March 2015 - Kimchi-devel - oVirt List Archives

RFC - Guest memory usage
by Daniel Henrique Barboza 27 Oct '15

27 Oct '15

I was looking at the following item in the backlog: Guests Stats: Display memory utilization (use virt-df or virt-top or ...) If I understood it right, the idea here is to show the -inner- memory allocation of the guest. If you have a VM with 4Gb of RAM running an Ubuntu, we want to know how much memory the Ubuntu OS and its processes are using. I've done an investigation and I haven't found any tool to accomplish this. "virt-top", "virsh dommemstat" and the libvirt API retrieves the information of the memory usage of the guest relative to the host. In the example mentioned before, supposing that the host has 64Gb of RAM, all these tools would show that the VM is using 12% of the host RAM. They do not dive in the VM and shows the actual mem usage of the Ubuntu and its processes running there. Haven't found anything useful in other MLs and forums. The common answer is 'run top in a terminal inside the VM', which of course does not suit us. My question is: any thoughts about how we can implement this feature? Because I am starting to think that, in the end, this kind of info is strict to the guest OS and can't be polled from the outside. Thanks!

5 5

[RFC] Split Kimchi into Base Framework and Virtualization layers
by Lucio Correia 05 May '15

05 May '15

Hi, Gustavo (on cc) and me have been analyzing how to split Kimchi Virtualization layer form its Base web framework. We would like to leverage all infrastructure kimchi base framework provides (URI routing, error handling, i18n, authentication, API validation, plugin tabs loading) in other projects, and contribute this back to kimchi-ginger project. The general idea for the web framework is that it provides basic UI with login and tabs menu, and each plugin installed extends the menu with some tabs. We started by making kimchi a plugin, like ginger is. We are way advanced on that and already have most of kimchi's UI running as a plugin over an almost fully running base web framework. Ginger needs minimal changes, mainly on import statements. We can provide changes needed back to ginger too. Currently, we are splitting kimchi specific stuff like this way: (after renaming the dir src/kimchi to src/backend *) * we are using the name "backend" just for disambiguation from virtualization layer (plugins/kimchi). We will need to define a better name for that basic framework. One suggestion is to call it "wok", the base where kimchi is cooked. {src/backend => plugins/kimchi}/API.json | 0 {src/backend/control/vm => plugins/kimchi/control}/Makefile.am | 7 +- {src/backend/control/vm => plugins/kimchi/control}/__init__.py | 2 +- {src/backend => plugins/kimchi}/control/config.py | 0 {src/backend => plugins/kimchi}/control/cpuinfo.py | 0 {src/backend => plugins/kimchi}/control/debugreports.py | 0 {src/backend => plugins/kimchi}/control/groups.py | 0 {src/backend => plugins/kimchi}/control/host.py | 0 {src/backend => plugins/kimchi}/control/interfaces.py | 0 {src/backend => plugins/kimchi}/control/networks.py | 0 {src/backend => plugins/kimchi}/control/peers.py | 0 {src/backend => plugins/kimchi}/control/storagepools.py | 0 {src/backend => plugins/kimchi}/control/storageservers.py | 0 {src/backend => plugins/kimchi}/control/storagevolumes.py | 0 {src/backend => plugins/kimchi}/control/templates.py | 0 {src/backend => plugins/kimchi}/control/users.py | 0 {src/backend => plugins/kimchi}/control/vm/Makefile.am | 0 {src/backend => plugins/kimchi}/control/vm/__init__.py | 0 {src/backend => plugins/kimchi}/control/vm/hostdevs.py | 0 {src/backend => plugins/kimchi}/control/vm/ifaces.py | 0 {src/backend => plugins/kimchi}/control/vm/snapshots.py | 0 {src/backend => plugins/kimchi}/control/vm/storages.py | 0 {src/backend => plugins/kimchi}/control/vms.py | 0 {src/backend => plugins/kimchi}/disks.py | 0 {src/backend => plugins/kimchi}/imageinfo.py | 0 {src/backend => plugins/kimchi}/iscsi.py | 0 {src/backend => plugins/kimchi}/isoinfo.py | 0 {src/backend => plugins/kimchi}/kvmusertests.py | 0 {src/backend => plugins/kimchi}/mockmodel.py | 0 {src/backend/control/vm => plugins/kimchi/model}/Makefile.am | 7 +- {src/backend/control/vm => plugins/kimchi/model}/__init__.py | 8 --- {src/backend => plugins/kimchi}/model/config.py | 0 {src/backend => plugins/kimchi}/model/cpuinfo.py | 0 {src/backend => plugins/kimchi}/model/debugreports.py | 0 {src/backend => plugins/kimchi}/model/diskutils.py | 0 {src/backend => plugins/kimchi}/model/featuretests.py | 0 {src/backend => plugins/kimchi}/model/groups.py | 0 {src/backend => plugins/kimchi}/model/host.py | 0 {src/backend => plugins/kimchi}/model/hostdev.py | 0 {src/backend => plugins/kimchi}/model/interfaces.py | 0 {src/backend => plugins/kimchi}/model/libvirtconnection.py | 0 {src/backend => plugins/kimchi}/model/libvirtstoragepool.py | 0 plugins/kimchi/model/model.py | 93 ++++++++++++++++++++++++++++ {src/backend => plugins/kimchi}/model/networks.py | 0 {src/backend => plugins/kimchi}/model/peers.py | 0 {src/backend => plugins/kimchi}/model/storagepools.py | 0 {src/backend => plugins/kimchi}/model/storageservers.py | 0 {src/backend => plugins/kimchi}/model/storagetargets.py | 0 {src/backend => plugins/kimchi}/model/storagevolumes.py | 0 {src/backend => plugins/kimchi}/model/templates.py | 0 {src/backend => plugins/kimchi}/model/users.py | 0 {src/backend => plugins/kimchi}/model/utils.py | 0 {src/backend => plugins/kimchi}/model/vmhostdevs.py | 0 {src/backend => plugins/kimchi}/model/vmifaces.py | 0 {src/backend => plugins/kimchi}/model/vms.py | 0 {src/backend => plugins/kimchi}/model/vmsnapshots.py | 0 {src/backend => plugins/kimchi}/model/vmstorages.py | 0 {src/backend => plugins/kimchi}/osinfo.py | 0 {src/backend => plugins/kimchi}/scan.py | 0 {config => plugins/kimchi/ui}/Makefile.am | 0 {config/ui => plugins/kimchi/ui/config}/Makefile.am | 0 config/ui/tabs.xml => plugins/kimchi/ui/config/tab-ext.xml | 0 {src/backend => plugins/kimchi}/vmtemplate.py | 0 The remaining generic stuff on the "backend" side, usable or already in use by other plugins, would be: backend |___model | |___model.py | |___plugins.py | |___tasks.py |___control | |___base.py | |___plugins.py | |___tasks.py | |___utils.py |___xmlutils | |___network.py | |___disk.py | |___utils.py | |___graphics.py | |___qemucmdline.py | |___interface.py |___asynctask.py |___auth.py |___basemodel.py |___cachebust.py |___distroloader.py |___exception.py |___i18n.py |___netinfo.py |___network.py |___objectstore.py |___proxy.py |___repositories.py |___rollbackcontext.py |___root.py |___screenshot.py |___server.py |___sslcert.py |___swupdate.py |___template.py |___utils.py |___vnc.py We don't want to make this a too long e-mail and lose the focus on the idea. We have further technical details to discuss in a follow up. Please let us know your thoughts on that, any directions and requirements you have on this, and how could we proceed with that plan. -- Lucio Correia Software Engineer IBM LTC Brazil

2 3

[RFC] Do not block operations due to ref_cnt
by Crístian Viana 03 Apr '15

03 Apr '15

Hi, I'd like to propose a change to how Kimchi uses the resource field "ref_cnt". Currently, "ref_cnt" - which stands for "reference count" - is one of the fields returned when looking up a storage volume. Its purpose is to indicate how many times that resource is being used at the moment. For example, if the resource /storagepools/pool/storagevolumes/vol has ref_cnt=1, it means that the disk is attached to 1 VM right now and thus it cannot be attached to another VM. I believe the original idea of this feature is to prevent the same resource from being attached more than once at the same time. However, IMO, that might not always be the desired behavior and there's no way to enforce it completely as those resources can be used outside of Kimchi, where "ref_cnt" doesn't exist. For example, if I have one VM which uses the disk "vol", I'm not able to attach it to another VM via Kimchi; but if I use another libvirt-based VM manager (e.g. virsh), I am able to attach that disk to a different VM. This becomes even trickier when we consider other operations, like snapshots, which can attach/detach disks while they're being reverted to. Also, suppose I might want to inspect one VM's disk from another VM, and then I'd need to attach one disk twice; Kimchi wouldn't allow that by stating that the disk is already in use. I propose Kimchi should stop using "ref_cnt" as a blocking method. The field may still exist for information/warning messages (e.g. "are you sure you want to attach this disk? it's already being used by another VM.") but no operation should be blocked because of it, as it is the case now. As inconsistencies with that value may happen and we have no way to make sure it will always work, we shouldn't annoy the user by stopping them from doing something that may be perfectly valid. Any feedback will be very welcome. Best regards, Crístian.

