]>
git.proxmox.com Git - pve-manager.git/log
Dietmar Maurer [Fri, 2 Jun 2017 08:07:33 +0000 (10:07 +0200)]
PVE::Replication - use new PVE::ReplicationState class
Dietmar Maurer [Fri, 2 Jun 2017 04:36:29 +0000 (06:36 +0200)]
PVE::Replication - simplify code
Move all state writes into run_replication()
Emmanuel Kasper [Mon, 29 May 2017 09:55:13 +0000 (11:55 +0200)]
Do not internationalize 'Pools' when used in Ceph Context
This is a Ceph domain term and should not be translated like OSD, ...
Galician had "Virtual Groups" for Pools, which does not apply for
Ceph Context.
Also most languages were using 'Pools' anyway.
Dominik Csapak [Mon, 29 May 2017 08:50:57 +0000 (10:50 +0200)]
remove "read-only" option from container root disk
with root-only enabled, you cannot start the conainer anymore,
so we remove it there entirely
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Thu, 1 Jun 2017 12:37:35 +0000 (14:37 +0200)]
fix #1400: do not use defaultButton in NotesEdit window
with a defaultButton set, we cannot enter a newline into a textarea
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 24 May 2017 10:30:47 +0000 (12:30 +0200)]
allow width and height parameter for vncshell
so that novnc can request a different screen size
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Wolfgang Bumiller [Thu, 1 Jun 2017 13:26:01 +0000 (15:26 +0200)]
replication: use a dedicated lock file
Otherwise $update_job_state's file_set_contents() call
drops our lock.
Dietmar Maurer [Wed, 31 May 2017 13:29:53 +0000 (15:29 +0200)]
make check: run replication_test6.pl
Dietmar Maurer [Wed, 31 May 2017 13:19:26 +0000 (15:19 +0200)]
PVE::Replication::replicate - call get_ssh_info on demand
Only call get_ssh_info() when needed, and do not pass $migration_network
for a simply remove_job - this produces just overhead and is not required.
If we always call get_ssh_info() we can never delete jobs using non-existing
targets.
Wolfgang Link [Wed, 31 May 2017 12:03:34 +0000 (14:03 +0200)]
Untaint the content of pve-replication-state.json.
JSON::decode will not untaint the raw text, so we get problems with the
last_sync, which we use many times in the code with open3.
Dietmar Maurer [Wed, 31 May 2017 11:54:40 +0000 (13:54 +0200)]
replication - allow removal of jobs with target on local node
Dietmar Maurer [Wed, 31 May 2017 07:38:42 +0000 (09:38 +0200)]
bump version to 5.0-11
Dietmar Maurer [Tue, 30 May 2017 12:40:08 +0000 (14:40 +0200)]
replication_test5.pl: add test for job removal
Dietmar Maurer [Mon, 29 May 2017 06:37:38 +0000 (08:37 +0200)]
PVE::API2::ReplicationConfig - implement delete
We just set the remove_job property in the configuration. Actual removal
is done asynchronous inside replicate().
Dietmar Maurer [Mon, 29 May 2017 05:49:17 +0000 (07:49 +0200)]
PVE::API2Tools::resolve_proxyto - new helper
Call proxyto_callback if that method is defined.
Dietmar Maurer [Wed, 24 May 2017 07:35:01 +0000 (09:35 +0200)]
PVE::Replication - aquire guest_migration_lock during replication
To block guest migration.
Dietmar Maurer [Tue, 23 May 2017 08:31:57 +0000 (10:31 +0200)]
PVE::Replication - implement replicate_volume()
Simply call PVE::Storage::storage_migrate()
Dietmar Maurer [Tue, 23 May 2017 08:31:18 +0000 (10:31 +0200)]
ReplicationTestEnv.pm: avoid warning about undefined value
Dietmar Maurer [Tue, 23 May 2017 06:15:28 +0000 (08:15 +0200)]
bin/test/Makefile: run all replication tests
Dietmar Maurer [Tue, 23 May 2017 05:40:30 +0000 (07:40 +0200)]
PVE::Replication::replicate - implement replicate
The actual volume replication is done in replicate_volume(), which is just
a stub for now.
I also added a regression test replication_test5.pl to verify basic
functions.
