opendevreview | Takashi Kajinami proposed openstack/ironic master: Enforce a supported value for [deploy] image_server_auth_strategy https://review.opendev.org/c/openstack/ironic/+/905639 | 02:51 |
---|---|---|
opendevreview | Takashi Kajinami proposed openstack/ironic master: Validate [deploy] image_server_auth_strategy https://review.opendev.org/c/openstack/ironic/+/905639 | 02:52 |
opendevreview | Takashi Kajinami proposed openstack/ironic master: [deploy] image_server_password should be secret https://review.opendev.org/c/openstack/ironic/+/905640 | 02:56 |
opendevreview | Takashi Kajinami proposed openstack/ironic master: Validate [deploy] image_server_auth_strategy https://review.opendev.org/c/openstack/ironic/+/905639 | 03:01 |
opendevreview | Takashi Kajinami proposed openstack/ironic master: Validate [deploy] image_server_auth_strategy https://review.opendev.org/c/openstack/ironic/+/905639 | 03:38 |
tonyb | iurygregory: Are you done with: https://review.opendev.org/c/openstack/project-config/+/904012 Did you regain access to your pypi account? | 05:44 |
rpittau | good morning ironic! o/ | 08:09 |
opendevreview | Takashi Kajinami proposed openstack/ironic master: Validate [deploy] image_server_auth_strategy https://review.opendev.org/c/openstack/ironic/+/905639 | 09:46 |
iurygregory | tonyb, not yet, https://github.com/pypi/support/issues/2708 | 11:23 |
iurygregory | good morning Ironic | 11:23 |
opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: RedfishFirmwareInterface - Unit Tests & More logs https://review.opendev.org/c/openstack/ironic/+/903379 | 14:28 |
opendevreview | Julia Kreger proposed openstack/ironic master: Revert "RBAC: Fix allocation check" https://review.opendev.org/c/openstack/ironic/+/905539 | 15:11 |
*** Continuity__ is now known as Continuity | 15:48 | |
rpittau | good night! o/ | 17:08 |
JayF | o/ | 17:12 |
iurygregory | hey JayF o/ | 17:15 |
JayF | howdy | 17:15 |
JayF | I thought for a sec you might need something then I realized my goodnight to riccardo is my hello to the channel for the day too :D | 17:16 |
iurygregory | yeah =) | 17:17 |
iurygregory | doing good | 17:17 |
JayF | That's a tough thing to be doing back from vaca this short of time :D | 17:18 |
JayF | usually takes a while to get my motor running at lease | 17:18 |
JayF | *least | 17:18 |
iurygregory | yeah, I'm still a bit slow to remember my priorities lol | 17:18 |
iurygregory | 5 things on my list, "ok this seems easy, let me work on this task" (it wasn't) =P | 17:19 |
opendevreview | Jay Faulkner proposed openstack/ironic master: DNM: Testing against devstack perf enhancements https://review.opendev.org/c/openstack/ironic/+/905743 | 17:35 |
opendevreview | Merged openstack/ironic master: Revert "RBAC: Fix allocation check" https://review.opendev.org/c/openstack/ironic/+/905539 | 18:40 |
opendevreview | Merged openstack/ironic master: [deploy] image_server_password should be secret https://review.opendev.org/c/openstack/ironic/+/905640 | 18:52 |
opendevreview | Merged openstack/ironic master: docs: add distribution specific chain of trust warning around grub https://review.opendev.org/c/openstack/ironic/+/904597 | 18:52 |
opendevreview | Merged openstack/ironic-python-agent master: Add missing headers to the inspection callback https://review.opendev.org/c/openstack/ironic-python-agent/+/905126 | 19:17 |
opendevreview | Merged openstack/ironic-python-agent master: Support several API and Inspector URLs https://review.opendev.org/c/openstack/ironic-python-agent/+/903999 | 19:17 |
samcat116 | Question for those here deploying Ironic via Kolla-ansible. Are you running the ironic tftp/http services on the "internal" interface (where DB, MQ, and other internal APIs run) like it defaults to, or on some other interface. | 19:18 |
opendevreview | Julia Kreger proposed openstack/ironic master: Revert "Revert "RBAC: Fix allocation check"" to use Unauthorized https://review.opendev.org/c/openstack/ironic/+/905784 | 19:19 |
TheJulia | dtantsur: fixed^ | 19:19 |
