| meena | somebody please unstale this: https://github.com/canonical/cloud-init/pull/2003 | 00:33 |
|---|---|---|
| -ubottu:#cloud-init- Pull 2003 in canonical/cloud-init "move get_ib_* functions to cloudinit.distros.networking" [Open] | 00:33 | |
| holmanb | I gotchu meena | 03:37 |
| holmanb | meena: re bug 269767: for the title might want to strike mention of LXC and mention instead that lxd is using qemu as the emulator | 03:43 |
| -ubottu:#cloud-init- Bug 269767 in usb-creator (Ubuntu) "Depends: syslinux" [Undecided, Fix Released] https://launchpad.net/bugs/269767 | 03:43 | |
| holmanb | bad ubottu | 03:43 |
| holmanb | I was refering to the freebsd bug, not whatever ubottu dragged up | 03:55 |
| cpaelzer | holmanb: hehe, bug numbers are techs "They-Who-Must-Not-Be-Named" :-P | 06:11 |
| meena | holmanb: I got six other FreeBSD VMs running just fine under qemu on that machine | 07:07 |
| meena | holmanb: https://github.com/freebsd/freebsd-src/pull/664 i got a weak start on a fix, but had no time for testing yet | 12:32 |
| -ubottu:#cloud-init- Pull 664 in freebsd/freebsd-src "apic: prevent divide by zero in CPU frequency init" [Open] | 12:32 | |
| ananke | hah. I spent couple hours beating my head agains the wall trying to figure out why a certain program would run just fine from cron or directly, but it would just exit without a single error when executed from a per-boot script | 13:30 |
| ananke | come to find out, it expected HOME env variable to exist | 13:30 |
| meena | ananke: I think somebody in here is working on a cc_env module | 14:21 |
| meena | https://github.com/canonical/cloud-init/pull/1965#issuecomment-1387581108 | 14:23 |
| -ubottu:#cloud-init- Pull 1965 in canonical/cloud-init "add proxy support for ansible galaxy" [Open] | 14:23 | |
| ananke | ohh, that plugin would be nice, since all of our systems are behind proxies, and configuring it everywhere is a bane of my existance | 14:28 |
| ananke | for this particular issue I'm just going to set HOME locally in the script, but it was such a pain to figure out what was going on. I was saving strace log files trying to find out why it would just exit | 14:29 |
| ananke | it was xfreerdp client btw | 14:30 |
| ananke | on the subject of environment variables, they can be such a pain. For example, we've found out that that various distros have limit on the length of variables that pam_env will accept from /etc/environment | 14:33 |
| ananke | for some distros it was 1024 bytes, for some it was 256 | 14:33 |
| ananke | we found out, because our NO_PROXY is about 1050 bytes | 14:33 |
| meena | ananke: oooooooooof. | 15:09 |
| meena | so what happened? did it truncate? or reject? | 15:09 |
| ananke | meena: it truncates it. | 15:30 |
| ananke | I forget which ones truncate to what length, but both debian and ubuntu are affected. | 15:31 |
| ananke | it was disappointing to see that, to say the least. we had to pivot from being able to rely on /etc/environment to having additional script in /etc/profile.d/ | 15:35 |
| waldi | i would say, everyone chooses their own problems. maybe reducing the size of NO_PROXY by using wildcards would be a good idea. but what do i know | 15:42 |
| ananke | waldi: NO_PROXY doesn't support wildcards | 15:56 |
| ananke | the whole proxy setup via env variables is a huge mess in the linux world. different rules govern no_proxy than http/s/ftp/all_proxy | 15:57 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!