| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
Right now we identify disks as partedDisk.model, which looks really bad
when using DM-based devices. Add a description to DiskDevice,
defaulting do partedDisk.model, and use it in the UI, so that subclasses
can easily replace this with more useful text.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make ifcfg configuration files getting parsed properly after update. The used
inotify configuration update mechanism (ifcfg-rh NM plugin) requires 1) writing
of new files out of /etc/sysconfig/network-scripts dir so that it doesn't
trigger parsing of the file too early (before all is written), and more
importantly, 2) removing of the old files before moving the new ones in so that
the new file gets parsed and the respective connection gets re-read and
eventually activated.
Also make sure that only device selected in UI has ONBOOT set to yes
in case of selection of another device after fail.
|
|
|
|
|
| |
We have lots of strings that we display to users which are unclear or
contain poor usage and grammar. That's bad.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds support for using FCoE during the installation. This patch
merely lays the initial ground work, there is more work todo:
- The system will not boot without manual help after the install, as
dracut / mkinitrd do not support FCoE yet
- If FCoE is not used for / but for example for /srv, then information
about the nic used for FCoE needs to be written in a to be defined
config file in the system, and rc.sysinit needs to be thought to read
this file and bring up FCoE SAN's / Fabrics not used for /
- kickstart support for FCoE still needs to be done
|
| |
|
|
|
|
|
|
|
| |
Fix of this UI flow: doing storage reset after root partitions discovery (e.g.
in partitioning step) and then going back and selecting upgrade. In place of
using root partition device objects obsoleted by storage reset, find root
partition devices of new device tree.
|
| |
|
|
|
|
|
| |
When manually adding an iscsi disk it will not show up in the where to
install the bootloader dropdown, this patch fixes this.
|
| |
|
|
|
|
|
|
|
|
| |
We adding a remote repo while performing a CD or DVD install, you're
presented with netconfig_dialog. The combo box listing network devices
used to show 'device_name - description'. Change this to 'device_name -
mac_address' to match our other interface listings in anaconda (plus,
the description available to us now is sort of useless).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(#503310) (#503681)
Default clearPartType to None so that all devices are discovered during storage
initialization step, and UI parttype can be set to its default "Replace existing
linux system". Also set clearPartType appropriately when going back from partition dialog
so that all devices are discovered during storage reset.
There is one case that can't be resolved this way: clearpart --all or --linux
is set in ks and "Create custom layout" is selected in parttype UI step. Because of ks
setting, storage initialization doesn't discover devices (RAID, LVM) on
partitions that are to be cleared and therefore they are not present in custom
partition dialog. This can be workarounded by going back and than again to
custom partitioning dialog. I think we should grey-out type of partitioning
combo set to value from ks clearpart command.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The i18n people have suggested using ngettext when we need to have
singular and plural forms of strings, where the count will vary as to
what we are reporting to the user.
I've made the changes they have suggested. I created a new lambda
function called P_() to use for the plural cases. P_() takes in three
parameters:
1) The singular form of the string.
2) The plural form of the string.
3) A count.
Here's an example:
....some loop runs doing stuff
bytesWritten = 47
msg = P_("Wrote %d byte.",
"Wrote %d bytes.",
bytesWritten) % (bytesWritten,)
print msg
The % substitution is correct at the end because P_() returns a single
string, so we only need the format string to account for that.
