Re: logging mount assembly and bootloader install, chroot missing?
On Sat, 2017-09-09 at 18:36 -0700, Adam Williamson wrote: > On Tue, 2017-08-08 at 12:22 -0600, Chris Murphy wrote: > > The program.log shows mount commands for assembly, followed by > > installation itself copying files to a path prefix /mnt/sysimage, > > and > > then bootloader setup. Isn't there a chroot happening before the > > bootloader setup? It's not in the log so I'm having some difficulty > > exactly replicating an installation environment while > > troubleshooting > > this bug: > > > > So the questions are: > > Is there a chroot? > > Is there a way to discover the command being used? > > And if there is a chroot, is there a way to log this command in the > > future? > > > > > > https://bugzilla.redhat.com/show_bug.cgi?id=1289752 > > Use the source, Luke: > > https://github.com/rhinstaller/anaconda/blob/master/pyanaconda/bootlo > ader.py > > and then search for 'root='. You'll find that most of the bootloader > installation-related commands run via iutil.execWithRedirect, with > root > set to iutil.getSysroot() . Of course, you can look in iutil.py to > see > what these do. If you *do* look there, you'll see that indeed the > relevant functions don't *currently* log the root argument. I think a > PR to add this wouldn't be unreasonable. I like this idea so I've created PR for that: https://github.com/rhinstaller/anaconda/pull/1191 Jirka ___ Anaconda-devel-list mailing list Anaconda-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/anaconda-devel-list
Re: default kickstart configuration (beyond interactive-defaults.ks)
On 05/23/2017 01:45 PM, Radek Vykydal wrote: Hello, I finally got to work on this and came up with this PR for master: https://github.com/rhinstaller/anaconda/pull/1076 We could consider it for f26 if it is useful for Atomic. Radek I closed the PR and I am moving the discussion back to the list. From the PR: > I'm thinking that it's probably best for now to find a place to hack in the logic > to say "if there's no payload specified, and we find ostree data in /ostree/repo, > pick that up automatically". - Are we able to pick all needed values specified by ostreerepo command from the data/repo we find? - How about other configuration (%post) from atomic interactive-defaults.ks - could it be moved into anaconda logic as well? So the interactive-defaults.ks mechanism is not needed... Radek On 15.2.2017 16:03, Colin Walters wrote: I'd like to fix: https://pagure.io/atomic-wg/issue/226 Right now for Atomic Host, we embed the ostree content inside the squashfs, with the goal that users of the ISO don't have to hit the network. We hook bits into interactive-defaults.ks; see this change for how it works: https://pagure.io/fedora-lorax-templates/c/e34c371501d6eac75925aae86bd737c8a43d8244?branch=master However, if one wants to then extract the ISO to do a kickstart install, the user must repeat the `ostreesetup` bits: https://lists.fedoraproject.org/archives/list/cl...@lists.fedoraproject.org/message/M6CEAGA3RGXZLSZ5GWSCTMMOV5OOXUUN/ Specifically from his attachment: ostreesetup --osname="fedora-atomic" --remote="fedora-atomic" --url="file:run/install/repo/content/repo" --ref="fedora-atomic/25/x86_64/docker-host" --nogpg which is really internal implementation details that I'd like the ability to change without breaking people's kickstarts, and conversely not require them to specify by default. Are there any facilities today for a "default kickstart" like /usr/share/anaconda/interactive-defaults.ks, except not just interactive? ___ Anaconda-devel-list mailing list Anaconda-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/anaconda-devel-list ___ Anaconda-devel-list mailing list Anaconda-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/anaconda-devel-list ___ Anaconda-devel-list mailing list Anaconda-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/anaconda-devel-list