Dietmar Maurer [Tue, 23 May 2017 05:37:39 +0000 (07:37 +0200)]
PVE::Replication::remote_finalize_local_job - new helper
Simple wrapper to call "pvesr finalize-local-job" on a remote
cluster node.
Dietmar Maurer [Tue, 23 May 2017 05:33:46 +0000 (07:33 +0200)]
PVE::Replication::remote_prepare_local_job - new helper
Simple wrapper to call "pvesr prepare-local-job" on a remote
cluster node.
Dietmar Maurer [Tue, 23 May 2017 05:27:57 +0000 (07:27 +0200)]
pvesr finalize-local-job: add helper to cleanup job
This will be called after replication on the target node. For now, this
just removes old snapshots.
Dietmar Maurer [Mon, 22 May 2017 10:39:05 +0000 (12:39 +0200)]
ReplicationTestEnv.pm: mock get_ssh_info and ssh_info_to_command
Dietmar Maurer [Mon, 22 May 2017 06:26:40 +0000 (08:26 +0200)]
ReplicationTestEnv.pm: mock storage content methods
Dietmar Maurer [Mon, 29 May 2017 08:31:04 +0000 (10:31 +0200)]
pvesr prepare-local-job: new helper
Prepare for starting a replication job. This is called on the target
node before replication starts. This call is for internal use, and
return a JSON object on stdout. The method first test if VM <vmid>
reside on the local node. If so, stop immediately. After that the
method scans all volume IDs for snapshots, and removes all replications
snapshots with timestamps different than <last_sync>. It also removes
any unused volumes.
Returns a hash with boolean markers for all volumes with existing
replication snapshots.
Dietmar Maurer [Thu, 18 May 2017 06:11:09 +0000 (08:11 +0200)]
pvesr run: add --verbose flag
We can now pass a $logfunc parameter to replicate() to
produce more verbose logs/output. This is especially useful
for regression testing.
Dietmar Maurer [Thu, 18 May 2017 05:46:50 +0000 (07:46 +0200)]
replication_test4.pl: Test replication job failure
Dietmar Maurer [Thu, 18 May 2017 05:14:42 +0000 (07:14 +0200)]
ReplicationTestEnv.pm: add job tracking log
To generate persistent logs with job state/config changes.
Dietmar Maurer [Thu, 18 May 2017 04:23:15 +0000 (06:23 +0200)]
replication_test3.pl: regression test for replication to same node
Dietmar Maurer [Wed, 17 May 2017 10:28:55 +0000 (12:28 +0200)]
PVE::Replication - use new calendar events instead of interval
And implement retry algorythm after failure:
$next_sync = $state->{last_try} + 5*60*$fail_count;
and limit to 3 failures.
Dietmar Maurer [Mon, 15 May 2017 10:28:33 +0000 (12:28 +0200)]
PVE/Replication.pm: save pid/ptime to running job state
So that we can check which job is running.
Dietmar Maurer [Mon, 15 May 2017 07:43:15 +0000 (09:43 +0200)]
replication_test2.pl: test replication scheduler
Dietmar Maurer [Mon, 15 May 2017 07:07:22 +0000 (09:07 +0200)]
add regression test environment for replication
Dietmar Maurer [Thu, 11 May 2017 07:02:01 +0000 (09:02 +0200)]
pvesr: add pve storage replication tool
Just added code to configure jobs. Replication itself is not
implemented.
Emmanuel Kasper [Thu, 18 May 2017 09:10:52 +0000 (11:10 +0200)]
Fix #1386: track postfix daemon status via postfix@-
Dominik Csapak [Wed, 3 May 2017 15:04:27 +0000 (17:04 +0200)]
optimize 'text' field in resourcestore
instead of doing a lot of string comparisons, we first check
the type with the most objects (vms/cts) via number operation
and in all other cases, we can simply use the type as property index
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 3 May 2017 15:04:29 +0000 (17:04 +0200)]
add suspend/resume events to updateTree
this makes adding/modifying a large number of items a lot faster
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 3 May 2017 15:04:28 +0000 (17:04 +0200)]
optimize resourcestore field collection
instead of doing weird if/elsif checks with nonsensical logic
we simply check the 3 properties we want (namely convert,calculate and
defaultValue) and add them to our field definition.
