since ceph 12.1.1 the (deprecated) parameter 'crush_ruleset' is removed
and replaced with 'crush_rule' while changing this, change from
integer to string so that we can later use the names of the rules
instead of the id
(for now there seems to be a bug that you can only use the name and
not the id)
Thomas Lamprecht [Tue, 25 Jul 2017 14:27:00 +0000 (16:27 +0200)]
dc/Backup: do not focus node selector on window focus
Do not focus anything here in the case he backup window gains focus
again. Else, if we click outside of the window the nodefield gets
always focused again - even if another field was focused previously.
This is a bit confusing, rather have no field focused at all -
especially as a blank value will be often used here.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Tue, 25 Jul 2017 14:26:59 +0000 (16:26 +0200)]
ComboGrid: open picker also when focusing an editable field
ComboBoxes provided by UI frameworks normally show their selection
possibilities once focused, ExtJS does this too but not on editable
(and thus filterable) input fields. Add a logic which allows this too
and streamlines the behavior of ComboGrid/Box like components.
To avoid a glitch, where clicking the "picker trigger" (the down
arrow which normally toggles the selection picker) would first focus
the element - thus expanding it - and only then causing a toggle
which collapses it again, we overwrite the 'Ext.form.trigger.Trigger'
onMouseDown so that we can skip a expansion on this event.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Tue, 25 Jul 2017 14:26:58 +0000 (16:26 +0200)]
FileStore: make editable to allow filtering
Making this ComboGrid based component editable allows implicit
filtering the backing store.
This is quite helpful on file selection, as for example the ISO file
selector from the create wizard is.
Clear the filter (and its displayed value) on store selection change,
as there the filter becomes void - at least this is a simple and easy
solution without glitches.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Tue, 25 Jul 2017 14:26:56 +0000 (16:26 +0200)]
FileSelector: adapt combogrid total and column width
By default the "Size" column ellipsed a lot entries, which is just
annoying. Adapt it so that all sizes can be viewed comfortably.
Further increase the total width of this combogrid, a lot less could
be seen there since theme change from the 4.X era.
Use an new empirical found out good working value.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
we now want to add a ceph-mgr daemon to every node where a ceph-mon
daemon runs, as per ceph documentation recommendation, because in
luminous the mgr daemons will not be automatically created/started
with a monitor anymore
we also give the createmon an optional id parameter, so that one
can set a custom id, and make the creation/removal of the manager
optional but the default
Improve a bit the esthetics of the User Settings window
* group the action buttons on the left, to distinguish them from the modal
'close' button
* display the action buttons as grey instead of blue, to be consistent
with the rest of the GUI
Do not display the abbreviated commit SHA-1 in the workspace version string
This string is ugly and unfit for human consumption for 99% of our users.
The same information is available in a number of place in the GUI via Node
Summary, Package Versions, Subscription (Report)
they restructured the json structure, so we have to adapt to it.
additionally, we now have a 'detail' property for the messages
which we show on a click of the 'info' button in the new actioncolumn
adapt ceph status detail to luminous and make it more robust
they restructured the json interface
(e.g. they removed the timechecks from ceph status)
so we have to generate those differently
also make the whole thing more robust to changes,
as in do not bail out if one thing is missing, but check the needed
data only when we need it and omit that part
Wolfgang Link [Tue, 18 Jul 2017 09:07:12 +0000 (11:07 +0200)]
Fix pveceph createosd
The --filestore flag is now required see doc of ceph.
If the --bluestore argument is given, a bluestore objectstore will be
created. If --filestore is provided, a legacy FileStore objectstore
will be created. If neither is specified, we default to BlueStore.
pvemailforward is a tiny oneliner, but for the sake of best
practices, let's use the build tools from the environment.
For example, with dpkg-buildpackage this will make us use
-D_FORTIFY_SOURCE=2 etc.
when having an lxc mountpoint/rootfs on a storage which does not
allow the 'rootfs' content (e.g. somebody disabled it after creating
a container there), trying to edit the mp/rootfs broke the gui
Thomas Lamprecht [Thu, 29 Jun 2017 14:37:32 +0000 (16:37 +0200)]
replication: reuse fitting translations
while it does not makes sense do over-reuse translations for the sake
of translating less, imo, here we can safely reuse already existing
ones and pull out the unit 'MB/s' from the gettext.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Thu, 29 Jun 2017 09:22:53 +0000 (11:22 +0200)]
replication; merge status and status text and add to dc mode
It was a bit strange to have two separate status columns, which do
effectively the same thing. So merge them together to save a bit
space and have less columns.
We do not need to translate 'Status Text' as a nice side effect.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 28 Jun 2017 07:19:49 +0000 (09:19 +0200)]
grid/replication: show textual description of default value
When a user sees this the first time and wants to add a job he
shouldn't be confused what the default value means, so display this
through the emptyText property, which does not get submitted to the
backend.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
tests: mock PVE::Cluster::get_members and adapt fail test
Commit 3385399339c94 ("replication: keep retrying every 30 minutes in
error state") changed the retry behavior to not stop after the 3rd error
and then stick to half-hour intervals. This needs to be reflected in the
tests. The numbers here match. (1900 + 30*60 = 3700).
Commit fd844180a7efa ("replication: don't sync to offline targets on
error states) changed the retry behavior to check whether the target
node is online. If this is not the case we fail right away. This
introduced a dependency on PVE::Cluster::get_members which we now need
to mock. Tests currently use node names "node{1,2,3}", so I just mock
those 3.
Thomas Lamprecht [Tue, 27 Jun 2017 13:01:26 +0000 (15:01 +0200)]
replication/ui: improve wording for job removal
Delete the strange sounding Note from the Removal dialog.
This was added to make sure that an user isn't confused that a Job is
still shown after 'removing' it. But it isn't needed as there is
already noted in the jobs status that it will get removed soon after
doing so, so the user sees that his action had the desired effect, no
extra note needed.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Emmanuel Kasper [Tue, 27 Jun 2017 07:52:39 +0000 (09:52 +0200)]
Add a Skip replication option on Disk Panel
This will allow to disable replication, when for instance we
add a disk on a non-replicatable storage.
The option is hidden in the wizard, because at that time no VM
replication has been set.
Dominik Csapak [Fri, 23 Jun 2017 09:36:37 +0000 (11:36 +0200)]
fix lxc hostname input window
we used a static emptytext at creation, which is wrong after editing
now copied from qemu/Options.js (the name), but instead of deleting the
hostname on the backend on an empty field, we set it to CT<VMID>
(this is also the default in the wizard)
Dominik Csapak [Fri, 23 Jun 2017 09:00:58 +0000 (11:00 +0200)]
fix reselection on vm/ct migrate in gui
there were a number of problems in how we reselected vms when they
migrated and selected:
we relied on the data in the configpanel to be static, but since
pveselnode is a reference, it actually changes when the resourcestore
changes
also, handling the reselection as a response to the statusstore failure
is very inconsistent, because we know exactly when a vm moves node in
the tree
this patch removes the whole configpanel vm migration code,
and lets the tree handle it
for this, we have to check if we have to reselect the entry,
deselect it manually and after the treeupdate reselect it
also in the right click selection check we have to check if it is
already selected, else the next selection will get canceled