JayF | samcat116: So you'll need the tftp/http services on the same interface that your "target" hosts are (e.g. the hosts Ironic will be deploying to) | 19:24 |
JayF | samcat116: what interface that is in a KA environment likely depends on how your environment is seutp | 19:24 |
samcat116 | Does that mean the same as the provisioning network? thats what I've done in the past, however in an OVN setup the port I have for the provisioning network is owned by OVN, so that interface can't have an IP on the controller | 19:27 |
wncslln | hello folks o/, anyone already deploy a baremetal with ironic and configure to use SGX instructions? | 19:36 |
JayF | I'm not sure how you'd get that setup in OVN; but I do know that the TFTP/HTTP servers need to be accessible from the target hosts. TheJulia has done some testing/documentation w/r/t running Ironic in OVN | 19:37 |
JayF | but there is some significant limitations I believe | 19:37 |
JayF | samcat116: assuming you've seen https://docs.openstack.org/ironic/latest/admin/ovn-networking.html | 19:37 |
samcat116 | yeah the OVN part of it is working just fine | 19:37 |
JayF | okay cool | 19:37 |
JayF | yeah, I just don't know the shape of it in OVN :/ sorry for not having more help for you | 19:38 |
samcat116 | Thats ok, it was more of an operational question for folks running ironic and OVN | 19:38 |
TheJulia | So the provisioning network just needs to be abel to reach the endpoint of services, nothing dynamically puts the services on a network, it is all rooted in the cloud's configuration | 19:38 |
TheJulia | Really, you would want the services on a separate network/interface and *not* internal ones | 19:39 |
samcat116 | It works if my provisioning network is routable to the aforementioned "internal" network, but I reaaaally don't want that to be routable | 19:39 |
TheJulia | because you want the provisioning network to be able to reach the resources | 19:39 |
TheJulia | yeah, I don't blame you | 19:39 |
TheJulia | you'd need some other network, really | 19:39 |
samcat116 | but the split is ironic's dnsmasq needs to listen on the interface for the provisioning network, but then http/tftp will on some other routable network | 19:40 |
samcat116 | I'm just trying to figure out why the Kolla-Ansible config defaults this way | 19:40 |
TheJulia | ironic's dnsmasq?! | 19:41 |
TheJulia | what is that dnsmasq serving? DHCP for the OVN network? | 19:42 |
samcat116 | sorry, Kolla-ansible spins up a dedicated dnsmasq container called "ironic_dnsmasq" for inspector to use | 19:42 |
TheJulia | okay, yeah, that is *entirely* separate, more so to support out of band introspection/hardware discovery use cases | 19:42 |
TheJulia | the model is you start by attaching the hardware to the provisioning network to start and then move the hardware from that network as time moves on | 19:42 |
TheJulia | with up to date OVN, as long as your not using IPv6, you can use the OVN dhcp service as well, just for extra context | 19:43 |
samcat116 | Yep im doing that just fine | 19:43 |
samcat116 | Its just figuring out what IPs the tftp/http server live on and what ovn dhcp therefore gives out | 19:44 |
samcat116 | I'd almost need two interfaces on each controller on the provisioning network, one for OVN to grab and one for tftp/http to listen on | 19:45 |
samcat116 | making it another network seems a bit silly | 19:45 |
TheJulia | for nodes being provisioned, ironic asserts that toneutron | 19:45 |
TheJulia | Or you want to be able to go "I want the tftp/http to be bound off this ovn network | 19:45 |
TheJulia | " | 19:46 |
samcat116 | yeah the second one. Basically say 192.168.1.0/24 is the provisioning network, there' | 19:46 |
samcat116 | there's not an easy way for tftp/http to listen on that if ovn owns the interface for that network on the controller | 19:46 |
TheJulia | Unfortunately we don't have an ability or tool to do that. We discussed that in part, but there was lots of mixed feelings in part because it requires carrying a lot of extra complexity when the address could just be a loopback on the host. | 19:51 |
