[15:22] general question, when raising a MR which fixes multiple Launchpad issues should multiple "LP: #NNNNN" lines be put in the PR description or a single line with comma separated numbers? [16:07] minimal: LP: #1858928, #1858929, #1858931, #1858932 [16:07] -ubottu:#cloud-init- Launchpad bug 1858932 in cloud-init "schema: add jsonschema definition to cc_yum_add_repo" [Low, Fix Released] https://launchpad.net/bugs/1858932 [16:07] -ubottu:#cloud-init- Launchpad bug 1858929 in cloud-init "schema: add jsonschema definition to cc_update_hostname" [Low, Fix Released] https://launchpad.net/bugs/1858929 [16:07] -ubottu:#cloud-init- Launchpad bug 1858931 in cloud-init "schema: add jsonschema definition to cc_write_files" [Low, Fix Released] https://launchpad.net/bugs/1858931 [16:07] -ubottu:#cloud-init- Launchpad bug 1858928 in cloud-init "schema: add jsonschema definition to cc_update_etc_hosts" [Low, Fix Released] https://launchpad.net/bugs/1858928 [16:07] oops sorry all [16:07] blame the bot ;) [16:07] ubottu: why be so noisy :) [16:08] blackboxsw: thanks [18:36] I don't suppose cloud-initramfs-growroot is part of cloud-init? [18:44] I'm tracking a bug in the latest Kali AMI, which involves missing flock. The issue in theory was solved, but somehow it resurfaces for me, when I apply all of the updates and make a new image. trying to figure out if it's something in our process that may be causing it, or Kali broke something [18:48] no, this is outside of cloud-init. and i think i know when it broke [18:49] I found a mention of it here: https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1860809.html , now trying to see if the bug fix is present or not [18:59] the hooks script present in cloud-initramfs-growroot is already the newest version (0.18.debian11). [18:59] oops, wrong paste [18:59] the hooks script present in cloud-initramfs-growroot 0.18.debian11 does not seem to contain the same stuff as https://git.launchpad.net/cloud-initramfs-tools/tree/growroot/hooks/growroot#n12 [19:01] ananke: wouldn't it be a case of adding a "dracut_install flock" to the growroot/dracut/modules.d/50growpart/install file? [19:04] minimal: possibly, I may ultimately do that to get us going. I'm trying to determine who's at fault first :) (me/us, kali, debian) [19:05] in case of kali/debian, there is no dracut though [19:05] ananke: well the other change you mentioned just checks if flock is present and gives and error if not, the change I'm suggesting (from a cursory glance) is to add flock to the initramfs, so avoid the error that it's not present ;-) [19:05] ok, well whatever manages the initramfs [19:05] initramfs-tools [19:06] minimal: are you referring to the bug url I posted? this seems to actually copy flock, rather than check for its presence [19:06] and sure enough, my initramfs lacks flock. I should check if their vanilla AMI has flock in initramfs [19:09] it doesn't have flock either, but resizes just fine. fun [19:58] ananke: I've never seen growpart run in an initramfs before [20:05] minimal: not sure if it's a new behavior or not, running out of initramfs that is [20:06] but it's certainly there, and has been for some time [20:11] still baffles me that their AMI works just fine on first boot, without that flock in initramfs. the problem likely happens with the update to cloud-initramfs-tools, their AMI comes with 0.18.debian8, but update sets it to 0.18.debian11 [20:12] and looks like there may be another issue, https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1027050 [20:12] -ubottu:#cloud-init- Debian bug 1027050 in cloud-initramfs-growroot "growing image fails to boot: /scripts/local-bottom/growroot: line 97: wait-for-root: not found" [Important, Open] [21:11] this whole thing is a hot mess. latest debian package also no longer includes hooks for sed/grep/rm, so growpart throws even more errors [21:46] smoser: I'd love your take on this issue, if you happen to be available sometime this week :)