February 2013 - vdsm-devel - Fedora Mailing-Lists

by Douglas Schilling Landgraf

Hello, On F17, there is a bug [1] affecting the installation of VDSM, we cannot modify the selinux booleans. The new policycoreutils package under testing repository resolves the problem. To promote it to a updates repo, we need 3 positive karmas [2] of tests, if you can help testing, please follow: https://bugzilla.redhat.com/show_bug.cgi?id=908773#c3 [1] https://bugzilla.redhat.com/show_bug.cgi?id=908773 [2] http://fedoraproject.org/wiki/Bodhi#Karma Thanks! -- Cheers Douglas

11 years, 2 months

1
0
0 / 0

ovirt-guest-agent: Package for Fedora has been accepted

by Vinzenz Feenstra

Hi, I am glad that I can announce that the ovirt guest agent has been also accepted as a package for Fedora. I have submitted the first packages for the updates-testing repository and if nothing goes wrong the packages will be available within a few days :-) https://admin.fedoraproject.org/updates/ovirt-guest-agent-1.0.6-4.fc17 https://admin.fedoraproject.org/updates/ovirt-guest-agent-1.0.6-4.fc18 <https://admin.fedoraproject.org/updates/ovirt-guest-agent-1.0.6-4.fc17> I am the maintainer of the fedora package, if you run into any issues please let me know so that we can take care of handling them. -- Regards, Vinzenz Feenstra | Senior Software Engineer RedHat Engineering Virtualization R & D Phone: +420 532 294 625 IRC: vfeenstr or evilissimo Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com

11 years, 2 months

1
0
0 / 0

Review Request: add JsonPpcClientAPI class

by shaohef＠linux.vnet.ibm.com

hi all, I have post a patch about Json Ppc Client API. http://gerrit.ovirt.org/11846 After this patch, we can call json API more easily. for example, if we want to call VM.shutdown. 1. Firstly, we can get the description from the vdsm/vdsm_api/vdsm-api.html, as follow. * VM.shutdown:* Request a graceful shutdown of a VM. Arguments vmID UUID <cid:part2.04060403.02020709@linux.vnet.ibm.com> The UUID of the VM delay int <cid:part3.02040007.06060401@linux.vnet.ibm.com> #optional Wait a number of seconds before shutting down message str <cid:part4.00050703.02040502@linux.vnet.ibm.com> #optional Message to display to VM users before shutting down Since: /4.10.0/ 2. Then we can call this API easily in our python code: insAPI = ApiJsonPpcClientAPI(reactorClient) insAPI.VM.shutdown(UUID, delay=10, message="shutdown in 10 seconds") -- Sheldon Feng(???)<shaohef(a)linux.vnet.ibm.com> IBM Linux Technology Center

11 years, 2 months

1
0
0 / 0

how to cause VNIC hot unplug to fail?

by Martin Pavlik

Hi All, I need to create a test for new VDSM NIC hotplug hooks. In order to do this I need to test a scenario where VNIC hotunplug fails. I do not know how to achieve such state. Could you point me in the right direction please? So far I tried to disable acpid in guest OS and acpi = off kernel parameter. It seems that hotunplug does not fail on running guest VM with no OS inside. TIA Regards Martin Pavlik RHEV Quality Engineer RHCSA Red Hat Czech Republic, s.r.o. Phone: +420 532 294 478 Direct call: 82 62478 Cell phone: +420 723 782 961 E-mail: mpavlik(a)redhat.com irc nick: mpavlik irc channels: #tlv #brno #rhev-dev #rhev-qe

11 years, 2 months

1
0
0 / 0

Re: [vdsm] May I apply for a user account on jenkins.ovirt.org to run VDSM functional tests?

by Ewoud Kohl van Wijngaarden

On Tue, Jan 29, 2013 at 06:15:08AM -0500, Eyal Edri wrote: > ----- Original Message ----- > > From: "Zhou Zheng Sheng" <zhshzhou(a)linux.vnet.ibm.com> > > To: infra(a)ovirt.org > > Cc: "ShaoHe Feng" <shaohef(a)linux.vnet.ibm.com> > > Sent: Tuesday, January 29, 2013 12:24:27 PM > > Subject: May I apply for a user account on jenkins.ovirt.org to run VDSM functional tests? > > > > Hi all, > > > > I notice there is no VDSM functional tests running in oVirt Jenkins. > > Currently in VDSM we have some XML-RPC functional test cases for > > iSCSI, > > localfs and glusterfs storage as well as creating and destroying VMs > > on > > those storage. Functional tests through JSON-RPC are under review. I > > also submit a patch to Gerrit for running the tests easily > > (http://gerrit.ovirt.org/#/c/11238/). More test cases will be added > > to > > improve test coverage and reduce the chance of regression. > > > > Some bugs that can not be covered by unit test can be caught by > > functional tests. I think it would be helpful to run these functional > > tests continuously. We can also configure the Gerrit trigger in > > Jenkins > > to run functional tests when someone verifies the patch or when it > > gets > > approved but not merged. This may be helpful to the maintainer. > > > > I've setup a Jenkins job for VDSM functional tests in my lab server. > > You > > can refer to the job configuration of my current setup > > (https://github.com/edwardbadboy/vdsm-jenkins/blob/master/config.xml). > > After my patch in Gerrit is accepted, the job configuration will be > > simpler and the hacks can be removed. May I apply a user account for > > creating job in the oVirt Jenkins? > > > > Hi Zhou, > Basically there shouldn't be any problem with that. > we have an option for giving a 'power-user' permissions for certain > users on oVirt misc projects to add and and configure jobs for thier > project. > > it requires knowledge in jenkins, which it seems that you have and > recognition from the team/other developers from the relevant project > (in this case, VDSM) that you are an active member of the project. > (just a formality essentially) > > I've added engine-devel list to this thread so anyone from vdsm team > can vote +1 for adding you as a power user for jenkins. > > once we'll receive a few +1 and not objections i'll create a user for > you and send you the details. > I think vdsm-devel is more relevant here.

