- Kimchi-devel - oVirt List Archives

[PATCH] Packaging: add the basemodel.py to src/kimchi/Makefile.am
by Paulo Vital 04 Feb '14

04 Feb '14

src/kimchi/basemodel.py is a common model builder and is imported by several models from src/kimchi/model, and it is not in src/kimchi/Makefile.am, so make install, make dist and make rpm do not install or package this file. Signed-off-by: Paulo Vital <pvital(a)linux.vnet.ibm.com> --- src/kimchi/Makefile.am | 1 + 1 file changed, 1 insertion(+) diff --git a/src/kimchi/Makefile.am b/src/kimchi/Makefile.am index ffcdcad..855eb9e 100644 --- a/src/kimchi/Makefile.am +++ b/src/kimchi/Makefile.am @@ -26,6 +26,7 @@ kimchi_PYTHON = \ __init__.py \ asynctask.py \ auth.py \ + basemodel.py \ cachebust.py \ disks.py \ distroloader.py \ -- 1.8.3.1

2 1

[ RFC V2 ] Manage guest CDROM devices
by Rodrigo Trujillo 03 Feb '14

03 Feb '14

Reference: https://github.com/kimchi-project/kimchi/wiki/Todo-1.2 https://github.com/kimchi-project/kimchi/wiki/customize-VM According to discussion in other thread: _API__definition_ **URI:** /vms/*:name*/storages * **GET**: Retrieve all storages attached to the VM * **POST**: Attatch new storage to specified virtual machine * dev : The guest device name of the storage * type: The type of the storage: cdrom, disk * path: Path of ISO or DISK. ### Sub-resource: storage **URI:** /vms/*:name*/storages/*:dev* * **GET**: Retrieve the storage information * dev: The guest device name of the storage * path: Path of the storage * type: Type of the storage * **PUT**: Update storage information * name: The name of the storage in the VM * path: Path of ISO or DISK * **DELETE**: Remove the storage _User __possibilities_ * List cdroms (GET - collection) * Add new cdrom (will be used to attach cd) (POST - collection) * Lookup information of a cdrom (GET - resource) * Update parameters of a cdrom (PUT - resource) * Delete a cdrom (will be used to detach) (DELETE - resource) - Will not be possible to detach and attach new cd on the same device. This operation will be done removing and adding new storage - Eject cd operation will not be provided _Perfomance/Tunning_ * Still to be defined

1 0

[PATCH V3 0/4] (WIP) Storagepool SCSI/FC
by Rodrigo Trujillo 03 Feb '14

03 Feb '14

Still working in test cases and mockmodel V3: - Changed API to only receive the scsi host name when creating new pool - Changed API to require LUN when creating new VM on SCSI pool - Created feature test and removed libvirt test function - Rebased with new model structure - Added error function handlers to UIs - Fixed LUN selection window V2: - Implements Fibre Channel devices discover in the host - Allow vms_create receive a volume to create the disk (if pool is SCSI) - Create basic UI to select SCSI Host when creating SCSI FC pool - Draft of UI to select LUN to create new VM when template has a SCSI pool configured. (Need help of UI guys here!) Rodrigo Trujillo (4): Storagepool SCSI/FC: Implement node devices API backend Storagepool SCSI/FC: Backend implementation Storagepool SCSI/FC: Implement UI for FC scsi_host pool Storagepool SCSI/FC: Modifies UI flow to select a LUN to new VM docs/API.md | 5 ++- src/kimchi/API.json | 14 ++++++- src/kimchi/control/host.py | 16 +++++++ src/kimchi/featuretests.py | 27 ++++++++++++ src/kimchi/model/config.py | 5 ++- src/kimchi/model/host.py | 54 ++++++++++++++++++++++++ src/kimchi/model/libvirtstoragepool.py | 45 ++++++++++++++++++++ src/kimchi/model/storagepools.py | 22 ++++++++-- src/kimchi/model/templates.py | 5 +++ src/kimchi/model/vms.py | 25 ++++++++++- src/kimchi/vmtemplate.py | 31 +++++++++++++- ui/js/src/kimchi.api.js | 24 +++++++++++ ui/js/src/kimchi.guest_add_main.js | 72 ++++++++++++++++++++++++++++++-- ui/js/src/kimchi.storagepool_add_main.js | 46 +++++++++++++++++++- ui/pages/i18n.html.tmpl | 4 ++ ui/pages/storagepool-add.html.tmpl | 12 ++++++ 16 files changed, 392 insertions(+), 15 deletions(-) -- 1.8.5.3

