| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
| |
kickstart.py.
|
| |
|
| |
|
|
|
|
|
|
|
| |
Since Upgrade.execute gets called on installs too (both the upgrade and
install kickstart commands have the same handler) we need to make sure
to set it to the right value. Otherwise, all kickstart installs end up
being upgrades.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
This is also the last thing in InstallData.write, so it can go away too.
|
| |
|
| |
|
|
|
|
|
| |
This patch also moves rootPassword into Users where it belongs and auth
into Security where it belongs.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
This also removes the hack where upgrade can be True, False, or None in order
to test whether we've seen the upgrade screen or not. Instead, it introduces
a global value hack. I consider this the slightly cleaner approach.
|
| |
|
| |
|
|
|
|
|
|
| |
We probably shouldn't be referring to attributes on the OptionParser object
when we've got a perfectly useful Anaconda object to use. This also allows
removing the displayMode attribute from the InstallData.
|
|
|
|
|
|
|
|
| |
Replace all usage of the isys C EDD code with the new storage
python EDD code. Note that this moves the sorting of partition.req_disks
from devices.py to paritioning.py, because sorting now needs access
to the storage object, this also has the added advantange that now we
always allocatePartitions in sorted drive order, even in interactive mode.
|
|
|
|
|
|
| |
This provides better support for excluding groups. Now, groups that were not
previously listed (because they were covered by a glob, for instance) can also
be removed.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently we rely on self.handler being defined in our execute() methods
of pykickstart command/data derived classes. But self.handler is is not
defined for BaseData derived classes, which we fix by some trickery to define
it anyways.
This patch replaces the trickery by using the anaconda parameter all
the execute() method of pykickstart command/data derived classes have, this
allows access to the same object through anaconda.id.ksdata, without requiring
the trickery.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
- Remove unnecessary semi colons in various places
- textw/upgrade_text.py: add a few missing imports
- Fix various missplaced doc strings
- Fix several double imports
- Fix AnacondaKSScript.run: Module 'os' has no 'exists' member
- users.py: Remove double definition of Users.setRootPassword
- yuminstall.py: correct typo reop -> repo
- yuminstall.py: _run: anaconda -> self.anaconda
|
|
|
|
|
|
|
|
| |
Since ignoredisk is an optional command and leaving it out of a kickstart
file means you don't want to ignore anything, there's no reason to stop
on that screen and wait for input if ignoredisk was not specified. However,
if this is an interactive kickstart install we do still need to stop and
ask.
|
|
|
|
| |
Otherwise, the autopart command won't actually do anything.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
Make iscsi,etc startup use the iscsi,etc Singletons, instead of the not
yet instantiated instData.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bring fcoe/iscsi/zfcp devices online when parsing the kickstart file
(instead of delaying this to the execute phase) so that the disks
can be identified by /dev/disk/by-path/foo, etc. in clearpart, ignoredisks,
etc. commands.
For the iscsi command this also improves error handling by passing
username and password to addTarget as is instead of only passing them if
both are present, using addTarget's error handling for a username without
a password (instead of simply ignoring the username or the password), and
also by catching IOErrors which will be thrown when logging in to the target
fails.
|
| |
|
|
|
|
|
|
| |
Most Fedora users don't need to see the advanced filtering UI with all its
multipath tabs and advanced storage buttons and search interfaces. So,
provide an introductory screen that allows skipping all that stuff.
|
|
|
|
|
|
| |
This UI is the screen that allows the user to select which devices will be
wiped during installation and which should just be mounted. It also allows
selecting which device the bootloader should be installed to.
|
|
|
|
|
|
|
|
|
|
| |
This UI allows the user to select which devices they would like to include
in the rest of the installation process, sorted out by their types. All
devices not checked will never again be referenced by anaconda, though we
may see their udev information in the logs from time to time. This UI
supports two different ways of looking at things: the simple UI for
regular Fedora users with basic devices, and the complex UI with many tabs
and filtering options.
|
|
|
|
|
|
|
|
|
|
| |
While doing some tests with iscsi and kickstart I noticed that currently
drives discovered through ibft (iscsi), edd (fcoe), or configured in the
rclinux script (zfcp), can not be used in kickstart's clearpart or
ignoredrive commands. This patches fixes this.
Note that the 3 added startup calls are complete no-ops on systems without
ibft / edd-fcoe info / pre configured zfcp.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Now you can use globs or non-device node names to specify disks in the
arguments to these commands, which should be a big help to large storage
configurations.
|
| |
|
|
|
|
|
|
| |
- Don't remove scriptlets when they've been written out to aid in debugging.
- Always log stdout/stderr.
- On errors, print the messages to anaconda.log as well.
|
|
|
|
|
| |
Among other problems, this means that all the partitioning commands can be
in a file generated from a %pre script again.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
kickstart.py never should have been written the way it previously was - with
all the code that actually did something in the parse methods. This
organization prevented the parse methods from being called until we had things
like Storage instances and led to the multiple pass hack.
This better design moves all the code that does anything into apply methods
(that pykickstart lacks, oh well) and keeps the parse methods only setting
values in KickstartCommand/KickstartData objects. This should allow doing
parsing very early, then calling the apply methods when we're set up and
therefore remove the reason for multiple passes.
This patch requires a pykickstart which can pass data objects up from deep
in dispatcher. Note also that this patch does not yet call any of the apply
methods, so kickstart is temporarily busted.
|
|
|
|
|
| |
This means we only have to specify those handlers and data objects we
require special versions of, not all of them.
|
| |
|
|
|
|
|
|
|
|
|
| |
Make MDRaidArrayDevice.__init__ raise a value exception when creating
a new (so non existing) raid set and there are not enough members for
the requested raid level.
And catch this value exception in the GUI raid dialog and kickstart raid
commands.
|
|
|
|
| |
This lets you set passwords to ssh into the anaconda environment.
|