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 3 May 2017 15:04:26 +0000 (17:04 +0200)]
optimize 'running' field
we do not need to check for qemu/lxc/node because the other types
do not have an uptime value
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Wolfgang Link [Thu, 27 Apr 2017 10:02:36 +0000 (12:02 +0200)]
Make a new systemd target.
This will ensure all storages are up before pveproxy is running.
Fabian Grünbichler [Wed, 3 May 2017 09:37:06 +0000 (11:37 +0200)]
ceph: init: rename disable-cephx to disable_cephx
for consistency
Fabian Grünbichler [Wed, 3 May 2017 08:25:24 +0000 (10:25 +0200)]
ceph: init: add warning to disable-cephx
Alexandre Derumier [Fri, 28 Apr 2017 07:52:59 +0000 (09:52 +0200)]
ceph: init: add optional disable-cephx option
Signed-off-by: Alexandre Derumier <aderumier@odiso.com>
Alexandre Derumier [Fri, 28 Apr 2017 07:52:57 +0000 (09:52 +0200)]
ceph: init: remove filestore xattr use omap
it's not used anymore since emperor
Signed-off-by: Alexandre Derumier <aderumier@odiso.com>
Dominik Csapak [Thu, 20 Apr 2017 09:18:41 +0000 (11:18 +0200)]
fix #1359: change to vm when double clicking in pool member view
like in the resource grid
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dietmar Maurer [Mon, 24 Apr 2017 05:46:07 +0000 (07:46 +0200)]
bump version to 5.0-10
Dietmar Maurer [Mon, 24 Apr 2017 05:24:12 +0000 (07:24 +0200)]
PVE/APLInfo.pm: avoid locale specific time stamps
Dietmar Maurer [Mon, 24 Apr 2017 05:15:01 +0000 (07:15 +0200)]
vzdump: do not generate locale specific time stamps
Avoid perl warning: Wide character in print
Wolfgang Bumiller [Thu, 20 Apr 2017 10:27:34 +0000 (12:27 +0200)]
bump version to 5.0-9
Wolfgang Bumiller [Thu, 20 Apr 2017 08:47:41 +0000 (10:47 +0200)]
statd: rebalance: don't use CpuSet::max_cpuids
We're already limiting CPUs to lxc/cpuset.effective_cpus,
so let's use the highest cpuid from that set as a maximum to
initialize the container count array.
Dietmar Maurer [Thu, 20 Apr 2017 04:25:43 +0000 (06:25 +0200)]
bump version to 5.0-8
Dominik Csapak [Wed, 19 Apr 2017 12:45:22 +0000 (14:45 +0200)]
filter templates in bulk start/stop
because we cannot start/stop templates
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 19 Apr 2017 12:43:55 +0000 (14:43 +0200)]
add skylake to cpu types
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Tue, 18 Apr 2017 14:55:49 +0000 (16:55 +0200)]
set default focus and default button for edit window
this patch sets following defaults for the edit window:
defaultFocus: 'field'
sets the focus on the first field it finds in its child items,
works for most edit windows, so that the cursor stands in a
textfield, or on a checkbox
defaultButton: 'submitbutton'
so that when someone presses enter, we submit the form
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dietmar Maurer [Fri, 14 Apr 2017 13:20:33 +0000 (15:20 +0200)]
bump version to 5.0-7
Dominik Csapak [Fri, 14 Apr 2017 08:09:24 +0000 (10:09 +0200)]
highlight errors in grids
we gave erroneous entries in grids the class .x-form-invalid-field
but since we changed to the "crisp" theme, this class did not have any
highlighting
so instead we use our own "pve-invalid-row" class
and give it a light red background
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Fri, 14 Apr 2017 08:09:23 +0000 (10:09 +0200)]
add right-click menu for nodes
to easier reach the bulk actions, create wizards, and shell
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Fri, 14 Apr 2017 08:09:22 +0000 (10:09 +0200)]
use nodename parameter to preselect the node in the wizard
this will be used for the ndoe right-click-menu
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Fri, 14 Apr 2017 08:09:21 +0000 (10:09 +0200)]
add a 'convert to template' in the vm action bar
this was only available on right clicking a vm, now it is also available
in the action bar
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Fri, 14 Apr 2017 08:09:20 +0000 (10:09 +0200)]
move qemu reset button into shutdown menu
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Fabian Grünbichler [Mon, 10 Apr 2017 14:25:10 +0000 (16:25 +0200)]
bump version to 5.0-6
Fabian Grünbichler [Mon, 3 Apr 2017 10:06:55 +0000 (12:06 +0200)]
Ceph: allow bigger size and min_size in API
Thomas Lamprecht [Wed, 22 Mar 2017 10:08:02 +0000 (11:08 +0100)]
ceph/pool: allow bigger size and min_size values
Some setups may want to use more replicas as three, while my proposed
maximal 7 replicas is also arbitrary it should cover all normal
usecases (i.e. 5 replicas) and a bit more, just to be sure.