3 3

[kimchi-devel][PATCHv3 0/6] Upload storage volume
by lvroyce＠linux.vnet.ibm.com 03 Apr '15

03 Apr '15

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> known issue: When using api to send requests without setting up session, cherrypy server will report error after hundred of calls, because race condition below: https://bitbucket.org/cherrypy/cherrypy/pull-request/50/fix-race-condition-… But it works well with session setup. v3>v1, Use libvirt upload api instead of r/w file Use lock to synchronize upload to prevent Royce Lv (6): Update docs and json schema of storage volume upload Update controller to make update accept formdata params Add lock facility for storage volume upload Update model for storage volume update Fix incomplete record when uploading update test case for storage volume upload docs/API.md | 7 ++-- src/kimchi/API.json | 19 +++++++++++ src/kimchi/control/base.py | 6 ++-- src/kimchi/i18n.py | 4 +++ src/kimchi/isoinfo.py | 5 ++- src/kimchi/mockmodel.py | 8 ++--- src/kimchi/model/storagevolumes.py | 61 +++++++++++++++------------------- src/kimchi/model/utils.py | 46 ++++++++++++++++++++++++++ tests/test_model.py | 67 ++++++++++++++++++++++++++++++++++++++ 9 files changed, 177 insertions(+), 46 deletions(-) -- 2.1.0

3 20

[PATCH] RFC patch to make nginx proxy optional
by Julien Goodwin 03 Apr '15

03 Apr '15

Sending this as an RFC patch, I expect people will want changes from this. Implements Issue #570. Julien Goodwin (1): Initial prototype, make nginx proxy optional. docs/Makefile.am | 1 + docs/apache.conf.ex | 35 +++++++++++++++++++++++++++++++++++ src/kimchi.conf.in | 3 +++ src/kimchi/config.py.in | 1 + src/kimchi/proxy.py | 6 ++++++ 5 files changed, 46 insertions(+) create mode 100644 docs/apache.conf.ex -- 2.1.4

4 18

[RFC] New API to create a random guest console password
by Aline Manera 02 Apr '15

02 Apr '15

Today user may set/change the guest console password and also its expiration time through Kimchi API. When passing an empty password, a random password is automatically generated. curl -u <user:password> -H "Content-Type: application/json" -H "Accept: application/json" http://localhost:8010/vms/blah -X PUT -d'{"graphics": {"passwd": ""}}' That way is difficult to handle when user wants to reset the guest password. We have a similar issue when we automatically change the passwdValidTo when it is expired - increasing it in 30 seconds. My proposal is simple: only change "passwd" and "passwdValidTo" when user wants to do it. curl -u <user:password> -H "Content-Type: application/json" -H "Accept: application/json" http://localhost:8010/vms/blah -X PUT -d'{"graphics": {"passwd": "123456", "passwdValidTo": "<some datetime format>"}}' curl -u <user:password> -H "Content-Type: application/json" -H "Accept: application/json" http://localhost:8010/vms/blah -X PUT -d'{"graphics": {"passwdValidTo": "<some datetime format>"}}' And make sure the passwdValidTo is only acceptable when there is a passwd set. And to reset those values, we only need to send an empty string: curl -u <user:password> -H "Content-Type: application/json" -H "Accept: application/json" http://localhost:8010/vms/blah -X PUT -d'{"graphics": {"password": "", "passwdValidTo": ""}}' And create a new API: POST /vms/blah/ticket to automatically generate a random password valid only for 30 seconds. curl -u <user:password> -H "Content-Type: application/json" -H "Accept: application/json" http://localhost:8010/vms/blah/ticket -X POST -d'{}' What do you think about it? Regards, Aline Manera

2 3

[RFC] Hotplug a new VM disk
by Crístian Viana 01 Apr '15

01 Apr '15

Hi, I'd like to propose a new feature to hotplug a new disk to a VM. This feature will be implemented only on the UI side. The backend already has the methods needed to make it work. Suppose a user wants to add a new disk to a running VM. Currently, they have to create an empty disk using other tools outside of Kimchi and put it into an existing storage pool. This might be challenging for some users, as it requires the use of other tools, possibly involving the command line (e.g. qemu-img). With this new feature, the user will only select the storage pool and the size of the new disk they want to add; Kimchi will create the disk in the appropriate pool and attach it to the VM. In order to perform this feature, the UI will have to be updated to allow the following steps: 1) Select an existing VM, "Actions" > "Edit". 2) Switch to the tab "Storage". 3) Click on the button "+". 4) On "Device type", select "disk" . 5) On "Storage pool", select any storage pool. *6) On "Storage volume", select "New...".** **7) Enter the desired disk size.* 8) Click on the button "Attach". Steps 6 and 7 are the ones related to this feature and they need to be implemented. The other steps already exist in the current UI. When the user performs step 8, the UI should send the following requests to the server: POST /storagepools/<pool>/storagevolumes {'name': <vol>, 'capacity': <size>} POST /vms/<vm>/storages {'name': <vol>} Those requests already work, they just have to be executed in that order in order to perform this new feature. Any feedback will be very welcome. Best regards, Crístian.

2 1

[RFC] Hotplug a VM network interface
by Crístian Viana 01 Apr '15

01 Apr '15

Hi, I'd like to propose a new feature to hotplug a network interface to a running VM. A similar feature is already available for stopped VMs by selecting a VM and clicking on "Actions" > "Edit", tab "Interface"; the user is able to add or remove a network interface in that dialog. However, if the VM is running, that tab disables the buttons to add and remove network interfaces. This feature will allow those two operations to be performed also on running VMs. The API for adding and removing network interfaces to a VM already exist: POST /vms/<vm>/ifaces {<iface data>} DELETE /vms/<vm>/ifaces/<iface> However, they currently raise an exception when the VM is running. This feature will use those same commands in order to hotplug a network interface, keeping the existing REST interface. Any feedback will be very welcome. Best regards, Crístian.

2 1

[RFC] Add suport for VM suspend/resume
by Crístian Viana 01 Apr '15

01 Apr '15

Hi, I'd like to propose two new operations on virtual machines: *suspend* and *resume*: Suspend ---------- From libvirt.org: /Suspend an active domain, the process is frozen without further access to CPU resources and I/O but the memory used by the domain at the hypervisor level will stay allocated./ REST syntax: POST /vms/<vm-name>/suspend Return status: 200: OK 4xx: error Return data: None Resume ---------- From libvirt.org: /Resume a suspended domain, the process is restarted from the state where it was frozen by calling "suspend"./ REST syntax: POST /vms/<vm-name>/resume Return status: 200: OK 4xx: error Return data: None -------------------- This feature has been partially implemented in previous patches ("[PATCH 0/3] VM suspend/resume", from Jan 21st), and its progress should be resumed soon. The UI will have to be updated accordingly by adding buttons to suspend and resume a VM, somewhat like the buttons which start and stop a VM. Any feedback will be very welcome. Best regards, Crístian.

2 1

[PATCH v2 0/3] Fix issue #461 (backend)
by Crístian Viana 01 Apr '15

01 Apr '15

This is the changelog between this and the previous patchset: - Don't allow mirrorlist and metalink to be specified at the same time; according to the YUM documentation, that situation is invalid. Crístian Viana (3): Handle empty variables when updating YUM repository Use more generic message in repo mirror list error issue #461: Add 'metalink' support for YUM repositories src/kimchi/API.json | 5 +++++ src/kimchi/i18n.py | 6 ++++-- src/kimchi/repositories.py | 31 +++++++++++++++++++++++++------ 3 files changed, 34 insertions(+), 8 deletions(-) -- 2.1.0

2 4

[PATCH V2] Update host number of cpus and total physical memory
by Jose Ricardo Ziviani 01 Apr '15

01 Apr '15

- The number of cpus and the total physical memory information displayed in the host section can change, so them cannot be static. This commit updates those values whenever a GET request hits the backend looking for host information. --- src/kimchi/model/host.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/src/kimchi/model/host.py b/src/kimchi/model/host.py index 4419bb3..588bb28 100644 --- a/src/kimchi/model/host.py +++ b/src/kimchi/model/host.py @@ -87,8 +87,8 @@ class HostModel(object): res['cpu_model'] = line.split(':')[1].strip() break - res['cpus'] = psutil.NUM_CPUS - res['memory'] = psutil.TOTAL_PHYMEM + res['cpus'] = 0 + res['memory'] = 0L # Include IBM PowerKVM name to supported distro names _sup_distros = platform._supported_dists + ('ibm_powerkvm',) @@ -102,6 +102,8 @@ class HostModel(object): return res def lookup(self, *name): + self.host_info['cpus'] = psutil._psplatform.get_num_cpus() + self.host_info['memory'] = psutil.phymem_usage().total return self.host_info def swupdate(self, *name): -- 1.9.1

3 3

[PATCH V3] Number of CPUs and Physical Memory
by Jose Ricardo Ziviani 01 Apr '15

01 Apr '15

V3: psutil has changed its interface to retrieve the number of CPUs, this new V3 updates the code while supporting the old version (still needed). Jose Ricardo Ziviani (1): Update host number of cpus and total physical memory src/kimchi/model/host.py | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) -- 1.9.1