1 4

[PATCH V2 0/7] (WIP) Storagepool SCSI/FC
by Rodrigo Trujillo 03 Feb '14

03 Feb '14

There is still some work to be done in this functionality. V2: - Implements Fibre Channel devices discover in the host - Allow vms_create receive a volume to create the disk (if pool is SCSI) - Create basic UI to select SCSI Host when creating SCSI FC pool - Draft of UI to select LUN to create new VM when template has a SCSI pool configured. (Need help of UI guys here!) Rodrigo Trujillo (7): Storagepool SCSI/FC: SCSI/Fibre Channel backend implementation Storagepool SCSI/FC: Function to check libvirt version Storagepool SCSI/FC: Assign SCSI fibre channel LUN as disk to a new guest Storagepool SCSI/FC: Implement node devices API backend Storagepool SCSI/FC: Implement UI when a FC scsi_host will be the pool Storagepool SCSI/FC: Create new VM with given SCSI FC LUN (backend) Storagepool SCSI/FC: Draft UI to allow user to select a LUN to new VM docs/API.md | 7 ++ src/kimchi/API.json | 19 ++++- src/kimchi/control/devices.py | 40 +++++++++++ src/kimchi/control/host.py | 3 + src/kimchi/model.py | 117 +++++++++++++++++++++++++++++-- src/kimchi/utils.py | 18 ++++- src/kimchi/vmtemplate.py | 48 ++++++++++++- ui/js/src/kimchi.api.js | 24 +++++++ ui/js/src/kimchi.guest_add_main.js | 32 +++++++++ ui/js/src/kimchi.storagepool_add_main.js | 40 ++++++++++- ui/pages/i18n.html.tmpl | 1 + ui/pages/storagepool-add.html.tmpl | 12 ++++ 12 files changed, 350 insertions(+), 11 deletions(-) create mode 100644 src/kimchi/control/devices.py -- 1.8.4.2

4 34

Re: [Kimchi-devel] RFC: "Add disk to existing logical storage> pool"
by Frank Novak 31 Jan '14

31 Jan '14

> Date: Fri, 31 Jan 2014 11:42:47 -0200 > From: Daniel H Barboza <danielhb(a)linux.vnet.ibm.com> > To: kimchi-devel(a)ovirt.org > Subject: [Kimchi-devel] RFC: "Add disk to existing logical storage > pool" > Message-ID: <52EBA857.2020302(a)linux.vnet.ibm.com> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Hello! > > I am almost delivering a first version of this feature but now I am > having doubts in how it might work or not. > > My implementation would use vgextend and add a disk to an existing LVM > pool. The question is that, in this process, all existing data in the > added partition will be deleted. Is this ok? We can warn the user about > it, of course. But I am not sure if this is the intended design. > > I've spoken with Aline a while ago and she agreed that the > implementation would function similar to what we have today when > creating a new LVM pool. We simply "do not care" about the potential > data loss when adding the disk to an existing pool. > > Any thoughts? > Certainly if we are going to trash data, we do need to warn the user and make sure they want to proceed... I don't think that is necessarily bad, should be expected for knowledgable admin... Not sure there's an easy answer for add w/o loosing data (well certainly backup/restore after) Primary use case driving this was really the case of adding a new physical disk to system and then add this to the storage pool.. I guess someone could want to take an in use disk from some other system, or elsewhere on the same system, and want ot add that to the pool, thinking that they could still use what was already on there... At this point, I think we stick to KISS... maybe we include the warning/caveat in the Help page for this as well... > > thanks! > > > > ------------------------------ > Cheers, Frank ----------------------------------------------------------------------------------------------------------------------- Frank Novak ( 诺帆 nuò、fān ) STSM, SCEM Open Hypervisor IBM Linux Technology Center US: fnovak(a)us.ibm.com ; Notes: Frank Novak/Watson/IBM @IBMUS cell : 919-671-7966 -------------------------------------------------------------------------------------------------------------------------

