| iurygregory | good morning Ironic | 11:35 |
|---|---|---|
| opendevreview | Maximilian Brandt proposed openstack/ironic-python-agent master: Add bmc_subnet information to introspection https://review.opendev.org/c/openstack/ironic-python-agent/+/918340 | 12:46 |
| opendevreview | Maximilian Brandt proposed openstack/ironic-python-agent master: Add bmc_subnet information to introspection https://review.opendev.org/c/openstack/ironic-python-agent/+/918340 | 12:46 |
| TheJulia | good morning | 13:13 |
| opendevreview | cid proposed openstack/ironic master: Link MACs with PXE config to pxe_enabled ports https://review.opendev.org/c/openstack/ironic/+/921532 | 14:26 |
| shardy | TheJulia: Hi! When you get a moment I was wondering if we can sync up re https://review.opendev.org/c/openstack/ironic-specs/+/906324 | 14:46 |
| shardy | I'm wondering what the latest status is, and if there are any preferred workarounds (particularly for the config-2 collision aspect) | 14:47 |
| TheJulia | o/ shardy | 14:51 |
| TheJulia | we ended up just locking out the case | 14:51 |
| TheJulia | in the dib code, so the spec is kind of pointless once we found a clean, non-breaking way to test/check | 14:51 |
| TheJulia | let me get you a link so you can see | 14:51 |
| TheJulia | A few changes: https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/920184 && https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/915170 && https://review.opendev.org/q/I9b74ec977fabc0a7f8ed6f113595a3f1624f6ee6 | 14:54 |
| opendevreview | Julia Kreger proposed openstack/ironic master: fix: Fix class typo for portgroup. Portgroup instead of PortGroup https://review.opendev.org/c/openstack/ironic/+/921534 | 14:56 |
| TheJulia | fungi: ^ is why I was interested in actually using a commercial tool so many moons ago. | 14:56 |
| shardy | TheJulia: awesome! Thanks so much for the work on this, I'll rebase our environments to align with this solution :) | 14:59 |
| TheJulia | so, in theory, you could use cloud-init with a code change, fwiw | 15:00 |
| TheJulia | we explicitly now uninstall it was well | 15:00 |
| TheJulia | and depending on the OS image, just reinstalling doesn't work | 15:00 |
| shardy | Thanks for the information, we're not currently using cloud-init but I know that's relevant e.g to the Sylva community usage of Metal3 | 15:01 |
| TheJulia | ... it should be possible to swap out the automatic uninstall for a flag | 15:02 |
| TheJulia | we ahve yet to actually merge the test job for all of this, but we know from the test results it is quite happy | 15:02 |
| TheJulia | it coudl be swapped to use cloud-init, I know from early on it worked, but just explicitly uninstalling cloud-init was much more of a "avoid all possible conflicts" | 15:03 |
| TheJulia | change | 15:03 |
| fungi | TheJulia: i don't remember that discussion any more. some sort of tool which would check for missing object definitions via static analysis? i guess there aren't any open source ones which do that? | 15:06 |
| TheJulia | Yeah, basically commercial vulnerability and static analysis tooling found something which found things our existing python tools didn't | 15:10 |
| TheJulia | granted, the commercial stuff also found a TON of false positives/invalid items | 15:13 |
| opendevreview | cid proposed openstack/ironic master: Link MACs with PXE config to pxe_enabled ports https://review.opendev.org/c/openstack/ironic/+/921532 | 15:15 |
| opendevreview | cid proposed openstack/networking-generic-switch master: Add VLAN and Port allowlist https://review.opendev.org/c/openstack/networking-generic-switch/+/888047 | 15:57 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Remove unused parameter in ansible driver https://review.opendev.org/c/openstack/ironic/+/921540 | 16:07 |
| opendevreview | Julia Kreger proposed openstack/ironic master: minor defensive check change to drac driver https://review.opendev.org/c/openstack/ironic/+/921542 | 16:30 |
| -opendevstatus- NOTICE: Gerrit will be restarted at around 17:45 UTC to pick up some small image updates | 16:45 | |
| opendevreview | cid proposed openstack/ironic master: Link MACs with PXE config to pxe_enabled ports https://review.opendev.org/c/openstack/ironic/+/921532 | 17:20 |
| opendevreview | Julia Kreger proposed openstack/ironic master: minor: additional tests for interaction around node lookup https://review.opendev.org/c/openstack/ironic/+/921549 | 17:31 |
| opendevreview | Julia Kreger proposed openstack/ironic master: ci: fix https boot artifact path https://review.opendev.org/c/openstack/ironic/+/921576 | 20:35 |
| TheJulia | looks like one of the mirrors are down/gone | 20:44 |
| TheJulia | I guess we'll see | 20:44 |
| cid | Yea. Probably just down | 20:46 |
| cid | Happy weekend o// | 20:47 |
| clarkb | TheJulia: yes we took that mirror down in order to redeploy the cloud. We've always said people shouldn't hardcode our mirrors as if they are truly public mirrors | 20:49 |
| clarkb | it should come back (hopefully next week) but will haev a new name | 20:49 |
| clarkb | CI jobs can refer to ansible variables to know what their local cloud mirror is | 20:49 |
| clarkb | zuul_site_mirror_fqdn is the ansible var | 20:50 |
| TheJulia | I think we tried that in the past and found only one of the mirrors had the actual efi assets | 20:53 |
| TheJulia | but happy to give it a spin if they do | 20:54 |
| clarkb | all of our mirrors serve the same content. The content is in afs which is a ditributed filesystem so they all see the same content within a few seconds of each other aiui | 20:54 |
| clarkb | the mirrors also run proxy caches for some content (but the distro packages are proper mirrors on afs) and those may diverge over time particularly for systems like pypi which serves through CDN and you get different local results | 20:55 |
| TheJulia | Oh, I didn't realize the mirror stuff was on AFS too | 20:55 |
| clarkb | its the primary reason we haev AFS because it allows us to ensure consistency across all the regions (then we don't get weird "it worked here but not there" problems due to distro package state) and for deb repos we only publish to the volume mirrors consume after reprepro has completed syncing and everything is in place for a working mirror | 20:58 |
| clarkb | we could do that for the rpm repos too but they don't have a tool like reprepro to confirm consistency before publishign as far as I know | 20:58 |
| TheJulia | hmm | 21:11 |
| opendevreview | Julia Kreger proposed openstack/ironic master: ci: fix https boot artifact path https://review.opendev.org/c/openstack/ironic/+/921576 | 21:12 |
| opendevreview | Julia Kreger proposed openstack/ironic master: ci: fix https boot artifact path https://review.opendev.org/c/openstack/ironic/+/921576 | 21:41 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!