2 2

[PATCH][RFC] Kimchi and relative URLs
by Frédéric Bonnard 01 Apr '15

01 Apr '15

From: Frederic Bonnard <frediz(a)linux.vnet.ibm.com> Hi, many people and distros use a subdirectory in the configuration of the system's webserver to configure different websites. The usual way to setup a website is thus to : 1. use the distro's webserver to serve all the websites 2. have a subdirectory to copy the configuration file that makes available a website by only restarting the webserver (and removing the website is just done by removing the configuration file and restart). Concerning this point, the classical configurations can be : a) a sub web location : https://website.org/webapp b) a virtual host : http://webapp.website.org a) and b) correspond each to different configuration files that are just dropped into : - Ubuntu/Debian : /etc/{apache,nginx,..}/sites-available/{webapp_subweb,webapp_virthost}.conf - Fedora/Opensuse : /etc/{httpd,nginx,..}/conf.d/ At the moment kimchi is launched with a private instance of nginx, not the system's installed one and it would be nice to have this improved and this is covered in another thread : http://lists.ovirt.org/pipermail/kimchi-devel/2015-February/009642.html Provided that previous patch, one can use the distro's webserver to run proxy kimchi and that fullfills 1. Then for 2. one need a configuration file for either b) (provided for apache in the link above ; I have also one for nginx ) and a), which I provide in this patch. But this nginx configuration file is not enough for a sub web location configuration as kimchi relies on absolute path based on / from the http://server/. I tried to make URLs paths relative so that kimchi doesn't have to know where it has been placed on the webserver, for example https://server/kimchi/ which will probably be the URL used in distros. I'm not a web developer and I tried to modify all failing requests with a a) configuration with relative paths in mind. So I'd like comments on the way this is done, and this may need more extensive testing (paths I didn't test). Though I used that code back on a virtualhost configuration and it worked as well. Thanks for your help, F. Frederic Bonnard (1): Making urls relative docs/Makefile.am | 1 + docs/nginx.conf.subsite.ex | 37 +++++++ src/kimchi/screenshot.py | 2 +- ui/css/theme-default/template_add.css | 20 ++-- ui/css/theme-default/topbar.css | 2 +- ui/js/src/kimchi.api.js | 176 +++++++++++++++++----------------- ui/js/src/kimchi.login.js | 2 +- ui/pages/guest.html.tmpl | 2 +- ui/pages/help/dita-help.xsl | 4 +- ui/pages/kimchi-ui.html.tmpl | 4 +- ui/pages/storagepool-add.html.tmpl | 2 +- ui/pages/tabs/storage.html.tmpl | 2 +- ui/pages/template-add.html.tmpl | 2 +- 13 files changed, 147 insertions(+), 109 deletions(-) create mode 100644 docs/nginx.conf.subsite.ex -- 1.9.1

4 9

[PATCH v3] test_model: Power arch fixes
by Daniel Henrique Barboza 01 Apr '15

01 Apr '15

changes in v3: - rebase to latest master changes in v2: - changed iso_path to reflect the new ISO being created (Ubuntu 14.04) Daniel Henrique Barboza (1): test/test_model: Power architecture fixes tests/iso_gen.py | 3 ++- tests/test_model.py | 12 +++++++----- 2 files changed, 9 insertions(+), 6 deletions(-) -- 1.9.3

2 3

[kimchi-devel][PATCH] Issue #623: Fix mismatch in host device details query
by lvroyce＠linux.vnet.ibm.com 01 Apr '15

01 Apr '15

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> Commit #902d2957 began to use splitted devices names, while get detailed device still use original names. Mismatched name results in error raised when listing host devices. Fix it by using splitted device name as well. Signed-off-by: Royce Lv <lvroyce(a)linux.vnet.ibm.com> --- src/kimchi/disks.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/kimchi/disks.py b/src/kimchi/disks.py index a092c17..5cf6885 100644 --- a/src/kimchi/disks.py +++ b/src/kimchi/disks.py @@ -64,7 +64,7 @@ def _get_dev_major_min(name): dev_list = _get_lsblk_devs(keys) for dev in dev_list: - if dev['name'] == name: + if dev['name'].split()[0] == name: maj_min = dev['maj:min'] break else: -- 1.9.1

2 2

[PATCH 0/3] Fix backend VM, POOL, NETWORK naming validation
by Rodrigo Trujillo 01 Apr '15

01 Apr '15

This patchset fixes the backend JSON name checkings when user creates a new VM, POOL or NETWORK. Rodrigo Trujillo (3): Prohibits slashes '/' in VM name Move slash checking in storagepool name from UI to backend Fix Network create name checking in backend (slashes and quotes) src/kimchi/API.json | 4 ++++ src/kimchi/i18n.py | 6 +++--- src/kimchi/model/utils.py | 3 ++- ui/js/src/kimchi.storagepool_add_main.js | 5 ----- ui/pages/i18n.json.tmpl | 1 - 5 files changed, 9 insertions(+), 10 deletions(-) -- 2.1.0

2 7

[PATCH] Github issue #630: fixing None.strip() error
by Daniel Henrique Barboza 31 Mar '15

31 Mar '15

This simple patch fixes the scenario in repositories.py where baseurl or mirrorlist is None and the code was executing baseurl.strip() and mirrorlist.strip(), leading to errors in tests/test_model.py (test_repository_update) and probably other bugs. Signed-off-by: Daniel Henrique Barboza <dhbarboza82(a)gmail.com> --- src/kimchi/repositories.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/kimchi/repositories.py b/src/kimchi/repositories.py index f1e1eb3..19c5810 100644 --- a/src/kimchi/repositories.py +++ b/src/kimchi/repositories.py @@ -275,10 +275,10 @@ class YumRepo(object): config = params.get('config', {}) mirrorlist = config.get('mirrorlist', None) - if len(baseurl.strip()) == 0: + if baseurl is not None and len(baseurl.strip()) == 0: baseurl = None - if len(mirrorlist.strip()) == 0: + if mirrorlist is not None and len(mirrorlist.strip()) == 0: mirrorlist = None if baseurl is None and mirrorlist is None: -- 1.9.3

2 3

[PATCH] issue #628: Set ref_cnt=1 on cloned disks
by Crístian Viana 31 Mar '15

31 Mar '15

Cloned disks currently don't have ref_cnt stored locally, so they may be attached again to another VM - or even to the same one. When cloning a VM, set their disks' ref_cnt to 1 to prevent them from being reattached. Fix issue #628 (Storage volume ref_cnt is not updated after vm clone). Signed-off-by: Crístian Viana <vianac(a)linux.vnet.ibm.com> --- src/kimchi/model/vms.py | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/src/kimchi/model/vms.py b/src/kimchi/model/vms.py index 86c173f..af5e765 100644 --- a/src/kimchi/model/vms.py +++ b/src/kimchi/model/vms.py @@ -449,6 +449,10 @@ class VMModel(object): vir_conn = self.conn.get() + def _delete_disk_from_objstore(path): + with self.objstore as session: + session.delete('storagevolume', path) + for i, path in enumerate(all_paths): try: vir_orig_vol = vir_conn.storageVolLookupByPath(path) @@ -518,6 +522,11 @@ class VMModel(object): xml = xml_item_update(xml, XPATH_DOMAIN_DISK_BY_FILE % path, new_vol['path'], 'file') + # set the new disk's ref_cnt + with self.objstore as session: + session.store('storagevolume', new_vol['path'], {'ref_cnt': 1}) + rollback.prependDefer(_delete_disk_from_objstore, new_vol['path']) + # remove the new volume should an error occur later rollback.prependDefer(self.storagevolume.delete, new_pool_name, new_vol_name) -- 2.1.0

2 2

[PATCH 0/3] Fix issue #461 (backend)
by Crístian Viana 31 Mar '15

31 Mar '15

This patchset fixes the backend part of issue #461. The UI should also be updated in order to use the metadata property. Other related fixes have been committed as well. Crístian Viana (3): Handle empty variables when updating YUM repository Use more generic message in repo mirror list error issue #461: Add 'metalink' support for YUM repositories src/kimchi/API.json | 5 +++++ src/kimchi/i18n.py | 5 +++-- src/kimchi/repositories.py | 25 +++++++++++++++++++------ 3 files changed, 27 insertions(+), 8 deletions(-) -- 2.1.0

3 9

RFC - New option 'Serial Console' in the template UI
by Jose Ricardo Ziviani 30 Mar '15

30 Mar '15

Hello everybody! I'd like to hear from you if it would be nice to have a new option in the Template UI to redirect the output to a serial console. Today, there is a field called 'Graphics' with two options [VNC, Spice]. What if we have [VNC, Spice, Serial Console (no-graphic)]? A drawback is the need of an external tool (like virsh console) to be able to take the console, as Kimchi doesn't provide one. What do you think? Thank you! -- Jose Ricardo Ziviani ----------------------------- Software Engineer Linux Technology Center - IBM

7 9

[PATCH] Fix a ZeroDivisionError when starting kimchi service in Qemu.
by Jose Ricardo Ziviani 30 Mar '15

30 Mar '15

When using kimchi inside a virtual machine it's possible to get a ZeroDivisionError exception because self.sockets is 0, so the division self.cores_per_socket = self.cores_present/self.sockets will fail. In this case, self.sockets is 0 because cores_present = 1 and threads_per_core = 4, which will truncate 0.25 to 0 during the integer division. This is a low priority corner case since kimchi is obviously not supposed to run under a virtual machine but for a plugin development, which will not use any virtualization capability, it might become handy. Signed-off-by: Jose Ricardo Ziviani <joserz(a)linux.vnet.ibm.com> --- src/kimchi/model/cpuinfo.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/kimchi/model/cpuinfo.py b/src/kimchi/model/cpuinfo.py index 3411ef5..d063425 100644 --- a/src/kimchi/model/cpuinfo.py +++ b/src/kimchi/model/cpuinfo.py @@ -84,6 +84,8 @@ class CPUInfoModel(object): if not rc: self.threads_per_core = int(out.split()[-1]) self.sockets = self.cores_present/self.threads_per_core + if self.sockets == 0: + self.sockets = 1 self.cores_per_socket = self.cores_present/self.sockets else: # Intel or AMD -- 1.9.1

2 2

Kimchi 1.4.1 is released!
by Aline Manera 27 Mar '15

27 Mar '15

On behalf of everyone who has worked hard on this release, I am pleased to announce the availability of *Kimchi 1.4.1*! This release adds many new features including: ✔ Bug fixes ✔ Add robot.txt file ✔ Disable SSLv3 ✔ Recognize Arch Linux ISO image ✔ Basic ppc65el support ✔ Add man page for kimchid command We have worked hard to ensure that Kimchi runs well on the most popular Linux distributions including: Fedora 21, Ubuntu 14.10, openSUSE 13.2, and RHEL 7.1. Kimchi uses standard Linux interfaces so it should run well on many other distributions too. You can easily grab this release in tarball format or via git: ✔https://github.com/kimchi-project/kimchi/archive/1.4.1.tar.gz <https://github.com/kimchi-project/kimchi/archive/1.4.0.tar.gz> ✔ git clonehttps://github.com/kimchi-project/kimchi.git There are also some packages available at: ✔http://kimchi-project.github.io/kimchi/downloads/ Go ahead! Give it a try and let us know what you think! Regards, Aline Manera