1 0

[PATCH 0/4] Host's software update support
by Paulo Vital 30 Jan '14

30 Jan '14

This patch set provides support to host's software update operations. At this point, an agnostic class is providing support to backend and REST API operations. In addition, YUM (for RHEL and Fedora), APT (for Debian and Ubuntu) and ZYPPER (for OpenSuse) specific classes are provided to support the operation os each software update system. There's no test case to check backend once the software update information is volatile to each system/box. To test the REST API, execute the following commands (all them are agnostic of the host's distro): 1) Get list of all packages to be updated in the host: $curl -H 'Content-type: application/json' -H 'Accept: application/json' http://localhost:8000/host/softwareupdate/ -X GET 2) Update the host system: $ curl -H 'Content-type: application/json' -H 'Accept: application/json' http://localhost:8000/host/softwareupdate/update -X POST -d '' Paulo Vital (4): Host's software update: Update API.md Host's software update: Update REST API Host's software update: Update backend. Host's software update: Update Makefile docs/API.md | 20 ++++ src/kimchi/Makefile.am | 1 + src/kimchi/control/host.py | 12 ++ src/kimchi/mockmodel.py | 12 ++ src/kimchi/model.py | 8 ++ src/kimchi/swupdate.py | 279 +++++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 332 insertions(+) create mode 100644 src/kimchi/swupdate.py -- 1.8.3.1

5 20

[PATCH V2]Avoid useless libvirt error log produced by featuretest
by apporc 30 Jan '14

30 Jan '14

v2-v1: 1. Just hide the portocol type error, leave other error messages as it is.(Thanks Cristian) 2. Unregister the error handler when necessary.(Thanks Cristian) 3. Keep back kimchi log "*** Running feature tests ***", etc.(Thanks Cristian) 4. Move libvirt error handler function inside of Featuretest.(Thanks Aline) v1: Avoid useless libvirt error log produced by featuretest apporc (1): Avoid useless libvirt error log produced by featuretests src/kimchi/featuretests.py | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) -- 1.7.9.5

5 8

[PATCH V2 1/2] Plugins: Fix plugins framework in kimchi
by Rodrigo Trujillo 30 Jan '14

30 Jan '14

If the plugin tab name string is not found in kimchi i18n file, the tab name is set as 'unknown'. This patch also add more messages to kimchis log, and fixes a problem with the plugins extra tabs file location Signed-off-by: Rodrigo Trujillo <rodrigo.trujillo(a)linux.vnet.ibm.com> --- src/kimchi/server.py | 13 ++++++++++--- ui/js/src/kimchi.main.js | 2 +- 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/src/kimchi/server.py b/src/kimchi/server.py index b820263..4fcb0df 100644 --- a/src/kimchi/server.py +++ b/src/kimchi/server.py @@ -28,6 +28,7 @@ import logging.handlers import os import sslcert + from kimchi import auth from kimchi import config from kimchi import model @@ -208,14 +209,19 @@ class Server(object): script_name = plugin_config['kimchi']['uri'] del plugin_config['kimchi'] + tabs_extra_file = config.get_plugin_tab_xml(plugin_name) plugin_config['/ui/config/tab-ext.xml'] = { 'tools.staticfile.on': True, - 'tools.staticfile.filename': - config.get_plugin_tab_xml(plugin_name), + 'tools.staticfile.filename': tabs_extra_file, 'tools.nocache.on': True} except KeyError: continue - + except IOError as e: + msg = "Failed to load plugin tabs file %s" %tabs_extra_file + cherrypy.log.error_log.error(msg) + cherrypy.log.error_log.error('Ignoring plugin "%s"', + plugin_name) + continue try: plugin_app = import_class(plugin_class)() except ImportError: @@ -223,6 +229,7 @@ class Server(object): plugin_class) continue cherrypy.tree.mount(plugin_app, script_name, plugin_config) + cherrypy.log('Plugin loaded: %s' %plugin_class) def _init_ssl(self, options): ssl_server = cherrypy._cpserver.Server() diff --git a/ui/js/src/kimchi.main.js b/ui/js/src/kimchi.main.js index cc8afee..8b84a34 100644 --- a/ui/js/src/kimchi.main.js +++ b/ui/js/src/kimchi.main.js @@ -211,7 +211,7 @@ kimchi.getTabHtml = function(url) { $(xmlData).find('tab').each(function() { var $tab = $(this); var titleKey = $tab.find('title').text(); - var title = i18n[titleKey]; + var title = i18n[titleKey] || titleKey; var path = $tab.find('path').text(); tabsHtml += "<li><a class='item' href=" + path + ">" + title + "</a></li>"; }); -- 1.8.4.2

