Dominik Csapak [Wed, 1 Jun 2016 14:17:07 +0000 (16:17 +0200)]
check if tabs exist before selecting them
when switching from a vm to a template, if you
have a tab selected which does not exists in a template
(for example console or task history) you break
the site
this patch checks if the wanted tab actually exists,
and leave it on default (the first) when it does not
Emmanuel Kasper [Wed, 1 Jun 2016 10:15:44 +0000 (12:15 +0200)]
Add custom syle to pveHelpButton
Next / OK are already displayed in blue, which is the 'call-to-action'
color we use everywhere.
To prevent stealing attention from these buttons, switch help button
to grey
Emmanuel Kasper [Wed, 1 Jun 2016 10:15:42 +0000 (12:15 +0200)]
Add support for onlineHelp in Creation Wizard
Inside a wizard, switching to a new tab will fire
the 'activate' event to the new tab, causing
the inputPanel of this tab to display its help in
the wizard window.
Thomas Lamprecht [Fri, 20 May 2016 10:07:39 +0000 (12:07 +0200)]
Status: report errors on socket creation problems
If the socket couldn't be created (e.g. FQDN not resolvable) we
continued witouth any hint, when actualy writing the data we then
die'd. The user then does not really know why, so report errors
if the socket creation failed.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Fri, 20 May 2016 10:07:38 +0000 (12:07 +0200)]
Status: allow IPs and move properties to base class
We only allowed servers with the dns-name format, as such status
server may often be in internal networks and with no hostname
(testing, small network so no dns, ...) do not limit the
configuration possibilities with no reason.
Also move the base property part to the base Status class, all
current plugins use server and port so no need for double
declaration of format/descriptions.
If a future plugin doesn't need them it can omit them by not
returning the respective properties in the options method
inherited by SectionConfig.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Dominik Csapak [Wed, 18 May 2016 13:55:21 +0000 (15:55 +0200)]
fix reset button with bind mounts
if we do not disable the storageselector with bind mounts,
it automatically sets the first storage which can hold
containers (because of allowBlank: false) and triggers
the dirty flag but clicking on reset does not delete
the value there
please note that even if ceph(osd|mon).target are currently not available in ceph packages,
they are not used by proxmox api currently, because we always specify an id to a service.
targets are used to start/stop all services at the same time. Starting speficic serviceid is
done through ceph-(osd|mon)@.service files.
V2 changelog :
- remove setsid
- use [A-Za-z0-9]{1,32} pattern for services
Emmanuel Kasper [Tue, 10 May 2016 13:01:54 +0000 (15:01 +0200)]
Enlarge Object Grid first column default size
The framework value of 100 is not enough for many of our description
fields.
Setting this at the parent component level allow to remove
the end component setting.
Dominik Csapak [Fri, 6 May 2016 06:48:29 +0000 (08:48 +0200)]
fix #979: check always for migrated vm
we only checked for grouping attributes for moved vms,
but this is only true in server view
e.g. in pool view the grouping attributes is pool
and the item is not reselected, resulting in a "file not found"
error (it looks for the conf on the wrong node)
From ExtJS4 to 6 the debug files have doubled in size from
4.4MB to 8.8MB
HTTPServer.pm compress these files on the fly before sending them
to the browser (as "Accept-Encoding: gzip" is the default in
browsers)
On very slow systems (AMD Bobcat for instance, as discussed in
https://forum.proxmox.com/threads/no-webfrontend-showed-in-browser-with-new-gui-pvetest.26951/#post-135699
the compression of this file can take up to 5 seconds, and this will
hit the HTTP Timeout (line 1629) we use for serving static files, provoking the
error:
curl -k --verbose --insecure --header "Accept-Encoding: gzip" https://slowhost:8006/pve2/ext6/ext-all-debug.js
curl: (18) transfer closed with 1991108 bytes remaining to read
Non-debug files are 4 times smaller, and since everybody will benefit of having
a faster initial load of PVE, there is no reason not to make this default.
Dominik Csapak [Mon, 2 May 2016 13:43:56 +0000 (15:43 +0200)]
fix combobox reset behaviour
on comboboxes/combogrids with multiselect,
if you deselect an item (but not the last),
the order of the selected items after resetting is
not the same as the original order
because of this, the reset button is still enabled
this happens, because extjs only adds the missing
values instead of overwriting the whole array
with this fix, we overwrite the reset function of
the comboboxes and only if the values do not match
(after a reset) do we clear and set the value again
so we only change the behaviour when it is necessary
Dominik Csapak [Mon, 2 May 2016 12:35:59 +0000 (14:35 +0200)]
fix combogrid multiselect bug
this fixes a nasty combogrid/multiselect/storeload bug
in which the value of a combogrid gets set to a string instead
of an array, under certain circumstances
we overwrite the getRawValue method of our ComboGrid,
and on multiselect ComboGrids, we just get the value out of
me.rawValue, else we call the method of the parent class
we mistakenly did not reset the originalvalues after the inital
load, which meant that on reset we would not get the correct values
furthermore we have to suspend only the change event and not all,
otherwise the functions for enabling/disabling the ok/reset buttons
do not trigger (on validitychange and dirtychange)
Do not display the internal value we use for storing ostypes
Strings like 'Linux 4.X/3.X/2.6 Kernel' are difficult
enough to visually parse, so don't append the internal
representation (l26) at the end: it does not bring any relevant
information to the user, and we usually don't display
internal representations in the GUI.
previously, we had a disk/memory usage column in the resource grid,
whose raw data were the bytes used, but we added a renderer
to display it as percentage
with this, the columns sorted by bytes and not by percentage,
which made the column rather confusing
with this patch, we add a real percentage column (where the
data is a float from 0 to 1) so it sorts correctly
also make the old columns show the used size in B/KiB/etc.
by default we still only show the percentage column
since the disk usage part was always zero with qemu guests,
leave it empty there
projects / pve-manager.git / log