1 0

[kimchi-devel][PATCHv2] Fix update vcpu count: Check available vcpu count before update
by lvroyce＠linux.vnet.ibm.com 26 Mar '15

26 Mar '15

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> Currently we just check vcpu count does not exceed host available count when topology specified. Actually vcpu count also need to be checked when only vcpu specified. Move this check outside topology check and add it to param update. Signed-off-by: Royce Lv <lvroyce(a)linux.vnet.ibm.com> --- src/kimchi/model/cpuinfo.py | 8 ++++++-- src/kimchi/model/templates.py | 6 +++--- tests/test_model.py | 2 +- tests/test_rest.py | 1 - 4 files changed, 10 insertions(+), 7 deletions(-) diff --git a/src/kimchi/model/cpuinfo.py b/src/kimchi/model/cpuinfo.py index 3411ef5..54bfad4 100644 --- a/src/kimchi/model/cpuinfo.py +++ b/src/kimchi/model/cpuinfo.py @@ -114,11 +114,15 @@ class CPUInfoModel(object): cores = topology['cores'] threads = topology['threads'] + self.check_vcpus(vcpus) + if not self.guest_threads_enabled: raise InvalidOperation("KCHCPUINF0003E") if vcpus != sockets * cores * threads: raise InvalidParameter("KCHCPUINF0002E") - if vcpus > self.cores_available * self.threads_per_core: - raise InvalidParameter("KCHCPUINF0001E") if threads > self.threads_per_core: raise InvalidParameter("KCHCPUINF0002E") + + def check_vcpus(self, vcpus): + if vcpus > self.cores_available * self.threads_per_core: + raise InvalidParameter("KCHCPUINF0001E") diff --git a/src/kimchi/model/templates.py b/src/kimchi/model/templates.py index 6bc8aca..14e5b37 100644 --- a/src/kimchi/model/templates.py +++ b/src/kimchi/model/templates.py @@ -53,7 +53,7 @@ class TemplatesModel(object): {'filename': iso, 'user': user, 'err': excp}) - cpu_info = params.get('cpu_info') + cpu_info = params.setdefault('cpu_info', dict()) if cpu_info: topology = cpu_info.get('topology') # Check, even though currently only topology @@ -68,8 +68,8 @@ class TemplatesModel(object): # exception if a topology is invalid. CPUInfoModel(conn=self.conn).\ check_topology(params['cpus'], topology) - else: - params['cpu_info'] = dict() + if params.get('cpus'): + CPUInfoModel(conn=self.conn).check_vcpus(params['cpus']) conn = self.conn.get() pool_uri = params.get(u'storagepool', '') diff --git a/tests/test_model.py b/tests/test_model.py index 0020022..138a7da 100644 --- a/tests/test_model.py +++ b/tests/test_model.py @@ -577,7 +577,7 @@ class ModelTests(unittest.TestCase): inst = model.Model(None, objstore_loc=self.tmp_store) - orig_params = {'name': 'test', 'memory': '1024', 'cpus': '1', + orig_params = {'name': 'test', 'memory': '1024', 'cpus': 1, 'cdrom': UBUNTU_ISO} inst.templates_create(orig_params) diff --git a/tests/test_rest.py b/tests/test_rest.py index 07e5733..4ecf3ce 100644 --- a/tests/test_rest.py +++ b/tests/test_rest.py @@ -932,7 +932,6 @@ class RestTests(unittest.TestCase): task_info = model.storagevolumes_create('pool-3', params) wait_task(self._task_lookup, task_info['id']) - storagevolume = json.loads(self.request( '/storagepools/kimchi_isos/storagevolumes/').read())[0] self.assertEquals('fedora.iso', storagevolume['name']) -- 1.9.1

2 1

[PATCH] Fix PowerOptions availability
by Jose Ricardo Ziviani 26 Mar '15

26 Mar '15

- When we run tuned-adm without parameters it returns code 2, which will fail our check condition and Power Options would never be displayed. This commit add a parameter that simply list the profiles, this will return 0. --- models/powermanagement.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/powermanagement.py b/models/powermanagement.py index 0bb3dcc..7600024 100644 --- a/models/powermanagement.py +++ b/models/powermanagement.py @@ -62,7 +62,7 @@ class PowerProfilesModel(object): return profiles def is_feature_available(self): - output, error, returncode = run_command(['tuned-adm']) + output, error, returncode = run_command(['tuned-adm', 'list']) return returncode == 0 -- 1.9.1

1 1

[PATCH v] Making urls relative
by Frédéric Bonnard 26 Mar '15

26 Mar '15

From: Frederic Bonnard <frediz(a)linux.vnet.ibm.com> On Wed, 11 Mar 2015 12:39:48 -0300, Aline Manera <alinefm(a)linux.vnet.ibm.com> wrote > My point is: if the main idea behind it is to allow user to connects > > Kimchi to the external web server, why do not add an option to do it for > > > the user? > 2) Add an option to Kimchi config file to allow user connects to an > > external web server. I'm a bit confused : do you mean, why don't we add an option so that kimchi starts nginx ? If this is your question, in a distro we would have kimchi package have a dependency on [nginx | apache]. So when you would install kimchi, it would install a web server as well and distros usually starts all services once installed, so, nginx would be already started. That is why, we would need an option, not to start the web server instead. I'm not sure I got you point :) sorry. In the same topic, that's why I'm using the patch in thread "RFC patch to make nginx proxy optional" [1] > And yes - we launch nginx (if we should or not do that it is other > > discussion) and we will continue doing that until something changes what > > > is not in our plans by now. I'm not willing to change that default :) . It'll just be nice to have an option to change that behaviour in other environments [1]. Default would be as it already is, but we could specify with that option to disable launching an external webserver or specify another one like apache. as Royce suggested [2] ? Here follows the patch without the configuration file I put the previous time. F. 1: http://lists.ovirt.org/pipermail/kimchi-devel/2015-February/009642.html 2: http://lists.ovirt.org/pipermail/kimchi-devel/2015-February/009655.html Frederic Bonnard (1): Making urls relative src/kimchi/screenshot.py | 2 +- ui/css/theme-default/template_add.css | 20 ++-- ui/css/theme-default/topbar.css | 2 +- ui/js/src/kimchi.api.js | 176 +++++++++++++++++----------------- ui/js/src/kimchi.login.js | 2 +- ui/pages/guest.html.tmpl | 2 +- ui/pages/help/dita-help.xsl | 4 +- ui/pages/kimchi-ui.html.tmpl | 4 +- ui/pages/storagepool-add.html.tmpl | 2 +- ui/pages/tabs/storage.html.tmpl | 2 +- ui/pages/template-add.html.tmpl | 2 +- 11 files changed, 109 insertions(+), 109 deletions(-) -- 1.9.1

2 5

[PATCH] Update host number of cpus and total physical memory information
by Jose Ricardo Ziviani 23 Mar '15

23 Mar '15

- The number of cpus and the total physical memory information displayed in the host section can change, so them cannot be static. This commit updates those values whenever a GET request hits the backend looking for host information. Signed-off-by: Jose Ricardo Ziviani <joserz(a)linux.vnet.ibm.com> --- src/kimchi/model/host.py | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/src/kimchi/model/host.py b/src/kimchi/model/host.py index 4419bb3..db14a28 100644 --- a/src/kimchi/model/host.py +++ b/src/kimchi/model/host.py @@ -87,8 +87,9 @@ class HostModel(object): res['cpu_model'] = line.split(':')[1].strip() break - res['cpus'] = psutil.NUM_CPUS - res['memory'] = psutil.TOTAL_PHYMEM + res['cpus'] = 0 + res['memory'] = 0L + self._update_host_info(res) # Include IBM PowerKVM name to supported distro names _sup_distros = platform._supported_dists + ('ibm_powerkvm',) @@ -101,7 +102,15 @@ class HostModel(object): return res + def _update_host_info(self, host_info): + if not isinstance(host_info, dict): + return + + host_info['cpus'] = psutil._psplatform.get_num_cpus() + host_info['memory'] = psutil.phymem_usage().total + def lookup(self, *name): + self._update_host_info(self.host_info) return self.host_info def swupdate(self, *name): -- 1.9.1

3 5

Entering hard code freeze for 1.4.1 release
by Aline Manera 23 Mar '15

23 Mar '15

Hi all, We are entering the hard code freeze phase for 1.4.1 release. Which means only critical bug fixes will be merged in the next days. Please, check which issues are still targeted for this release at: https://github.com/kimchi-project/kimchi/milestones/Kimchi%201.4.1 I'd also want to ask your help to update the translations files. Thanks, Aline Manera

2 3

[PATCH 0/3 v2] unit tests - Power arch fixes
by Daniel Henrique Barboza 20 Mar '15

20 Mar '15

v2: - make the import change in test_osinfo.py - changed self.kimchi_iso path in test_model.py This patch set contains fixes to several unit tests that had somewhat the same root cause - the differences between x86 and power/ppc64le in osinfo.py template configurations. Daniel Henrique Barboza (3): tests/test_osinfo.py: fixes for Power architecture Kimchi tests: Power system fixes - removing hardcoded values test/test_model: Power architecture fixes src/kimchi/osinfo.py | 9 +++++++++ tests/iso_gen.py | 3 ++- tests/test_mockmodel.py | 4 +++- tests/test_model.py | 12 +++++++----- tests/test_osinfo.py | 36 ++++++++++++++++++++++++++---------- tests/test_rest.py | 6 ++++-- tests/test_vmtemplate.py | 8 ++++++-- 7 files changed, 57 insertions(+), 21 deletions(-) -- 1.9.3