Some strings have been changed slightly to make it easier for
translations to other languages, particularly when choosing plural forms.
|
|
|
|
|
| |
Show unknown partition types as "Unknown" instead of "None" in the
partition editor window.
|
|
|
|
|
|
|
| |
The use of a unique id frees us from having to worry about parted renaming
partitions whenever we remove one. It will also provide a mechanism with
which we can more reliably track protected partitions, currently
problematic for the same reason.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Seriously use VNC already.
|
|
|
|
|
| |
This does not allow you to click on a freespace slice in the graph and then
edit, new, delete, etc. but it does prevent tracebacks. One step at a time.
|
|
|
|
| |
This was causing failures to set large(r) extent size with small(er) VGs.
|
|
|
|
|
|
|
|
| |
This fixes the problem where highlighting the partition in the slice graph
doesn't allow you to click "edit", and where selecting from the tree view
doesn't highlight the partition in the graph. The underlying problem was
that our list stores storage objects, but we were passing in a pyparted
object to do the comparison with.
|
|
|
|
|
|
| |
Give the hostname entry field focus when we get to that screen in
anaconda. Pressing Enter in the entry box will advance to the next
screen.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch contains the following small fixes:
- There is no need to convert None as username/pass into an empty string
pylibiscsi will happily take either
- Only set the initiator name from ibft if the ibft flag is present
- Show an error when no username is specified, but a password /
reverse username / pass is given
- Do not backtrace when the following happens:
1) Manually add iscsi disk
2) enter wrong IP / username without pass
3) Fix this *and* change the initiator name initially choosen
- Allow having only a reverse password without a reverse username
|
|
|
|
|
|
| |
In createPEOptionMenu(...), store in the actualPE list only the physical
extents (PE) displayed by the combo-box, not all the possible physical
extents.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
This only fixes the traceback, but not the deeper cause which
is that we are not editing original format request, but our last
request (edit), so e.g. in 2nd edit "Original filesystem type"
label and "Migrate" option are wrong (reflecting our last
request, not the original format). I added some more conditions
so that we do not display the info and do not offer migration
in such cases. Note that the checks also affect non-raid
existing partition editing in the same sense. The conditions
should be kept when the behavior is fixed.
|
|
|
|
|
|
|
| |
If you have an existing disk layout and choose custom partitioning and
want to resize an existing logical volume and filesystem, targetSize was
getting the new size value before an ActionFormatResize() was created
for the device, which resulted in a traceback.
|
| |
|
| |
|
|
|
|
| |
...but only they are reformatting the device.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
When carrying over the format's attributes we do not consider the
possiblility that the partition is encrypted, in which case we want to
preserve the attributes of the LUKSDevice's - not the PartitionDevice's -
format. This was causing a traceback whenever a user tried to edit a
preexisting encrypted partition.
|
|
|
|
| |
It was broken (empty) when editing non-preexisting array.
|
|
|
|
|
|
| |
I didn't want to add another 0/'0'/'RAID0' tweaking so I made
mdraid.raid_levels and mdraid deviceclass level attribute to contain
constants for levels defined in mdraid.py.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
parted and lvm have a different idea of the exact size of an LV, so
when using currentSize to see if we need to resize, we will always
create a ResizeAction, which will later on fail as "lvm lvresize"
returns an error status if the requested resize is a nop.
So use size instead, this is safe as we never change the size of
existing lv's.
This fixes a traceback with a lvmresize exception when trying to use
an existing LV during install without changing its size.
|
|
|
|
|
|
| |
Currently if you create a new partition from the UI and then later
decide to change the size using edit partition, the size does not
get changed. This patch fixes this.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We may think a partition has a valid filesystem when it really does not.
This is easily reproduced with the following steps:
1) Create a partition on some device.
2) Create an NTFS filesystem on that partition.
3) Delete the partition from the device.
4) Divide the device in to two equal sized partitions, but do not
create a filesystem on either.
Boot the installer and notice that the first partition is assumed to be
NTFS, but it's unmountable. In getExistingSize(), we get an FSError
traceback when trying to mount the filesyste.
The problem is that we haven't zeroed out the new partition or created a
new filesystem on top of it. The old filesystem data is still on the
disk and is misleading udev and friends.
The solution in this patch is to catch the FSError from
getExistingSize(). If the mount failed, assume the partition is empty
and continue.
|
| |
|
| |
|
| |
|
|
|
|
| |
We may not have formatcb in all cases, don't assume it exists.
|