=== bahama- is now known as bahamat === bahamat is now known as bahama- === nikolay_ is now known as hjckr === nikolay_ is now known as hjckr === stw__ is now known as stw === bahama- is now known as bahamat [16:13] hope smoser doesn't think this review https://github.com/canonical/cloud-init/pull/2043#pullrequestreview-1320148515 was not-nice [16:13] -ubottu:#cloud-init- Pull 2043 in canonical/cloud-init "send dhcp-client-identifier with dhclient command for InfiniBand ports" [Open] [18:51] i didn't think it was not nice; And I, for one, welcome our infiniband overlords [18:57] InfiniBand is great. Some of our InfiniBand code, however.... [18:58] https://github.com/canonical/cloud-init/blob/5d1d2544ab1f070ab9810779ba914298d44c06b2/cloudinit/net/__init__.py#L908 [18:59] holmanb: we have maybe 120 lines of infiniband related code, and i hate every single character in it [20:12] heh [20:18] holmanb: ah, the is_virtual stuff is being reverted? I guess I can stop writing testcases for it on Alpine lol [20:20] minimal: i guess that can stay, but the current user is being reverted for sure [20:21] meena: well the #2045 PR is removing your is_virtual for FreeBSD [20:21] I worked so hard on that ;( [20:22] (on outsourcing the fixing of tests of that pr) [20:23] also looks like in general is only a single DS specified (or DS + 'None') then that DS is assumed (so no 'None' fallback?) [20:24] meena: yeah same for me, writing is_virtual for Alpine was straightforward but I've spent ages on getting tests for it working lol [20:29] minimal: meena: that's the proposal as it stands now, with this fix we won't need those functions [20:31] holmanb: what's get_proc_env(1) do? [20:33] eh, I didn't write that stuff [20:33] looks like https://github.com/canonical/cloud-init/blob/main/cloudinit/util.py#L2374 [20:34] meena: actually, wait - that's a funky way to get env vars in python [22:26] nah, that's a funky way of getting envvars in Linux [22:27] most other Unices have gotten rid of /proc, because it's an unholy mess [22:27] but also, what kind of env does pid 1 have, anyway?? [22:49] meena: just to double check - was there any reason for is_virtual() besides openstack baremetal support? [22:50] holmanb: not thati know, but if there comes reason again, we can easily resurrect it [22:50] meena: I thought that the PR was prompted by the openstack PR, but if not I don't want to remove it [22:50] cool, thx