2 5

[PATCH] Specify user when changing VM disks permission
by Crístian Viana 20 Mar '15

20 Mar '15

Before starting a virtual machine, Kimchi updates the VM disks permissions so libvirt can read them appropriately. However, the correct user isn't being passed as a parameter, and the function never actually changes the file's ACL to something useful. Set the correct user when changing disk files' permissions. Signed-off-by: Crístian Viana <vianac(a)linux.vnet.ibm.com> --- src/kimchi/model/vms.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/src/kimchi/model/vms.py b/src/kimchi/model/vms.py index 4c5f443..86c173f 100644 --- a/src/kimchi/model/vms.py +++ b/src/kimchi/model/vms.py @@ -33,6 +33,7 @@ from kimchi import model, vnc from kimchi.config import READONLY_POOL_TYPE, config from kimchi.exception import InvalidOperation, InvalidParameter from kimchi.exception import NotFoundError, OperationFailed +from kimchi.kvmusertests import UserTests from kimchi.model.config import CapabilitiesModel from kimchi.model.tasks import TaskModel from kimchi.model.templates import TemplateModel @@ -946,8 +947,10 @@ class VMModel(object): xml = dom.XMLDesc(0) xpath = "/domain/devices/disk[@device='cdrom']/source/@file" isofiles = xpath_get_text(xml, xpath) + + user = UserTests.probe_user() for iso in isofiles: - run_setfacl_set_attr(iso) + run_setfacl_set_attr(iso, user=user) dom = self.get_vm(name, self.conn) try: -- 2.1.0

2 1

[PATCH] issue #518: Simplify template URL verification
by Crístian Viana 20 Mar '15

20 Mar '15

When using Kimchi via the UI, the browser checks whether the template URL is valid using a regular expression. However, that regexp doesn't match every valid URL out there and it causes problems on some browsers (e.g. the latest Firefox ESR). Use a simpler URL validation when creating new templates (i.e. only validate the possible web protocols). Even if a non-existing/invalid URL passes this validation, it will still be checked in the backend. Fix issue #518 (Unable to download ISO to storage pool on Firefox 31.3.0 ESR). Signed-off-by: Crístian Viana <vianac(a)linux.vnet.ibm.com> --- ui/js/src/kimchi.template_add_main.js | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/ui/js/src/kimchi.template_add_main.js b/ui/js/src/kimchi.template_add_main.js index 6223c96..46c4f84 100644 --- a/ui/js/src/kimchi.template_add_main.js +++ b/ui/js/src/kimchi.template_add_main.js @@ -425,19 +425,9 @@ kimchi.template_add_main = function() { }; kimchi.template_check_url = function(url) { - var protocols = "((https|http|ftp|ftps|tftp)?://)?", - userinfo = "(([0-9a-z_!~*'().&=+$%-]+:)?[0-9a-z_!~*'().&=+$%-]+@)?", - ip = "(\\d{1,3}\.){3}\\d{1,3}", - domain = "([0-9a-z_!~*'()-]+\.)*([0-9a-z][0-9a-z-]{0,61})?[0-9a-z]\.[a-z]{2,6}", - port = "(:\\d{1,5})?", - address = "(/[\\w!~*'().;?:@&=+$,%#-]+)+", - domaintype = [ protocols, userinfo, domain, port, address ], - ipType = [ protocols, userinfo, ip, port, address ], - validate = function(type) { - return new RegExp('^' + type.join('') + '$'); - }; + var reg = /(https|http|ftp|ftps|tftp):\/\//; if (url.constructor === String) { - return validate(domaintype).test(url) || validate(ipType).test(url); + return reg.test(url); } return false; }; -- 2.1.0

2 1

[PATCH v2 0/3] Fix storage volume issues
by Crístian Viana 20 Mar '15

20 Mar '15

This is the changelog between this and the previous patchset: - Instead of removing the ISO permission check at all, update it so the file's ACL is correctly updated before checking its permission; Crístian Viana (3): issue #565: Allow a template's ISO to be a block device issue #564: Parse logical volumes to find out their actual formats Update ISO file's ACL before checking its permission src/kimchi/isoinfo.py | 7 +++++-- src/kimchi/model/storagevolumes.py | 19 ++++++++++++++++++- src/kimchi/model/templates.py | 20 ++++++++++++-------- src/kimchi/vmtemplate.py | 10 ++++++++-- 4 files changed, 43 insertions(+), 13 deletions(-) -- 2.1.0

2 4

[PATCH 0/2] Fix ISO streaming and tests cases
by Aline Manera 20 Mar '15

20 Mar '15

Depends on [Kimchi-devel] [PATCH] Fix issue #597: Get the right arch when using libvirt Test driver Aline Manera (2): Bug fix: Consider server is being started up on servermethod function Fix test_model.py to reset Singleton classes when initializing the tests src/kimchi/utils.py | 3 ++- tests/test_model.py | 51 ++++++++++++++++++++++++++++++++++----------------- 2 files changed, 36 insertions(+), 18 deletions(-) -- 2.1.0

3 5

[PATCH] Use the right firewall-cmd command to open Kimchi ports
by Aline Manera 20 Mar '15

20 Mar '15

When trying to open Kimchi ports following the instructions on README on RHEL 7.1 I got the following error: [root@rhel7-vm ~]# sudo firewall-cmd --add-port=8000 --permanent bad port (most likely missing protocol), correct syntax is portid[-portid]/protocol The same behavior was identified on Fedora 21. So update the firewall-cmd commands to properly guide user on opening Kimchi ports. Signed-off-by: Aline Manera <alinefm(a)linux.vnet.ibm.com> --- docs/README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/README.md b/docs/README.md index 4b14814..5407264 100644 --- a/docs/README.md +++ b/docs/README.md @@ -145,9 +145,9 @@ If you cannot access Kimchi, take a look at these 2 points: Kimchi uses by default the ports 8000, 8001 and 64667. To allow incoming connections: For system using firewalld, do: - sudo firewall-cmd --add-port 8000 --permanent - sudo firewall-cmd --add-port 8001 --permanent - sudo firewall-cmd --add-port 64667 --permanent + sudo firewall-cmd --add-port=8000/tcp --permanent + sudo firewall-cmd --add-port=8001/tcp --permanent + sudo firewall-cmd --add-port=64667/tcp --permanent sudo firewall-cmd --reload For openSUSE systems, do: -- 2.1.0

2 2

[PATCH] Fix issue #597: Get the right arch when using libvirt Test driver
by Aline Manera 20 Mar '15

20 Mar '15

libvirt Test driver only supports i686 arch. So fix feature tests to get the right arch according to libvirt driver. Signed-off-by: Aline Manera <alinefm(a)linux.vnet.ibm.com> --- src/kimchi/model/featuretests.py | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/src/kimchi/model/featuretests.py b/src/kimchi/model/featuretests.py index f22fada..4aac8ed 100644 --- a/src/kimchi/model/featuretests.py +++ b/src/kimchi/model/featuretests.py @@ -20,13 +20,13 @@ import cherrypy import libvirt import lxml.etree as ET +import platform import socket import subprocess import threading from lxml.builder import E -from kimchi import osinfo from kimchi.rollbackcontext import RollbackContext from kimchi.utils import kimchi_log, run_command, servermethod @@ -102,8 +102,9 @@ class FeatureTests(object): @staticmethod def libvirt_supports_iso_stream(conn, protocol): - domain_type = 'test' if conn.getType().lower() == 'test' else 'kvm' - arch = osinfo.defaults['arch'] + conn_type = conn.getType().lower() + domain_type = 'test' if conn_type == 'test' else 'kvm' + arch = 'i686' if conn_type == 'test' else platform.machine() arch = 'ppc64' if arch == 'ppc64le' else arch xml = ISO_STREAM_XML % {'name': FEATURETEST_VM_NAME, 'domain': domain_type, 'protocol': protocol, @@ -198,8 +199,9 @@ class FeatureTests(object): with RollbackContext() as rollback: FeatureTests.disable_libvirt_error_logging() rollback.prependDefer(FeatureTests.enable_libvirt_error_logging) - domain_type = 'test' if conn.getType().lower() == 'test' else 'kvm' - arch = osinfo.defaults['arch'] + conn_type = conn.getType().lower() + domain_type = 'test' if conn_type == 'test' else 'kvm' + arch = 'i686' if conn_type == 'test' else platform.machine() arch = 'ppc64' if arch == 'ppc64le' else arch dom = conn.defineXML(SIMPLE_VM_XML % {'name': FEATURETEST_VM_NAME, 'domain': domain_type, -- 2.1.0

3 3

[PATCH] Fix issue #617: Base URL may be an empty string when updating YUM repo
by Aline Manera 20 Mar '15

