| *** pmannidi has joined #openstack-dib | 00:28 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move blockdevicesetupexception.py into blockdevice.py https://review.openstack.org/461633 | 00:45 |
|---|---|---|
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: block_device: reorder imports https://review.openstack.org/461632 | 00:45 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 00:45 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 00:53 |
| openstackgerrit | Merged openstack/diskimage-builder master: exec_sudo: check cmd for str, log output and raise exception https://review.openstack.org/463427 | 01:25 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Remove unused val_else_none https://review.openstack.org/461634 | 01:53 |
| openstackgerrit | Merged openstack/diskimage-builder master: Remove unused val_else_none https://review.openstack.org/461634 | 03:24 |
| *** aparnav has joined #openstack-dib | 04:13 | |
| *** andreas-f has joined #openstack-dib | 04:27 | |
| *** jamielennox has quit IRC | 04:34 | |
| *** jamielennox has joined #openstack-dib | 04:34 | |
| openstackgerrit | Merged openstack/diskimage-builder master: block_device: reorder imports https://review.openstack.org/461632 | 04:53 |
| *** chhavi has joined #openstack-dib | 05:00 | |
| *** pmannidi has quit IRC | 05:07 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move dib-block-device implementation into cmd.py https://review.openstack.org/461635 | 05:18 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 05:18 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move YAML parsing into cmd.py; default to env https://review.openstack.org/461636 | 05:23 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move blockdevicesetupexception.py into blockdevice.py https://review.openstack.org/461633 | 05:23 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move dib-block-device implementation into cmd.py https://review.openstack.org/461635 | 05:23 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 05:23 |
| *** pmannidi has joined #openstack-dib | 05:24 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Create BlockDeviceCmd object https://review.openstack.org/461638 | 05:49 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 05:49 |
| yolanda | hi ianw , good morning | 05:50 |
| ianw | yolanda: hey | 05:50 |
| ianw | so after you guys had a look at those patches, i've been swizzling, squashing and putting several of them underneath the bd refactor | 05:51 |
| yolanda | so you need to abandon some of them? | 05:52 |
| yolanda | about the suse failure, see our comments above, is due to a bad call to dib-block-device, because it's not getting it inside the chroot | 05:53 |
| openstackgerrit | zgdeng proposed openstack/diskimage-builder master: WIP: Add zip element as s390 architecture bootloader https://review.openstack.org/443548 | 05:53 |
| ianw | yolanda: things are merging | 05:56 |
| *** appu103 has joined #openstack-dib | 05:56 | |
| ianw | sorry it's taken much longer to merge than i'd hoped | 05:58 |
| *** aparnav has quit IRC | 05:59 | |
| ianw | yolanda: i think the most practical thing to do will be to just pass env vars in | 06:00 |
| ianw | i mean, i thought we could do something like install diskimage_builder inside the chroot, and have dib-python find it from sys.path | 06:00 |
| ianw | i don't know if we want to do that | 06:01 |
| yolanda | i need to have this available as well, for enumerating the mount points in my patch | 06:01 |
| yolanda | https://review.openstack.org/447076 | 06:01 |
| ianw | yeah ... or some sort of rpc mechanism? | 06:03 |
| ianw | one of the bind-mounts could have a socket, we have a listener ... | 06:04 |
| ianw | urgh | 06:04 |
| ianw | i think it would be best to work on this as a follow-on to the bd refactor | 06:06 |
| ianw | even if we squash the changes in after, we should consider the solution separately i think | 06:07 |
| *** pmannidi has quit IRC | 06:07 | |
| yolanda | yes, it's a complex one | 06:19 |
| *** appu103 has quit IRC | 06:20 | |
| *** appu103 has joined #openstack-dib | 06:20 | |
| *** pmannidi has joined #openstack-dib | 06:21 | |
| *** appu103 has quit IRC | 06:38 | |
| *** appu103 has joined #openstack-dib | 06:38 | |
| openstackgerrit | Merged openstack/diskimage-builder master: Move blockdevicesetupexception.py into blockdevice.py https://review.openstack.org/461633 | 06:52 |
| openstackgerrit | Merged openstack/diskimage-builder master: Move dib-block-device implementation into cmd.py https://review.openstack.org/461635 | 07:04 |
| *** appu103 has quit IRC | 07:22 | |
| openstackgerrit | Merged openstack/diskimage-builder master: Move YAML parsing into cmd.py; default to env https://review.openstack.org/461636 | 07:26 |
| openstackgerrit | Merged openstack/diskimage-builder master: Create BlockDeviceCmd object https://review.openstack.org/461638 | 07:27 |
