 | |
| *** milan has quit IRC | 18:06 | |
| sambetts | o/ milan_ | 18:06 |
| leitan | mikal: njoy ! | 18:10 |
| leitan | already gone :P | 18:10 |
| vdrok | good night everyone! | 18:14 |
| fellypefca | Good night vdrok | 18:14 |
| nicodemos | night vdrok | 18:17 |
| *** Sukhdev has joined #openstack-ironic | 18:23 | |
| *** dprince has joined #openstack-ironic | 18:25 | |
| *** markguz_ has joined #openstack-ironic | 18:26 | |
| *** Sukhdev_ has joined #openstack-ironic | 18:28 | |
| *** tesseract has quit IRC | 18:28 | |
| *** Sukhdev_ has quit IRC | 18:29 | |
| *** markguz has quit IRC | 18:29 | |
| *** Sukhdev has quit IRC | 18:29 | |
| *** bnemec has quit IRC | 18:35 | |
| *** Sukhdev has joined #openstack-ironic | 18:35 | |
| *** rama_y has joined #openstack-ironic | 18:36 | |
| *** markguz_ has quit IRC | 18:41 | |
| *** rwsu has quit IRC | 18:42 | |
| *** markguz has joined #openstack-ironic | 18:42 | |
| *** rama_y has quit IRC | 18:43 | |
| *** e0ne has joined #openstack-ironic | 18:44 | |
| leitan | i dont image what guys with 300gb disk wait for the cleaning process | 18:44 |
| * leitan waiting 45 minutes from test to test | 18:44 | |
| *** parlos has quit IRC | 18:46 | |
| dtantsur | leitan: for testing, I usually switch to metadata cleaning only | 18:47 |
| leitan | well well well helloooo, rushing to check that configgg | 18:48 |
| *** mgoddard has joined #openstack-ironic | 18:48 | |
| sambetts | yeah its in the ironic.conf to disable full erase | 18:48 |
| * dtantsur learned about http://superuser.openstack.org/articles/deploying-ironic-openstack-tripleo/ in the meantime | 18:48 | |
| sambetts | dtantsur: does that invole having a separate pool of equipment to enroll into the overcloud? | 18:50 |
| sambetts | dtantsur: or can the undercloud's pool and overcloud's pool be the same? | 18:50 |
| sambetts | first of those is "easy" second of those not so much | 18:50 |
| dtantsur | sambetts: depending on your definition of "pool". node list does not overlap, if that's what you ask | 18:50 |
| dtantsur | I guess the easy one | 18:50 |
| * sambetts has been working on some concepts for an overcloud/undercloud model where the baremetals can either be deployed by the undercloud as new computes or by the overcloud as baremetals | 18:51 | |
| sambetts | its starts to mess up your brain | 18:52 |
| openstackgerrit | Dmitry Tantsur proposed openstack/python-ironicclient master: Switch the deprecated "ironic" CLI to "latest" API version by default https://review.openstack.org/515064 | 18:52 |
| leitan | erase_devices_priority = 0 | 18:52 |
| leitan | erase_devices_metadata_priority = 10 | 18:52 |
| leitan | dtantsur sambetts for what im seeing settings that ones | 18:52 |
| leitan | should be enough | 18:52 |
| leitan | on the [deploy] section | 18:53 |
| dtantsur | sambetts: oh yeah, that would be fun | 18:53 |
| sambetts | I believe thats the right settings yup | 18:53 |
| dtantsur | leitan: looks good | 18:53 |
| dtantsur | time to go, see you tomorrow | 18:53 |
| sambetts | o/ dtantsur | 18:53 |
| leitan | have a good one | 18:53 |
| *** dtantsur is now known as dtantsur|afk | 18:53 | |
| *** fellypefca has quit IRC | 18:55 | |
| *** markguz_ has joined #openstack-ironic | 18:58 | |
| *** ricardoas has quit IRC | 19:00 | |
| * sambetts --> home | 19:01 | |
| sambetts | night all o/ | 19:01 |
| *** markguz has quit IRC | 19:01 | |
| *** sambetts is now known as sambetts|afk | 19:02 | |
| leitan | sambetts|afk: night and thanks ! | 19:02 |
| rloo | jlvillal: let me know if you want to discuss https://review.openstack.org/#/c/515225/. I looked at it again, and I feel like the original is easier to understand/maintain. but i could be biased since I reviewed the orig, and maybe don't have an open-enuf-mind to change my thinking to your new way. | 19:02 |
| patchbot | patch 515225 - python-ironicclient - Simplify logic for API version '1' and improve uni... | 19:02 |
| *** openstackgerrit has quit IRC | 19:03 | |