20 Mar '15

Signed-off-by: Aline Manera <alinefm(a)linux.vnet.ibm.com> --- src/kimchi/i18n.py | 2 +- src/kimchi/repositories.py | 16 +++++++++++----- 2 files changed, 12 insertions(+), 6 deletions(-) diff --git a/src/kimchi/i18n.py b/src/kimchi/i18n.py index df5422f..e4e1a89 100644 --- a/src/kimchi/i18n.py +++ b/src/kimchi/i18n.py @@ -298,7 +298,7 @@ messages = { "KCHREPOS0010E": _("GPG key must be a URL pointing to the ASCII-armored file."), "KCHREPOS0011E": _("Could not update repository %(repo_id)s."), "KCHREPOS0012E": _("Repository %(repo_id)s does not exist."), - "KCHREPOS0013E": _("Specify repository base URL or mirror list in order to create a YUM repository."), + "KCHREPOS0013E": _("Specify repository base URL or mirror list in order to create or update a YUM repository."), "KCHREPOS0014E": _("Repository management tool was not recognized for your system."), "KCHREPOS0015E": _("Repository %(repo_id)s is already enabled."), "KCHREPOS0016E": _("Repository %(repo_id)s is already disabled."), diff --git a/src/kimchi/repositories.py b/src/kimchi/repositories.py index f826ac9..f1e1eb3 100644 --- a/src/kimchi/repositories.py +++ b/src/kimchi/repositories.py @@ -1,7 +1,7 @@ # # Project Kimchi # -# Copyright IBM, Corp. 2014 +# Copyright IBM, Corp. 2014-2015 # # This library is free software; you can redistribute it and/or # modify it under the terms of the GNU Lesser General Public @@ -269,19 +269,25 @@ class YumRepo(object): if repo_id not in repos.repos.keys(): raise NotFoundError("KCHREPOS0012E", {'repo_id': repo_id}) - config = params.get('config', {}) entry = repos.getRepo(repo_id) baseurl = params.get('baseurl', None) + config = params.get('config', {}) mirrorlist = config.get('mirrorlist', None) + if len(baseurl.strip()) == 0: + baseurl = None + + if len(mirrorlist.strip()) == 0: + mirrorlist = None + + if baseurl is None and mirrorlist is None: + raise MissingParameter("KCHREPOS0013E") + if baseurl is not None: validate_repo_url(baseurl) entry.baseurl = baseurl - if mirrorlist == '': - mirrorlist = None - if mirrorlist is not None: validate_repo_url(mirrorlist) entry.mirrorlist = mirrorlist -- 2.1.0

3 3

[PATCH] Fix issue #621: Catch libvirt exception when network does not have bridge name
by Aline Manera 20 Mar '15

20 Mar '15

The bridged network does not have a bridge name (<bridge> on XML) <network> <name>bridged</name> <uuid>70441e61-2f87-44c4-836b-5cc817fcf025</uuid> <forward dev='eth0' mode='bridge'> <interface dev='eth0'/> </forward> </network> and then libvirt raises an exception when we try to get the bridge name. File "/usr/lib64/python2.7/site-packages/libvirt.py", line 2780, in bridgeName if ret is None: raise libvirtError ('virNetworkGetBridgeName() failed', net=self) libvirtError: internal error: network 'bridged-eth0' does not have a bridge name. Signed-off-by: Aline Manera <alinefm(a)linux.vnet.ibm.com> --- src/kimchi/model/networks.py | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/src/kimchi/model/networks.py b/src/kimchi/model/networks.py index 7265c64..92aed92 100644 --- a/src/kimchi/model/networks.py +++ b/src/kimchi/model/networks.py @@ -210,7 +210,17 @@ class NetworksModel(object): br_xml = create_vlan_tagged_bridge_xml(br_name, interface, vlan_id) conn = self.conn.get() - if br_name in [net.bridgeName() for net in conn.listAllNetworks()]: + bridges = [] + for net in conn.listAllNetworks(): + # Bridged networks do not have a bridge name + # So in those cases, libvirt raises an error when trying to get + # the bridge name + try: + bridges.append(net.bridgeName()) + except libvirt.libvirtError, e: + kimchi_log.error(e.message) + + if br_name in bridges: raise InvalidOperation("KCHNET0010E", {'iface': br_name}) with RollbackContext() as rollback: -- 2.1.0

3 3

[PATCH] Fix issue #620: Allow use iSCSI/SCSI volume on Template
by Aline Manera 20 Mar '15

20 Mar '15

When selecting an iSCSI/SCSI volume to create a Template, the disk size is calculated according to the selected volume. In that case only the volume information should be used: { ... "disks":[ { "volume":"unit:0:0:2", "index":0 } ] } Also to get and display the disk size for user we should query the selected volume. This patch also created function to call GET /storagepools/<pool> API. That way we can easily get the pool type without the need to query the information on all pools returned by GET /storagepools API Signed-off-by: Aline Manera <alinefm(a)linux.vnet.ibm.com> --- ui/js/src/kimchi.api.js | 15 +++++- ui/js/src/kimchi.template_edit_main.js | 86 ++++++++++++++++++---------------- 2 files changed, 59 insertions(+), 42 deletions(-) diff --git a/ui/js/src/kimchi.api.js b/ui/js/src/kimchi.api.js index 2abe8f5..838b8d3 100644 --- a/ui/js/src/kimchi.api.js +++ b/ui/js/src/kimchi.api.js @@ -1,7 +1,7 @@ /* * Project Kimchi * - * Copyright IBM, Corp. 2013-2014 + * Copyright IBM, Corp. 2013-2015 * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. @@ -806,6 +806,19 @@ var kimchi = { }); }, + getStoragePool: function(poolName, suc, err) { + var url = kimchi.url + 'storagepools/' + encodeURIComponent(poolName); + kimchi.requestJSON({ + url : url, + type : 'GET', + contentType : 'application/json', + timeout: 2000, + dataType : 'json', + success : suc, + error : err + }); + }, + getStoragePoolVolume: function(poolName, volumeName, suc, err) { var url = kimchi.url + 'storagepools/' + encodeURIComponent(poolName) + '/storagevolumes/' + encodeURIComponent(volumeName); kimchi.requestJSON({ diff --git a/ui/js/src/kimchi.template_edit_main.js b/ui/js/src/kimchi.template_edit_main.js index a557ca5..85f76cd 100644 --- a/ui/js/src/kimchi.template_edit_main.js +++ b/ui/js/src/kimchi.template_edit_main.js @@ -89,51 +89,54 @@ kimchi.template_edit_main = function() { $('select', '#form-template-storage').change(function() { var selectedItem = $(this).parent().parent(); var tempStorageNameFull = $(this).val(); - var tempType; - var tempStorageName =tempStorageNameFull.split('/')[0]; - var scsiCap; - $.each(result, function(index, storageEntities) { - if (tempStorageName === storageEntities.name) { - selectedItem.find('.template-storage-type').val(storageEntities.type); - scsiCap = storageEntities.capacity / Math.pow(1024, 3); - tempType = storageEntities.type; + var tempName = tempStorageNameFull.split('/'); + var tempStorageName = tempName[0]; + $('.template-storage-name').val(tempStorageNameFull); + kimchi.getStoragePool(tempStorageName, function(info) { + tempType = info.type; + selectedItem.find('.template-storage-type').val(tempType); + if (tempType === 'iscsi' || tempType === 'scsi') { + kimchi.getStoragePoolVolume(tempStorageName, tempName[tempName.length-1], function(info) { + volSize = info.capacity / Math.pow(1024, 3); + $('.template-storage-disk', selectedItem).attr('readonly', true).val(volSize); + }); + } else { + $('.template-storage-disk', selectedItem).attr('readonly', false); } }); - if (tempType === 'iscsi' || tempType === 'scsi') { - $('.template-storage-disk', selectedItem).attr('readonly', true).val(scsiCap); - } else { - $('.template-storage-disk', selectedItem).attr('readonly', false); - } - $('.template-storage-name').val(tempStorageNameFull); }); }; if ((origDisks && origDisks.length) && (origPool && origPool.length)) { splitPool = origPool.split('/'); - var defaultPool; + var defaultPool = splitPool[splitPool.length-1]; var defaultType; - $.each(result, function(index, poolEntities) { - if (poolEntities.name === splitPool[splitPool.length-1]) { - defaultType = poolEntities.type; - defaultPool = splitPool[splitPool.length-1] - } - }); - if (origDisks[0]['volume']) { - defaultPool = defaultPool + '/' + origDisks[0]['volume']; - } - $.each(origDisks, function(index, diskEntities) { - var storageNodeData = { - viewMode : '', - editMode : 'hide', - storageName : defaultPool, - storageType : defaultType, - storageDisk : diskEntities.size - } - addStorageItem(storageNodeData); + + kimchi.getStoragePool(defaultPool, function(info) { + defaultType = info.type; + $.each(origDisks, function(index, diskEntities) { + var storageNodeData = { + viewMode : '', + editMode : 'hide', + storageName : defaultPool, + storageType : defaultType, + storageDisk : diskEntities.size + } + + if (diskEntities.volume) { + kimchi.getStoragePoolVolume(defaultPool, diskEntities.volume, function(info) { + var volSize = info.capacity / Math.pow(1024, 3); + var nodeData = storageNodeData + nodeData.storageName = defaultPool + '/' + diskEntities.volume; + nodeData.storageDisk = volSize; + addStorageItem(nodeData); + $('.template-storage-disk').attr('readonly', true); + }); + } else { + addStorageItem(storageNodeData); + } + }); }); - if(defaultType === 'iscsi' || defaultType === 'scsi') { - $('.template-storage-disk').attr('readonly', true); - } } $('#template-edit-storage-add-button').button({ @@ -249,10 +252,11 @@ kimchi.template_edit_main = function() { //Fix me: Only support one storage pool now var storages = $('.template-tab-body .item', '#form-template-storage'); var tempName = $('.template-storage-name', storages).val(); + var tmpItem = $('#form-template-storage .item'); tempName = tempName.split('/'); - var tempNameHead =tempName[0]; + var tempNameHead = tempName[0]; var tempNameTail = tempNameHead; - if(tempNameHead === 'iscsi' || tempNameHead =='scsi') { + if($('.template-storage-type', tmpItem).val() === 'iscsi' || $('.template-storage-type', tmpItem).val() == 'scsi') { tempNameTail = tempName[tempName.length-1]; } tempName = '/storagepools/' + tempNameHead; @@ -260,12 +264,12 @@ kimchi.template_edit_main = function() { $.each(editableFields, function(i, field) { /* Support only 1 disk at this moment */ if (field == 'disks') { - var tmpItem = $('#form-template-storage .item'); - origDisks[0].size = Number($('.template-storage-disk', tmpItem).val()); - if($('.template-storage-type', tmpItem).val() === 'iscsi' || $('.template-storage-type', tmpItem).val() =='scsi') { + if($('.template-storage-type', tmpItem).val() === 'iscsi' || $('.template-storage-type', tmpItem).val() == 'scsi') { + origDisks[0]['size'] && delete origDisks[0]['size']; origDisks[0]['volume'] = tempNameTail; } else { origDisks[0]['volume'] && delete origDisks[0]['volume']; + origDisks[0].size = Number($('.template-storage-disk', tmpItem).val()); } data[field] = origDisks; } -- 2.1.0