Thomas Lamprecht [Wed, 22 Mar 2017 10:08:01 +0000 (11:08 +0100)]
ceph/pool: change default replica to 3/2
size/min_size 2/1 is unsafe as it can result in split brains, so we
should not recommend it as default.
If the user really wants it he can still set it explicit.
Dominik Csapak [Fri, 24 Mar 2017 13:19:11 +0000 (14:19 +0100)]
fix #1317: check wizard form also on validitychange
with the commit
40342aa6c4278497b9ec8c9cce8739f9b29c2e8f
we introduced a validator on the guestidselector, but did not
notice that the wizard checks the validity on the field change event,
but the selector gets valid/invalid in an api callback(so a little later)
with this patch, we now validate the field correctly with validate()
and also listen on the validitychange event to catch it
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Fabian Grünbichler [Fri, 24 Mar 2017 07:58:20 +0000 (08:58 +0100)]
fix start on boot
fix the fix for #1024
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Fabian Grünbichler [Wed, 22 Mar 2017 09:52:43 +0000 (10:52 +0100)]
bump version to 5.0-5
Fabian Grünbichler [Wed, 22 Mar 2017 07:23:23 +0000 (08:23 +0100)]
use new repository URL
Fabian Grünbichler [Tue, 21 Mar 2017 12:39:19 +0000 (13:39 +0100)]
bump version to 5.0-4
Dominik Csapak [Mon, 20 Mar 2017 15:17:41 +0000 (16:17 +0100)]
fix #1307: dont use language dependent separator in gui
by default, extjs submits the language dependent decimal separator,
e.g. ',' for german and '.' for english
but we always want '.'
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Tue, 21 Mar 2017 11:33:55 +0000 (12:33 +0100)]
add beta text with link to bugtracker
for the 5.0 beta, add a link to the bugtracker
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Fabian Grünbichler [Tue, 21 Mar 2017 12:35:25 +0000 (13:35 +0100)]
ceph: create default keys when creating monitor
this used to be a separate service, but got dropped in
Kraken in favour of calling ceph-create-keys manually when
deploying a node.
Fabian Grünbichler [Mon, 20 Mar 2017 12:25:11 +0000 (13:25 +0100)]
pveceph: use luminous by default
Dietmar Maurer [Fri, 17 Mar 2017 11:13:13 +0000 (12:13 +0100)]
bump version to 5.0-3
Dietmar Maurer [Fri, 17 Mar 2017 10:27:18 +0000 (11:27 +0100)]
use 'U' to encode undefined values for RRD graphs
rrdtools 1.5 and newer seems to require this.
Thomas Lamprecht [Wed, 15 Mar 2017 15:47:42 +0000 (16:47 +0100)]
remove backup locks when starting all VMs on boot
If on bootup one of our VMs is locked by an backup we safely can
assume that this backup job does not run anymore and that the lock
has no reason anymore and just hinders uptime of services.
As at this time we (the node) have quorum so we may safely assume
that we have a consistent view of the cluster and all our VMs really
belong to us. We just need to ensure that we do not run into an
automatic backup jobs, so execute our code with VZDumps lock or
timeout.
Log in the Task and Sys log that we removed the lock, so that an
admin easily sees that there may be need for cleaning leftovers from
an interrupted backup.