| *** bnemec has joined #openstack-ironic | 19:03 | |
| *** markguz_ has quit IRC | 19:03 | |
| *** markguz has joined #openstack-ironic | 19:04 | |
| *** efried_rollin is now known as efried | 19:07 | |
| *** markguz_ has joined #openstack-ironic | 19:08 | |
| *** markguz has quit IRC | 19:08 | |
| *** crushil has quit IRC | 19:11 | |
| *** crushil has joined #openstack-ironic | 19:11 | |
| *** catintheroof has quit IRC | 19:16 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 19:18 | |
| *** catintheroof has joined #openstack-ironic | 19:18 | |
| *** markguz_ has quit IRC | 19:20 | |
| *** markguz has joined #openstack-ironic | 19:20 | |
| *** ijw has quit IRC | 19:21 | |
| *** AlexeyAbashkin has quit IRC | 19:22 | |
| *** etingof has quit IRC | 19:24 | |
| *** Goneri has quit IRC | 19:25 | |
| *** Goneri has joined #openstack-ironic | 19:26 | |
| *** jcoufal_ has joined #openstack-ironic | 19:27 | |
| *** jcoufal has quit IRC | 19:30 | |
| *** jcoufal has joined #openstack-ironic | 19:31 | |
| *** jcoufal_ has quit IRC | 19:33 | |
| jlvillal | rloo: Sure, I'd be happy to discuss the patch | 19:34 |
| rloo | jlvillal: ok, let me know what you think after you read my comments. | 19:35 |
| *** ijw has joined #openstack-ironic | 19:36 | |
| *** mgoddard has quit IRC | 19:36 | |
| *** adreznec has joined #openstack-ironic | 19:36 | |
| jlvillal | rloo: Okay I have read the comments :) | 19:38 |
| rloo | jlvillal: thoughts? | 19:39 |
| jlvillal | rloo: Well on the when we go to "2" thing. I'm kind of like, fix it when that happens. | 19:39 |
| jlvillal | rloo: There will need to be quite a few changes for that. | 19:39 |
| rloo | jlvillal: ewww | 19:39 |
| jlvillal | rloo: A LOT of changes. | 19:39 |
| rloo | jlvillal: really? a lot of changes? | 19:39 |
| jlvillal | rloo: Well like line 31 for example | 19:40 |
| rloo | jlvillal: seems like it would be in that one place, where we change 1 to 1.something and 2 to 2.latestversion | 19:40 |
| rloo | jlvillal: yeah, we'd need to add support for 2.x | 19:40 |
| jlvillal | rloo: And of course all of the code changes for the new API. | 19:40 |
| rloo | jlvillal: that is regardless of this code | 19:40 |
| rloo | jlvillal: i feel that your change ties 1 with 'latest' too tightly | 19:41 |
| jlvillal | rloo: And the changes to my patch would be relatively minor. | 19:41 |
| rloo | jlvillal: minor, but i suspect in two diff places. | 19:41 |
| jlvillal | rloo: Yeah, just change '1' to '2' and then figure out what '1' should be. | 19:41 |
| jlvillal | rloo: Well my follow-on patch: https://review.openstack.org/#/c/515459/5 makes it in 1 place | 19:41 |
| patchbot | patch 515459 - python-ironicclient - Refactor check for 'latest' version | 19:41 |
| jlvillal | rloo: I could combine those. But I was trying not to do too much in one patch. | 19:42 |
| jlvillal | rloo: Follow on patch: https://review.openstack.org/#/c/515459/5/ironicclient/osc/plugin.py | 19:42 |
| patchbot | patch 515459 - python-ironicclient - Refactor check for 'latest' version | 19:42 |
| jlvillal | rloo: There is one function that takes care of determining latest | 19:42 |
| rloo | jlvillal: with that latest patch, where would you put code to convert '1' to '1.x', after we had '2'? | 19:43 |
| *** crushil has quit IRC | 19:44 | |
| jlvillal | rloo: I'm not sure. Either rename that function to something more generic. _convert_special_versions() or a new function? | 19:44 |
| jlvillal | rloo: Off the top of my head. | 19:44 |
| rloo | jlvillal: i can try reviewing this later, in a week or two or three, but right now, it doesn't sit well with me. Maybe someone else can +2 it for you. If it had been an easy/obvious thing, I'd be good with it, but I feel like there's a lot of patches that need reviewing, and I shouldn't spend more time on this. | 19:46 |
| jlvillal | rloo: Okay | 19:46 |
| *** bnemec has quit IRC | 19:53 | |
| *** makowals has joined #openstack-ironic | 19:55 | |