2 2

[PATCH] Add support to recognize RHEL-LE distro and version
by Jose Ricardo Ziviani 20 Mar '15

20 Mar '15

- When creating a new template, RHEL little-endian was not being reconized. This commit fixes the regex to support such distro. --- src/kimchi/isoinfo.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/kimchi/isoinfo.py b/src/kimchi/isoinfo.py index b5a1769..818ffe3 100644 --- a/src/kimchi/isoinfo.py +++ b/src/kimchi/isoinfo.py @@ -110,7 +110,7 @@ iso_dir = [ ('opensuse', '12.2', 'openSUSE-DVD-i586-Build0167|openSUSE-DVD-x86_640167'), ('rhel', '4.8', 'RHEL/4-U8'), - ('rhel', lambda m: m.group(1), 'RHEL[_/-](\d+\.\d+)'), + ('rhel', lambda m: m.group(2), 'RHEL(-LE)?[_/-](\d+\.\d+)'), ('debian', lambda m: m.group(1), 'Debian (\d+\.\d+)'), ('ubuntu', lambda m: m.group(2), '[Uu]buntu(-Server)? (\d+\.\d+)'), ('fedora', lambda m: m.group(1), 'Fedora[ -](\d+)'), -- 1.9.1

2 2

[kimchi-devel][PATCH 0/2] Peer discovery fixes
by lvroyce＠linux.vnet.ibm.com 20 Mar '15

20 Mar '15

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> Royce Lv (2): Update doc for federation Deregister peer information when kimchid exit docs/README-federation.md | 4 +++- src/kimchi/model/peers.py | 12 ++++++++++++ 2 files changed, 15 insertions(+), 1 deletion(-) -- 1.9.1

2 4

[PATCH] Assign HTML labels to their corresponding elements
by Crístian Viana 20 Mar '15

20 Mar '15

Somea UI form fields have labels which aren't assigned to those fields, which makes it more difficult to click and interact with them. Assign the labels to their corresponding elements, so when the user clicks on a label, the element also receives the click event. Also, some indentation levels have been changed. The elements which have their labels assigned to are: - The text box where the user types the file path in "New Template" > "Local image file"; - All form elements in "Templates" > "Actions" > "Edit", tab "Processor"; - The checkbox used to select a device in "Storage pools" > "Actions" > "Extend"; Signed-off-by: Crístian Viana <vianac(a)linux.vnet.ibm.com> --- ui/pages/tabs/storage.html.tmpl | 2 +- ui/pages/template-add.html.tmpl | 4 +++- ui/pages/template-edit.html.tmpl | 20 ++++++++++++++++---- 3 files changed, 20 insertions(+), 6 deletions(-) diff --git a/ui/pages/tabs/storage.html.tmpl b/ui/pages/tabs/storage.html.tmpl index dbbaef9..6af7327 100644 --- a/ui/pages/tabs/storage.html.tmpl +++ b/ui/pages/tabs/storage.html.tmpl @@ -128,7 +128,7 @@ </script> <script id="logicalPoolExtendTmpl" type="html/text"> <div> - <input type="checkbox" value="{path}" name="devices"> + <input type="checkbox" value="{path}" name="devices" id="{name}"> <label for="{name}">{path}</label> </div> </script> diff --git a/ui/pages/template-add.html.tmpl b/ui/pages/template-add.html.tmpl index eaafaa7..3b82c6c 100644 --- a/ui/pages/template-add.html.tmpl +++ b/ui/pages/template-add.html.tmpl @@ -130,7 +130,9 @@ <h2 class="step-title">$_("Local Image File")</h2> </header> <div class="body"> - <span>File Path:</span><input type="text"><button class="ui-button-primary">$_("Create")</button> + <label for="vm-image-local-text">File Path:</label> + <input type="text" id="vm-image-local-text" /> + <button class="ui-button-primary">$_("Create")</button> </div> </div> diff --git a/ui/pages/template-edit.html.tmpl b/ui/pages/template-edit.html.tmpl index 7177bf0..3a84870 100644 --- a/ui/pages/template-edit.html.tmpl +++ b/ui/pages/template-edit.html.tmpl @@ -118,11 +118,23 @@ <div class="template-tab-body"></div> </form> <form id="form-template-processor"> - <div>$_("CPU Number"):<input type="text" value="1" id="cpus"></div> - <div class="manual"><input type="checkbox">$_("Manually set CPU topology")</div> + <div> + <label for="cpus">$_("CPU Number"):</label> + <input type="text" value="1" id="cpus" /> + </div> + <div class="manual"> + <input type="checkbox" id="cpus-check" /> + <label for="cpus-check">$_("Manually set CPU topology")</label> + </div> <div class="topology hide"> - <div>$_("Cores"):<input type="text" value="1" id="cores"></div> - <div>$_("Threads"):<select id="threads"></select></div> + <div> + <label for="cores">$_("Cores"):</label> + <input type="text" value="1" id="cores" /> + </div> + <div> + <label for="threads">$_("Threads"):</label> + <select id="threads"></select> + </div> </div> </form> </div> -- 2.1.0

2 2

[PATCH 0/3] unit tests - Power arch fixes
by Daniel Henrique Barboza 20 Mar '15

20 Mar '15

This patch set contains fixes to several unit tests that had somewhat the same root cause - the differences between x86 and power/ppc64le in osinfo.py template configurations. Daniel Henrique Barboza (3): tests/test_osinfo.py: fixes for Power architecture Kimchi tests: Power system fixes - removing hardcoded values test/test_model: Power architecture fixes src/kimchi/osinfo.py | 9 +++++++++ tests/iso_gen.py | 3 ++- tests/test_mockmodel.py | 4 +++- tests/test_model.py | 10 ++++++---- tests/test_osinfo.py | 36 ++++++++++++++++++++++++++---------- tests/test_rest.py | 6 ++++-- tests/test_vmtemplate.py | 8 ++++++-- 7 files changed, 56 insertions(+), 20 deletions(-) -- 1.9.3

2 8

[PATCH 0/3] Fix storage volume issues
by Crístian Viana 20 Mar '15

20 Mar '15

Crístian Viana (3): Remove ISO permission check when creating a template issue #565: Allow a template's ISO to be a block device issue #564: Parse logical volumes to find out their actual formats src/kimchi/i18n.py | 5 --- src/kimchi/isoinfo.py | 7 +++- src/kimchi/kvmusertests.py | 75 -------------------------------------- src/kimchi/model/storagevolumes.py | 19 +++++++++- src/kimchi/model/templates.py | 12 ------ src/kimchi/vmtemplate.py | 10 ++++- 6 files changed, 31 insertions(+), 97 deletions(-) delete mode 100644 src/kimchi/kvmusertests.py -- 2.1.0

2 6

[PATCH] tests/test_osinfo.py: fixes for Power architecture
by Daniel Henrique Barboza 20 Mar '15

20 Mar '15

- osinfo tests weren't considering config differences between x86 and Power archs, resulting in errors. - a new test was added to explictly test the case where an unknown distro/version should return the configuration of an old distro/version after the lookup. Signed-off-by: Daniel Henrique Barboza <dhbarboza82(a)gmail.com> --- tests/test_osinfo.py | 29 ++++++++++++++++++++++++----- 1 file changed, 24 insertions(+), 5 deletions(-) diff --git a/tests/test_osinfo.py b/tests/test_osinfo.py index d5e90b4..f2837f0 100644 --- a/tests/test_osinfo.py +++ b/tests/test_osinfo.py @@ -30,13 +30,23 @@ class OSInfoTests(unittest.TestCase): self.assertEquals('unknown', entry['os_version']) self.assertEquals(['default'], entry['networks']) + def _check_old_distro_diskbus_nic_multi_arch(self, disk_bus, nic_model): + if _get_arch() == 'x86': + self.assertEquals(disk_bus, 'ide') + self.assertEquals(nic_model, 'e1000') + else: + self.assertEquals(disk_bus, 'scsi') + self.assertEquals(nic_model, 'spapr-vlan') + def test_old_distros(self): old_versions = {'debian': '5.0', 'ubuntu': '7.04', 'opensuse': '10.1', 'centos': '5.1', 'rhel': '5.1', 'fedora': '15'} for distro, version in old_versions.iteritems(): entry = lookup(distro, version) - self.assertEquals(entry['disk_bus'], 'ide') - self.assertEquals(entry['nic_model'], 'e1000') + self._check_old_distro_diskbus_nic_multi_arch( + entry['disk_bus'], + entry['nic_model'] + ) def test_modern_bases(self): for distro, version in modern_version_bases[_get_arch()].iteritems(): @@ -45,10 +55,19 @@ class OSInfoTests(unittest.TestCase): self.assertEquals(entry['nic_model'], 'virtio') def test_modern_distros(self): - modern_versions = {'debian': '7.0', 'ubuntu': '12.04', - 'opensuse': '12.3', 'centos': '6.4', 'rhel': '6.3', - 'fedora': '18', 'gentoo': '12.1'} + # versions based on ppc64 modern distros + modern_versions = {'ubuntu': '14.04', 'opensuse': '13.1', + 'rhel': '6.5', 'fedora': '19', 'sles': '11sp3'} for distro, version in modern_versions.iteritems(): entry = lookup(distro, version) self.assertEquals(entry['disk_bus'], 'virtio') self.assertEquals(entry['nic_model'], 'virtio') + + def test_lookup_unknown_distro_version_returns_old_distro(self): + distro = 'unknown_distro' + version = 'unknwon_version' + entry = lookup(distro, version) + self._check_old_distro_diskbus_nic_multi_arch( + entry['disk_bus'], + entry['nic_model'] + ) -- 1.9.3