Addresses bug #1024
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 15 Mar 2017 15:47:41 +0000 (16:47 +0100)]
get_filtered_vmlist: save VM conf in own subhash
small refactoring in get_filtered_vmlist: save a VMs config in its
own subhash to avoid collisions with other data which we want to save
in the vmid list, for now this is only `type` but in the next patch
I want to save also the class
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 15 Mar 2017 14:46:07 +0000 (15:46 +0100)]
include also HA VMs if a vm list is explicit specified
Else the user would not had specified them explicit
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 15 Mar 2017 14:46:06 +0000 (15:46 +0100)]
VMSelctor: add HA state and allow to filter with it
This allows also to filter after HA states, so one may bulk control
HA managed services more easily or do bulk actions just on unmanaged
VMs
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 15 Mar 2017 14:46:05 +0000 (15:46 +0100)]
PVEResource store: respect defaulValue of elements
If we explicitly set a fields defaultValue we really want it in the
Model.
Also needed for a further patch
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Wed, 15 Mar 2017 14:46:04 +0000 (15:46 +0100)]
ResourceStore: fix typo s/Colum/Columns/
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Fabian Grünbichler [Wed, 15 Mar 2017 11:03:00 +0000 (12:03 +0100)]
update sources.list to stretch
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Fabian Grünbichler [Wed, 15 Mar 2017 11:02:59 +0000 (12:02 +0100)]
APT: update packages and package names
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Dietmar Maurer [Tue, 14 Mar 2017 06:59:01 +0000 (07:59 +0100)]
Makefile: use "ssh -X" for upload
Dietmar Maurer [Tue, 14 Mar 2017 06:57:15 +0000 (07:57 +0100)]
bump version to 5.0-2
Dietmar Maurer [Tue, 14 Mar 2017 06:34:22 +0000 (07:34 +0100)]
control.in: change description to avoid lintian warning
Avoids warning: description-synopsis-starts-with-article
Dietmar Maurer [Tue, 14 Mar 2017 06:29:00 +0000 (07:29 +0100)]
depend on libc6 to avoid lintian error missing-dependency-on-libc
Dietmar Maurer [Tue, 14 Mar 2017 05:56:09 +0000 (06:56 +0100)]
pveproxy: do not serve /usr/share/vncterm - no longer required
Note: We removed the java applet
Dietmar Maurer [Tue, 14 Mar 2017 05:39:04 +0000 (06:39 +0100)]
install lintian overrides file
Wolfgang Bumiller [Mon, 13 Mar 2017 13:23:52 +0000 (14:23 +0100)]
bump version to 5.0-1
Wolfgang Bumiller [Mon, 13 Mar 2017 12:06:09 +0000 (13:06 +0100)]
bump version to 4.4-13
Fabian Grünbichler [Thu, 9 Mar 2017 09:34:57 +0000 (10:34 +0100)]
fix ceph.service for Ceph Jewel >= 10.2.6
having our ceph.service pulled in by ceph.target does not
work anymore, because "systemctl start ceph.target" hangs
forever on ceph-common upgrades. multi-user.target seems to
work as well, and we are ordered after pve-cluster anyway.
only replace the old ceph.service if it is an exact match.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Wolfgang Bumiller [Tue, 7 Feb 2017 14:39:06 +0000 (15:39 +0100)]
buildsys: depend on lsb-base
Wolfgang Bumiller [Tue, 7 Feb 2017 14:38:33 +0000 (15:38 +0100)]
buildsys: make job safety
Wolfgang Bumiller [Tue, 7 Feb 2017 14:38:54 +0000 (15:38 +0100)]
remove space in service file
Dominik Csapak [Fri, 10 Mar 2017 09:12:25 +0000 (10:12 +0100)]
fix diffstore issue with clusterlog
extjs cannot "convert" and id from other fields, so the ids in the
diffstore and the realstore are different and we re-add every element on
every update
to mitigate this, we generate the id (which is "uid:hostname") in the
backend, and simply use it in the frontend
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Tue, 21 Feb 2017 09:53:15 +0000 (10:53 +0100)]
use validator for GuestIDSelector and optimize labels
instead of using markInvalid, use a validator, which also marks the
field dirty and marks the form correctly invalid.
also optimize the label/error message assignments
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 15 Feb 2017 14:00:24 +0000 (15:00 +0100)]
show nodeselector errors only if a node is selected
otherwise we get the error:
'Node seems to be offline'
if we do not have a node selected
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Dominik Csapak [Wed, 15 Feb 2017 13:55:50 +0000 (14:55 +0100)]
check earlier if the logview is destroyed
because extjs 6.2 handles destroying of components differently (namely
asynchronous), we have to check earlier if the view is destroyed, else
we get a 'cannot access ~ of undefined' error
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>