| *** rwsu has joined #openstack-ironic | 19:56 | |
| *** crushil has joined #openstack-ironic | 19:56 | |
| *** Goneri has quit IRC | 20:14 | |
| *** jcoufal_ has joined #openstack-ironic | 20:18 | |
| *** jcoufal__ has joined #openstack-ironic | 20:20 | |
| *** jcoufal has quit IRC | 20:20 | |
| *** jcoufal_ has quit IRC | 20:22 | |
| *** rwsu has quit IRC | 20:23 | |
| *** dprince has quit IRC | 20:26 | |
| *** jcoufal__ has quit IRC | 20:27 | |
| *** Goneri has joined #openstack-ironic | 20:30 | |
| *** bnemec has joined #openstack-ironic | 20:35 | |
| *** openstackgerrit has joined #openstack-ironic | 20:35 | |
| openstackgerrit | Merged openstack/ironic master: Migrate python-oneviewclient validations to Ironic OneView drivers https://review.openstack.org/468428 | 20:35 |
| *** rwsu has joined #openstack-ironic | 20:42 | |
| *** adreznec has quit IRC | 20:51 | |
| *** windsn has joined #openstack-ironic | 20:52 | |
| *** makowals has quit IRC | 20:53 | |
| *** adreznec has joined #openstack-ironic | 20:53 | |
| *** derekh has quit IRC | 20:59 | |
| *** mat128 has quit IRC | 21:00 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient master: Simplify logic for API version '1' and improve unit tests https://review.openstack.org/515225 | 21:05 |
| *** trown is now known as trown|outtypewww | 21:08 | |
| *** crushil has quit IRC | 21:08 | |
| *** priteau has quit IRC | 21:13 | |
| *** bfournie has quit IRC | 21:13 | |
| *** catintheroof has quit IRC | 21:16 | |
| *** adreznec has quit IRC | 21:20 | |
| *** leitan has quit IRC | 21:21 | |
| *** adreznec has joined #openstack-ironic | 21:22 | |
| openstackgerrit | Anup Navare proposed openstack/ironic-python-agent master: [Test]DNM Testing if TinyIPA builds with py3 https://review.openstack.org/509652 | 21:25 |
| *** crushil has joined #openstack-ironic | 21:29 | |
| *** e0ne has quit IRC | 21:29 | |
| *** adreznec has quit IRC | 21:31 | |
| *** adreznec has joined #openstack-ironic | 21:32 | |
| *** sasha22 has joined #openstack-ironic | 21:35 | |
| *** sasha21 has quit IRC | 21:35 | |
| *** rcernin has joined #openstack-ironic | 21:42 | |
| *** crushil has quit IRC | 21:46 | |
| *** Sukhdev has quit IRC | 21:51 | |
| *** adreznec has quit IRC | 21:55 | |
| *** etingof has joined #openstack-ironic | 21:55 | |
| *** adreznec has joined #openstack-ironic | 21:57 | |
| *** rwsu has quit IRC | 21:57 | |
| *** etingof has quit IRC | 22:00 | |
| *** rpioso is now known as rpioso|afk | 22:02 | |
| *** windsn has quit IRC | 22:13 | |
| *** jlabarre has quit IRC | 22:15 | |
| *** baoli has quit IRC | 22:19 | |
| openstackgerrit | Merged openstack/ironic master: Add ability to provide configdrive when rebuilding https://review.openstack.org/501761 | 22:27 |
| *** markguz has quit IRC | 22:36 | |
| *** markguz has joined #openstack-ironic | 22:37 | |
| *** markguz has quit IRC | 22:41 | |
| *** ijw has quit IRC | 22:50 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient master: Simplify logic for API version '1' and improve unit tests https://review.openstack.org/515225 | 22:54 |
| *** bfournie has joined #openstack-ironic | 22:55 | |
| *** adreznec has quit IRC | 22:57 | |
| *** adreznec has joined #openstack-ironic | 22:59 | |
| *** 5EXAAPFFA has joined #openstack-ironic | 23:01 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient master: Simplify logic for API version '1' and improve unit tests https://review.openstack.org/515225 | 23:07 |
| *** bfournie has quit IRC | 23:11 | |
| *** charcol has joined #openstack-ironic | 23:15 | |
| *** mikal has quit IRC | 23:28 | |
| *** bfournie has joined #openstack-ironic | 23:43 | |
| *** 5EXAAPFFA has quit IRC | 23:50 | |
| *** jaypipes has quit IRC | 23:54 | |
| *** links has joined #openstack-ironic | 23:56 | |
| openstackgerrit | Merged openstack/ironic master: Ensure ping actually succed https://review.openstack.org/516656 | 23:56 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