| yolanda | things are moving | 07:30 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Take --params from environment https://review.openstack.org/461637 | 07:39 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move to subparsers https://review.openstack.org/461639 | 07:39 |
| ianw | yolanda: ^ those two will conflict with the bd refactor. i'll rebase it on them. both were approved last night, as follow-ons | 07:40 |
| yolanda | so main issue now seems to be the usage of dib-block-device command | 07:44 |
| yolanda | maybe we can find some workaround for suse, then work on it later | 07:44 |
| ianw | yolanda: i'd support just using env variables at this point? | 07:48 |
| ianw | a comma separated list? | 07:48 |
| yolanda | ianw, not sure if i follow you. You mean exporting the generated vars? | 07:50 |
| ianw | in https://review.openstack.org/#/c/444586/52/diskimage_builder/elements/zypper/post-install.d/10-mkinitrd | 07:52 |
| ianw | isn't FS_TYPE set? or we can set it outside? | 07:53 |
| ianw | quite honestly i have no idea what's happening in there :) | 07:53 |
| yolanda | so, that's also one of the problems i reported in one of my comments | 07:53 |
| yolanda | FS_TYPE still seems to be used and needs to be exported | 07:54 |
| yolanda | but then it's conflicting with the mkfs type option | 07:54 |
| yolanda | so you could set mkfs-> type=xfs for example, but if FS_TYPE is not set to xfs, it still continues picking ext4 | 07:54 |
| ianw | but FS_TYPE doesn't make sense in a multiple partition world? | 07:54 |
| yolanda | indeed | 07:54 |
| ianw | sort of what i got to in https://review.openstack.org/463243 | 07:55 |
| yolanda | also see my comments on https://review.openstack.org/#/c/444586/, patchset 36 | 07:56 |
| ianw | why, xfs doesn't understand "-m"? | 07:57 |
| yolanda | it fails on older versions, works with the latest updated package on centos7 | 07:57 |
| yolanda | but i mean, in that situation, it was trying to set optiosn for ext4 because i didn't export FS_TYPE=xfs, although in my block-device.yaml i was setting / to xfs | 07:58 |
| ianw | ooohhh | 07:58 |
| ianw | yolanda: so that should just be gated? | 08:00 |
| ianw | "if DIB_BLOCK_DEVICE_CONFIG; then don't set fs options at all from FS_TYPE" | 08:00 |
| ianw | i think we might actually need the checking like in https://review.openstack.org/#/c/463243/ | 08:01 |
| yolanda | but then we need a way to extract the FS_TYPE that was set on root, then it's where dib-block-device is needed | 08:01 |
| ianw | right, but we could just do that, and set DIB_ROOT_FSTYPE | 08:01 |
| ianw | or something ... even if it's hackish | 08:01 |
| ianw | as long as it's documented ... | 08:02 |
| ianw | and, we could replace it with a better more generic way for elements to get info inside the chroot later on | 08:02 |
| yolanda | that will work for our need. At the point we need to use FS_TYPE on the image create, it's outside the chroot, so dib-block-device shall work | 08:03 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 08:11 |
| openstackgerrit | Andreas Florath proposed openstack/diskimage-builder master: Make python entry points accessible from elements https://review.openstack.org/463934 | 08:22 |
| andreas-f | This can be the base for the fix for suse (and later on for bootloader refactoring). | 08:24 |
| andreas-f | But I'm not sure about... Some hackish things were needed here. | 08:25 |
| andreas-f | waiting for the CI results. | 08:27 |
| *** hashar has joined #openstack-dib | 08:29 | |
| openstackgerrit | Andreas Florath proposed openstack/diskimage-builder master: Make python entry points accessible from elements https://review.openstack.org/463934 | 08:31 |
| yolanda | andreas-f, i'm also trying to collect FS_TYPE from the block device instead of the exported FS_TYPE var, but still WIP | 08:42 |
| ianw | andreas-f: that is ... quite a hack | 08:43 |
| ianw | i can't see it will work without a source tree? | 08:43 |
| ianw | where's the pip install coming from? | 08:43 |
| ianw | and it will require all sorts of dependencies inside the chroot | 08:44 |
| andreas-f | Ups: your are right: a source tree is needed. | 08:44 |
| andreas-f | Do you have any other idea? | 08:44 |
| andreas-f | (IMHO just copying things around will not work.) | 08:44 |
| ianw | no | 08:45 |
| ianw | i feel like maybe some sort of socket rpc mechanism might work | 08:45 |
| ianw | if dib-block-device was setup to listen as a daemon on a common socket before the chroot | 08:45 |
| ianw | inside the chroot, we could make a simple request/response | 08:46 |
| andreas-f | Yes - that would be a possibility. | 08:47 |
| ianw | but, i also think just env vars we need for now is probably the KISS approach | 08:47 |
| andreas-f | But would need major changes / extensions for the dib-block-device.. | 08:47 |
| ianw | we have so much new stuff going on, writing daemons etc is not worth it at this point | 08:47 |
| andreas-f | Yes - env vars would work for the suse problem - But I'm not convinced that this is enough for the bootstrap. | 08:48 |
| *** aparnav has joined #openstack-dib | 08:51 | |
| *** isaacb has joined #openstack-dib | 08:52 | |