samcat116 | Oh thats true I could put it on the loopback | 19:54 |
TheJulia | a loopback, you just want $something to be able to respond to the traffic someplace in a way which is routable | 19:54 |
TheJulia | as long as everything in the core of the networking knows how to reach $addresses | 19:54 |
opendevreview | Merged openstack/bifrost master: Uplift default Ansible version to 8.x https://review.opendev.org/c/openstack/bifrost/+/903950 | 20:04 |
opendevreview | Merged openstack/ironic stable/2023.1: Fix system scoped manageable node network failure https://review.opendev.org/c/openstack/ironic/+/905087 | 20:38 |
tonyb | iurygregory: Okay cool. I've "subscribed" to that issue so hopefully I wont need to pester you ;P | 20:55 |
iurygregory | tonyb, in any case, happy to merge the change upstream and we can open again when I need to delete the branch | 21:24 |
iurygregory | TheJulia, hey you still around? | 21:27 |
iurygregory | trying to understand your comment in https://review.opendev.org/c/openstack/ironic/+/903379/4/ironic/tests/unit/drivers/modules/redfish/test_firmware.py | 21:27 |
TheJulia | I am stepping f away for a little bit, be back in 15 minutes | 21:28 |
TheJulia | Err, stepping away for | 21:28 |
iurygregory | ack =) | 21:29 |
tonyb | iurygregory: It's fine as is I just didn't want it left that way if the need had passed | 21:47 |
iurygregory | tonyb, ack =) | 21:47 |
TheJulia | iurygregory: o/ | 21:47 |
TheJulia | iurygregory: I guess what I'm trying to get across, and what I suspecct dmitry is also trying to get across, is we shouldn't be expecting a sushy error to be raised up when calling task.driver.firmware.update() | 21:48 |
iurygregory | I understand, but at least in the test we have present test_missing_simple_update_action ( the exception that will happen when simple update service attempts to execute the action ) so the idea would be to add a try / catch a re-raise as RedfishError? | 21:53 |
iurygregory | https://review.opendev.org/c/openstack/ironic/+/903379/4/ironic/drivers/modules/redfish/firmware.py#179 https://review.opendev.org/c/openstack/ironic/+/903379/4/ironic/drivers/modules/redfish/firmware.py#209 | 21:53 |
TheJulia | Yeah, I think that is reasonable | 21:53 |
iurygregory | ack =) will change it | 21:54 |
TheJulia | Something that is *within* ironic's exceptions :) | 21:54 |
iurygregory | thanks for the help TheJulia o/ | 21:54 |
iurygregory | now time to go to the gym, get some dinner and be back to work in 3hrs =) | 21:55 |
JayF | I always think of it like this, and I have no idea if this is the right way to frame it | 21:56 |
JayF | stuff that goes in node history / last_error is sorta part of our API | 21:56 |
JayF | and should live in Ironic, not clients, because the clients can change independent of Ironic version | 21:56 |
opendevreview | Julia Kreger proposed openstack/ironic master: Disable legacy RBAC policy by default. https://review.opendev.org/c/openstack/ironic/+/902009 | 22:07 |
opendevreview | Merged openstack/sushy master: Handle a different related properties for missing TransferProtocolType https://review.opendev.org/c/openstack/sushy/+/904888 | 22:56 |
opendevreview | Merged openstack/ironic-python-agent-builder master: Switch IPA building to Debian Bookworm https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/903164 | 23:34 |
opendevreview | Merged openstack/ironic-python-agent-builder stable/2023.2: Fix building images for Bullseye and Bookworm https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/903550 | 23:35 |
opendevreview | Merged openstack/ironic-python-agent-builder master: Drop TripleO job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/904491 | 23:38 |
opendevreview | Julia Kreger proposed openstack/ironic master: Handle Power On/Off for child node cases https://review.opendev.org/c/openstack/ironic/+/896570 | 23:42 |
* TheJulia calls it a day | 23:47 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!