3 10

Sprint 3 deadlines
by Aline Manera 30 Jan '14

30 Jan '14

Hi all, As we agreed in last scrum meeting <http://kimchi-project.github.io/kimchi/meetings/kimchi.scrum.2014-01-29-13.…> (Jan 29th), from now on all the sprints will have deadlines. Which means the RFC or first patch series needs to be send to mail list at the begging of the sprint. Following are the dates I recommend everyone to follow for sprint 3. *02/03 09:00 (**UTC-03:00*) - RFC or first patch series By this date all tasks planned in the sprint needs to have RFC or patch on mail list. *02/10 **09:00 (**UTC-03:00*) - Next patch version Address comments made on RFC and moving to final patch series version. It only applies to the tasks planned for the sprint. All other patches are welcome any time - bug fixes, improvements, etc. Regards, Aline Manera

1 0

[PATCH v2 1/2] Break the 'sample' plugin's monolithic model into several smaller ones
by Mark Wu 30 Jan '14

30 Jan '14

From: Zhou Zheng Sheng <zhshzhou(a)linux.vnet.ibm.com> In kimchi the model is always a monolithic class. This is not good from the point of modulization, and it is not necessary to put everything into one big model. Unfortunately the problem will not get solved unless we conduct enormous refactor against the whole controller-model architecture. This patch tries to mitigate the problem by firstly splitting the model class into some smaller ones, than automatically mapping all sub-model instance methods to a RootModel instance. As a demonstration, this patch splits the 'sample' plugin's model. Signed-off-by: Zhou Zheng Sheng <zhshzhou(a)linux.vnet.ibm.com> --- plugins/__init__.py | 23 ++++++++++ plugins/sample/model.py | 109 ++++++++++++++++++++++++++++++------------------ src/kimchi/model.py | 26 ++++++++++++ 3 files changed, 117 insertions(+), 41 deletions(-) diff --git a/plugins/__init__.py b/plugins/__init__.py index e69de29..4103a65 100644 --- a/plugins/__init__.py +++ b/plugins/__init__.py @@ -0,0 +1,23 @@ +# +# Project Kimchi +# +# Copyright IBM, Corp. 2014 +# +# Authors: +# Zhou Zheng Sheng <zhshzhou(a)linux.vnet.ibm.com> +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + + diff --git a/plugins/sample/model.py b/plugins/sample/model.py index 9a2f22f..e606176 100644 --- a/plugins/sample/model.py +++ b/plugins/sample/model.py @@ -21,79 +21,106 @@ # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA from kimchi.exception import InvalidOperation, NotFoundError +from kimchi.model import RootModel -class Model(object): +class CirclesModel(object): + def __init__(self): + self._circles = {} + + def create(self, params): + name = params['name'] + if name in self._circles: + raise InvalidOperation("Circle %s already exists" % name) + self._circles[name] = Circle(params['radius']) + return name + def get_list(self): + return sorted(self._circles) + + +class CircleModel(object): + def __init__(self, parent_model): + # Circel and Circles models are friends, it's OK to share _circles. + self._circles = parent_model._circles + + def lookup(self, name): + try: + circle = self._circles[name] + except KeyError: + raise NotFoundError("Circle %s not found" % name) + return {'radius': circle.radius} + + def update(self, name, params): + if name not in self._circles: + raise NotFoundError("Circle %s not found" % name) + self._circles[name].radius = params['radius'] + return name + + def delete(self, name): + try: + del self._circles[name] + except KeyError: + pass + + +class RectanglesModel(object): def __init__(self): - self.rectangles = {} - self.circles = {} + self._rectangles = {} - def rectangles_create(self, params): + def create(self, params): name = params['name'] - if name in self.rectangles: + if name in self._rectangles: raise InvalidOperation("Rectangle %s already exists" % name) - self.rectangles[name] = Rectangle(params['length'], params['width']) + self._rectangles[name] = Rectangle(params['length'], params['width']) return name - def rectangles_get_list(self): - return sorted(self.rectangles) + def get_list(self): + return sorted(self._rectangles) + - def rectangle_lookup(self, name): +class RectangleModel(object): + def __init__(self, parent_model): + self._rectangles = parent_model._rectangles + + def lookup(self, name): try: - rectangle = self.rectangles[name] + rectangle = self._rectangles[name] except KeyError: raise NotFoundError("Rectangle %s not found" % name) return {'length': rectangle.length, 'width': rectangle.width} - def rectangle_update(self, name, params): - if name not in self.rectangles: + def update(self, name, params): + if name not in self._rectangles: raise NotFoundError("Rectangle %s not found" % name) try: - self.rectangles[name].length = params['length'] + self._rectangles[name].length = params['length'] except KeyError: pass try: - self.rectangles[name].width = params['width'] + self._rectangles[name].width = params['width'] except KeyError: pass return name - def rectangle_delete(self, name): + def delete(self, name): try: - del self.rectangles[name] + del self._rectangles[name] except KeyError: pass - def circles_create(self, params): - name = params['name'] - if name in self.circles: - raise InvalidOperation("Circle %s already exists" % name) - self.circles[name] = Circle(params['radius']) - return name - def circles_get_list(self): - return sorted(self.circles) - - def circle_lookup(self, name): - try: - circle = self.circles[name] - except KeyError: - raise NotFoundError("Circle %s not found" % name) - return {'radius': circle.radius} +class Model(RootModel): + def __init__(self): + circles = CirclesModel() + circle = CircleModel(circles) - def circle_update(self, name, params): - if name not in self.circles: - raise NotFoundError("Circle %s not found" % name) - self.circles[name].radius = params['radius'] - return name + rectangles = RectanglesModel() + rectangle = RectangleModel(rectangles) - def circle_delete(self, name): - try: - del self.circles[name] - except KeyError: - pass + return super(Model, self).__init__( + [circle, circles, rectangle, rectangles]) class Rectangle(object): diff --git a/src/kimchi/model.py b/src/kimchi/model.py index ea528c6..47f2d22 100644 --- a/src/kimchi/model.py +++ b/src/kimchi/model.py @@ -1853,3 +1853,29 @@ class LibvirtConnection(object): # However the values need to be considered wisely to not affect # hosts which are hosting a lot of virtual machines return conn + + +class RootModel(object): + ''' + This model squashes all sub-model's public callable methods to itself. + + Model methods are not limited to get_list, create, lookup, update, delete. + Controller can call generate_action_handler to generate new actions, which + call the related model methods. So all public callable methods of a + sub-model should be mapped to this model. + ''' + def __init__(self, model_instances): + for model_instance in model_instances: + cls_name = model_instance.__class__.__name__ + if cls_name.endswith('Model'): + method_prefix = cls_name[:-len('Model')].lower() + else: + method_prefix = cls_name.lower() + + callables = [m for m in dir(model_instance) + if not m.startswith('_') and + callable(getattr(model_instance, m))] + + for member_name in callables: + m = getattr(model_instance, member_name, None) + setattr(self, '%s_%s' % (method_prefix, member_name), m) -- 1.8.4.2

3 5