| === Ionic_ is now known as 32NABA83T | ||
| === 32NABA83T is now known as Ionic | ||
| === ePierre-afk is now known as ePierre | ||
| xnox | wgrant, that bit works; faking raw dasd track data -> no solution yet. And some z/vm is currently penciled as required. | 04:12 |
|---|---|---|
| xnox | wgrant, if not in launchpad, it would mean manually setting up and hooking up z/vms to jenkins environments. | 04:12 |
| xnox | pain all around. | 04:12 |
| wgrant | xnox: There has to be a way to fake that. | 04:12 |
| xnox | wgrant, i am escalating that question to ibm. | 04:13 |
| wgrant | xnox: How do we manage to produce an image if it's not possible to fake a DASD? | 04:17 |
| xnox | wgrant, we produce an image for virtio (scsi) at the moment. | 04:17 |
| xnox | wgrant, I have been creating dasd images by hand, in a z/vm out of bound for L1CC until it died. | 04:18 |
| xnox | wgrant, and now we are rescoping to produce dasd images proper. | 04:18 |
| wgrant | xnox: Is the problem just that loopback devices can't presently emulate the DASD ioctls? | 04:18 |
| xnox | wgrant, that. | 04:18 |
| xnox | wgrant, and there seemed to be no tool to ~= $ qemu-img convert -> into raw track data. But also pinging ibm if we can somehow not ship raw track data to them. | 04:19 |
| wgrant | xnox: The bootloader installer grows a feature that takes ECKD geometry as a file | 04:19 |
| wgrant | Or whatever | 04:19 |
| xnox | wgrant, i can install eckd bootloader, it has all flags to override everything. | 04:19 |
| wgrant | Ah, so it's just the image format? | 04:20 |
| xnox | it's just previously they wanted "dd of the /dev/dasda when said dasd is in raw track access mode, rather than normal mode" | 04:20 |
| xnox | wgrant, yeah. | 04:20 |
| wgrant | xnox: I wonder if that's just to capture the bootloader properly. | 04:21 |
| wgrant | Surely the precise ECKD layout of the ext4 volume doesn't matter. | 04:21 |
| xnox | wgrant, i wonder if i should pull out hexeditor... | 04:21 |
| xnox | wgrant, we do binary patch some of our cloud images to make them bootable.... | 04:21 |
| xnox | it's not like ibm is going to change ECKD format any day soon | 04:22 |
| wgrant | Heh, quite | 04:22 |
| wgrant | I think it's certainly worth examining the ECKD layout of the images that you build to see what weirdness exists. | 04:22 |
| wgrant | The chance that this can't be handled in userspace on a normal block device seems negligible. | 04:23 |
| wgrant | And I'm not going to develop new build scheduling technology just because nobody's spent the day to write that userspace conversion tool :) | 04:23 |
| xnox | wgrant, i was hoping to simply keep the devirt builders; and have a livefs build that is devirt and nothing else har har har but yeah, should be doable in userspace. | 04:38 |
| ricotz | cjwatson, hi, if you have a moment, could you adjust the size of this PPA to e.g. 32GB? https://launchpad.net/%7Emozillateam/+archive/ubuntu/firefox-next/+packages -- Thank you very much | 12:44 |
| cjwatson | ricotz: done | 12:45 |
| nottrobin | cjwatson: Do you know how do I do a pull/merge request in Launchpad Git? | 16:13 |
| nacc | nottrobin: clone a repo, make a change, push to a repo, submit an MP on the web UI | 16:13 |
| nottrobin | nacc: could you help me find the URL for that? I've been looking... I want to merge this branch into master https://git.launchpad.net/usn.ubuntu.com/log/?h=add-dockerfile | 16:14 |
| nacc | nottrobin: start here: https://code.launchpad.net/usn.ubuntu.com | 16:15 |
| nacc | click add-dockerfile | 16:15 |
| nacc | nottrobin: then propose for merging | 16:15 |
| nottrobin | ah thanks | 16:15 |
| nacc | nottrobin: the other UI is the repo, not the Laucnhpad view of the repo (if that makes sense) | 16:15 |
| nottrobin | yep makes sense. Thanks. | 16:16 |
| nacc | nottrobin: np | 16:16 |
| pivotal | Question regarding 3rd party PPA packages: is there a way for one PPA to depend on another PPA package, via the "Build Depends" of debian/control? We tried adding a PPA repo in the rules file, and it failed. | 22:59 |
| nacc | pivotal: the ppas need to depend on each other | 23:09 |
| nacc | pivotal: it's in the ppa definition (iirc) | 23:09 |
| nacc | pivotal: so when the package builds in the second ppa, it will use packages from the first ppa | 23:09 |
| nacc | pivotal: definitely not in the source package itself | 23:10 |
| wgrant | pivotal: As nacc said, dependencies on another PPA are a property of the dependent PPA, not of the source package. See the "Edit PPA dependencies" on the PPA page. | 23:16 |
| pivotal | Thanks wgrant. Could you share me the link for the one you just specified? | 23:25 |
| pivotal | Never mind. Found it. Thanks. | 23:26 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!