2 1

[PATCH] test/vmtemplate.py: fixes for Power architecture
by Daniel Henrique Barboza 16 Mar '15

16 Mar '15

- changing 'fields' variable to a dict instead of an array to manipulate specific fields easier - changing 'memory', 'disk_bus' and 'nic_model' fields when running the tests in a Power host Signed-off-by: Daniel Henrique Barboza <dhbarboza82(a)gmail.com> --- tests/test_vmtemplate.py | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/tests/test_vmtemplate.py b/tests/test_vmtemplate.py index 550bb2a..372f54a 100644 --- a/tests/test_vmtemplate.py +++ b/tests/test_vmtemplate.py @@ -22,6 +22,7 @@ import unittest import uuid +from kimchi.osinfo import _get_arch from kimchi.vmtemplate import VMTemplate from kimchi.xmlutils.utils import xpath_get_text @@ -35,16 +36,21 @@ class VMTemplateTests(unittest.TestCase): os.unlink(self.iso) def test_minimal_construct(self): - fields = (('name', 'test'), ('os_distro', 'unknown'), - ('os_version', 'unknown'), ('cpus', 1), - ('memory', 1024), ('networks', ['default']), - ('disk_bus', 'ide'), ('nic_model', 'e1000'), - ('graphics', {'type': 'vnc', 'listen': '127.0.0.1'}), - ('cdrom', self.iso)) + fields = {'name': 'test', 'os_distro': 'unknown', + 'os_version': 'unknown', 'cpus': 1, + 'memory': 1024, 'networks': ['default'], + 'disk_bus': 'ide', 'nic_model': 'e1000', + 'graphics': {'type': 'vnc', 'listen': '127.0.0.1'}, + 'cdrom': self.iso} + + if _get_arch() in ('power', 'ppc64le'): + fields['memory'] = 1280 + fields['disk_bus'] = 'scsi' + fields['nic_model'] = 'spapr-vlan' args = {'name': 'test', 'cdrom': self.iso} t = VMTemplate(args) - for name, val in fields: + for name, val in fields.iteritems(): self.assertEquals(val, t.info.get(name)) def test_construct_overrides(self): -- 1.9.3

1 1

[PATCH] tests/test_mockmodel.py: fixes for Power architecture
by Daniel Henrique Barboza 16 Mar '15

16 Mar '15

Testing for the right amount of default memory when running the test in a Power system. Signed-off-by: Daniel Henrique Barboza <dhbarboza82(a)gmail.com> --- tests/test_mockmodel.py | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tests/test_mockmodel.py b/tests/test_mockmodel.py index 542f845..ed3a0f9 100644 --- a/tests/test_mockmodel.py +++ b/tests/test_mockmodel.py @@ -28,6 +28,7 @@ import kimchi.mockmodel from utils import get_free_port, patch_auth, request, run_server from utils import wait_task from kimchi.control.base import Collection, Resource +from kimchi.osinfo import _get_arch test_server = None @@ -155,10 +156,13 @@ class MockModelTests(unittest.TestCase): 'io_throughput', 'io_throughput_peak')) info = model.vm_lookup(u'test-vm') + mem_expected = 1024 + if _get_arch() in ('power', 'ppc64le'): + mem_expected = 1280 self.assertEquals(keys, set(info.keys())) self.assertEquals('shutoff', info['state']) self.assertEquals('test-vm', info['name']) - self.assertEquals(1024, info['memory']) + self.assertEquals(mem_expected, info['memory']) self.assertEquals(1, info['cpus']) self.assertEquals('images/icon-vm.png', info['icon']) self.assertEquals(stats_keys, set(info['stats'].keys())) -- 1.9.3

1 1

[PATCH v3 0/5] Create VMs Asynchronously
by Christy Perez 13 Mar '15

13 Mar '15

v3 Changes: Instead of adding a description key to the task, just append the action desciptor to the target_uri. If a guest has a large disk, and uses a filesystem that requires preallocation, it can take several minutes to create a VM. During that time, kimchi is tied up by the VM creation. This patch changes the VMs Collection to be an AsyncCollection. Another change required for this was to create a more granular way to query tasks. Currently it is not possible (using the API) to query tasks for the same collection or resource type that may have different operations. For example, VM cloning is also an asynchronous operation. For the guests tab, the UI was querying all running tasks and displaying them with the Cloning label. This picked up VMs that were being created as well. For more information about how the tasks can be queried, see the updated API doc and the UI change. Christy Perez (5): Granular Task Queries: Backend Granular task query test updates Granular Task Queries: UI Create guests asynchronously: Backend Async vm creation test updates docs/API.md | 16 ++++++++ src/kimchi/control/vms.py | 4 +- src/kimchi/mockmodel.py | 7 ++-- src/kimchi/model/debugreports.py | 4 +- src/kimchi/model/host.py | 4 +- src/kimchi/model/storagepools.py | 2 +- src/kimchi/model/storagevolumes.py | 12 +++--- src/kimchi/model/vms.py | 33 +++++++++++++--- src/kimchi/model/vmsnapshots.py | 5 ++- tests/test_authorization.py | 23 +++++++----- tests/test_mockmodel.py | 12 ++++-- tests/test_model.py | 60 +++++++++++++++++++---------- tests/test_model_storagevolume.py | 2 +- tests/test_rest.py | 77 ++++++++++++++++++++++++++++---------- ui/js/src/kimchi.guest_main.js | 4 +- 15 files changed, 188 insertions(+), 77 deletions(-) -- 2.1.0

1 6

[kimchi-devel][PATCH 0/2] Fix error usage of feature tests
by lvroyce＠linux.vnet.ibm.com 13 Mar '15

13 Mar '15

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> Feature tests depend on running kimchi server should not be used in model tests, while coincidently, command "qemu-io -r http://localhost:8080/images/icon-fedora.png -c 'read -v 0 512'" return code is 0 so we cannot count on the return code. This patch forces module usage without running kimchi server to return false, and fixes testcase accordingly. REF: https://github.com/kimchi-project/kimchi/issues/579 https://github.com/kimchi-project/kimchi/issues/580 Royce Lv (2): Avoid using server dependent feature tests when server not running Fix wrong usage of feature tests src/kimchi/model/featuretests.py | 5 ++++- src/kimchi/utils.py | 8 ++++++++ tests/test_model.py | 16 +++++++--------- 3 files changed, 19 insertions(+), 10 deletions(-) -- 1.9.3

3 6

[PATCH] Remove slash "/" filter in template name when create VM
by Rodrigo Trujillo 13 Mar '15

13 Mar '15

Users are allowed to create or update template names with "slash", this generates an error when he tries to create a guest with that template because the guest paramters validation prohibits slashes in templates names. This patch fixes this problem. Signed-off-by: Rodrigo Trujillo <rodrigo.trujillo(a)linux.vnet.ibm.com> --- src/kimchi/API.json | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/kimchi/API.json b/src/kimchi/API.json index 0cfa20c..f507251 100644 --- a/src/kimchi/API.json +++ b/src/kimchi/API.json @@ -233,7 +233,7 @@ "template": { "description": "The URI of a template to use when building a VM", "type": "string", - "pattern": "^/templates/[^/]+/?$", + "pattern": "^/templates/(.*?)/?$", "required": true, "error": "KCHVM0012E" }, -- 2.1.0

3 2

[PATCH v3] Issue #587: Man page submission for kimchid
by Frédéric Bonnard 13 Mar '15

13 Mar '15

From: Frederic Bonnard <frediz(a)linux.vnet.ibm.com> Hi, I added the federation information into a section within the manpage with a bit of cosmetic. F. Frederic Bonnard (1): Issue #587: Man page submission for kimchid COPYING | 5 +- Makefile.am | 2 + contrib/kimchi.spec.fedora.in | 4 ++ contrib/kimchi.spec.suse.in | 4 ++ docs/Makefile.am | 1 + docs/kimchid.8 | 156 ++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 170 insertions(+), 2 deletions(-) create mode 100644 docs/kimchid.8 -- 1.9.1

2 3

[PATCH V2] Remove slash "/" filter in template name when create VM
by Rodrigo Trujillo 13 Mar '15

13 Mar '15

Users are allowed to create or update template names with "slash", this generates an error when he tries to create a guest with that template because the guest paramters validation prohibits slashes in templates names. This patch fixes this problem. Signed-off-by: Rodrigo Trujillo <rodrigo.trujillo(a)linux.vnet.ibm.com> --- src/kimchi/API.json | 2 +- src/kimchi/utils.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/kimchi/API.json b/src/kimchi/API.json index 0cfa20c..f507251 100644 --- a/src/kimchi/API.json +++ b/src/kimchi/API.json @@ -233,7 +233,7 @@ "template": { "description": "The URI of a template to use when building a VM", "type": "string", - "pattern": "^/templates/[^/]+/?$", + "pattern": "^/templates/(.*?)/?$", "required": true, "error": "KCHVM0012E" }, diff --git a/src/kimchi/utils.py b/src/kimchi/utils.py index f1ef12c..0d5f988 100644 --- a/src/kimchi/utils.py +++ b/src/kimchi/utils.py @@ -45,7 +45,7 @@ task_id = 0 def _uri_to_name(collection, uri): - expr = '/%s/(.*?)/?$' % collection + expr = '/%s/(.*?)$' % collection m = re.match(expr, uri) if not m: raise InvalidParameter("KCHUTILS0001E", {'uri': uri}) -- 2.1.0

3 2