| ianw | i think we want to keep dib dependencies running in the chroot to a minimum | 08:53 |
| ianw | see greg's work on busybox systems etc | 08:53 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Clear __init__.py from cmd move https://review.openstack.org/463943 | 08:53 |
| yolanda | mm, is root-fstype not a valid symbol? | 08:53 |
| yolanda | mm, seems that no... i'll work a bit more on this | 08:55 |
| *** appu103 has joined #openstack-dib | 09:04 | |
| *** hashar has quit IRC | 09:07 | |
| *** aparnav has quit IRC | 09:07 | |
| yolanda | andreas-f, so i'm thinking in implementing a root-fs-type symbol in getval, and that could parse the mkfs->type from the config yaml. We could use it to get FS_TYPE instead of relying on the env var, what do you think? | 09:18 |
| yolanda | will not fix the suse issue, because it's on chroot, but will fix the issues i have with the disk-image-create and the MKFS_OPTIONS | 09:19 |
| *** appu103 has quit IRC | 09:25 | |
| *** aparnav has joined #openstack-dib | 09:25 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 09:30 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 09:31 |
| openstackgerrit | Merged openstack/diskimage-builder master: Take --params from environment https://review.openstack.org/461637 | 09:37 |
| *** brault|away is now known as brault | 09:40 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 09:42 |
| yolanda | mm, this seems to have worked, but i'm getting a different error now on unmount | 09:55 |
| yolanda | i can create xfs images properly now, without having to export the FS_TYPE | 10:07 |
| yolanda | and didn't give any error, don't know why the IBM test failed | 10:08 |
| ianw | yolanda: the ibm test is failing always | 10:28 |
| ianw | we need to fix up the bootloader, but i'm waiting till we get this in to work on it! :) | 10:29 |
| yolanda | ah good to know then | 10:30 |
| yolanda | but i remember having seen +1 from that test, is a new issue? | 10:30 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 10:34 |
| openstackgerrit | Merged openstack/diskimage-builder master: Clear __init__.py from cmd move https://review.openstack.org/463943 | 10:35 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move to subparsers https://review.openstack.org/461639 | 10:50 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 10:55 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move to subparsers https://review.openstack.org/461639 | 11:04 |
| *** aparnav has quit IRC | 11:20 | |
| yolanda | i see the green in the tests finally! | 11:35 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Apply setfiles on all mountpoints https://review.openstack.org/447076 | 11:40 |
| *** LiCeUser1 has joined #openstack-dib | 11:43 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Apply setfiles on all mountpoints https://review.openstack.org/447076 | 11:44 |
| *** LiCeUser1 is now known as abarbaneld | 11:59 | |
| *** abarbaneld has left #openstack-dib | 12:00 | |
| *** abarbaneld has joined #openstack-dib | 12:00 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Apply setfiles on all mountpoints https://review.openstack.org/447076 | 12:34 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Add lvm management to diskimage-builder https://review.openstack.org/444403 | 12:44 |
| *** chhavi_ has joined #openstack-dib | 13:10 | |
| *** chhavi has quit IRC | 13:12 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Add lvm management to diskimage-builder https://review.openstack.org/444403 | 13:16 |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Apply setfiles on all mountpoints https://review.openstack.org/447076 | 14:08 |
| *** yolanda has quit IRC | 14:14 | |
| *** yolanda has joined #openstack-dib | 14:16 | |
| *** isaacb has quit IRC | 14:56 | |
| *** isaacb has joined #openstack-dib | 15:16 | |
| *** chhavi_ has quit IRC | 15:37 | |
| *** isaacb has quit IRC | 16:09 | |
| *** chhavi_ has joined #openstack-dib | 16:24 | |
| *** andreas-f has quit IRC | 19:01 | |
| *** chhavi_ has quit IRC | 19:06 | |
| openstackgerrit | yolanda.robla proposed openstack/diskimage-builder master: Add lvm management to diskimage-builder https://review.openstack.org/444403 | 19:37 |
| *** andreas-f has joined #openstack-dib | 20:00 | |
| *** abarbaneld has quit IRC | 21:22 | |
| *** jamielennox is now known as jamielennox|away | 22:42 | |
| openstackgerrit | Merged openstack/diskimage-builder master: Move to subparsers https://review.openstack.org/461639 | 23:02 |
| *** jamielennox|away is now known as jamielennox | 23:22 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Remove args from BlockDevice() init https://review.openstack.org/461640 | 23:36 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Move config out of parameter YAML file https://review.openstack.org/463242 | 23:40 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Refactor: block-device filesystem creation, mount and fstab https://review.openstack.org/444586 | 23:44 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!