11 years, 2 months

5
5
0 / 0

Meeting minutes, Jan 28

by Dan Kenigsberg

Following are the notes that I've taking during the call. I'm pretty sure that I'm missing something important, such as another blocker BZ, so please reply with corrections/addtions. - Adam: going to introduce Python binding for the new Vdsm API - Sharad has joined our call for the first time. Welcome! Interested in backup/restore functionality in oVirt, to facilitate integration with TSM, presumably http://www-142.ibm.com/software/products/us/en/tivostormana/ - Federico warns that live snapshot is not enough, since we lack live merge. After a year of daily backups, each qcow chain would be 365 hops deep. ovirt-3.2: - We should probably revert http://gerrit.ovirt.org/9315 (integrate zombie reaper in supervdsmServer) as it tickles a bug in multiprocessing. See http://lists.ovirt.org/pipermail/users/2013-January/011857.html for more details in the thread starting in http://lists.ovirt.org/pipermail/users/2013-January/011747.html (thanks for raising the issue and finding the culprit, Dead Horse) - Danken believes that all important el6-specific issues reported by dreyou are now handled in the master and ovirt-3.2 branches. Please report this list if he's wrong. - Federico reports that udev is going to revert to its Fedora 18alpha behavior so that we can keep our integration with it. I'd love to see the udev BZ listed in our tracker https://bugzilla.redhat.com/showdependencytree.cgi?id=881006&hide_resolved=1 as I do not recall it. Federico, could you add it? We should probably require a newer udev release. - Federico reports that we must take a short-but-important patch from Lee Yarwood: http://gerrit.ovirt.org/#/c/11281/ Cross-distro: - Adam: a guy from IBM is now listing issues that block running Vdsm on Ubuntu. Yay! - Saggi suggests that IBM contribute an Ubuntu slave for Jenkins, so that each and every patch is tested not to introduce Ubuntu regressions - just as Red Hat has recently done with EL6. - Danken: yes, we have got el6-based unit-testing running. If you are still getting an unjustified X, you may need to rebase on current master. - Federico: NetworkManager 9 has bridge support, which leads the way to NM-based implementation of Vdsm's configNetwork. - Danken: Toni is working on a suggestion for refactoring configNetwork in such a way that multiple implementations (e.g. ifcfg-based, NM-based) can coexist. Stay stuned for his Wiki page on the subject. Other refactoring going on: - Vinzenz has begun breaking the horrible libvirtvm+vm to edible-size pieces. Reviews are most welcome, particularly from Mark Wu, whose http://gerrit.ovirt.org/10054/ only awaits verification tick. Functional Testing: - Adam: a guy from IBM (Zhou, I presume) is running the functional tests via Jenkins on his laptop. Yay! - Adam: plans to add a functional test for the getVdsCaps verb, as he found out that an evil Vdsm developer has added values to the caps without updating the schema. - P.S. FlowID is now in for storage verbs as well (thanks, Douglas). I hope Haim is happier now. Happy coding!

11 years, 2 months

3
3
0 / 0

A question about the SPM operation permission in VDSM

by shuming＠linux.vnet.ibm.com

Hi, In the VDSM code about some SPM operations like HSM.deleteImage(), It is found that VDSM doesn't check if the operation will be launched on a SPM host or not. It only checks if the storage pool is already acquired by one SPM host, but not necessary the same host as the SPM operation is delivered to. The code is like this: HSM.deleteImage() { ... HSM._spmSchedule() { self.validateSPM(spUUID) <--- Only check if the storage pool was acquired by one host, but not necessary this host } ... } So it really depends on the node management application AKA ovirt-engine to dispatch the SPM operations to the right VDSM host. And the VDSM host itself doesn't check if it is the SPM host which can execute the operation. To me, it is a bit broken. When the engine query the VDSM host who is the SPM host, it can get the right one. However, the host may be broken for some reason after the engine believes it is the SPM host and the host loses the SPM privilege, another host will take the SPM role. Then the engine continue to send the SPM operations to the broken host. As a result, the SPM operation will be launched on a non-SPM host. So I think there is a small window of racing to corrupt the VDSM hosts meta data. I think VDSM host should check if it is SPM before the SPM job is scheduled. If the host lost the SPM role already, It should fail the RPC call from the engine to let the engine to retry the operation after engine knows the failure of the former call. -- --- 舒明 Shu Ming Open Virtualization Engineerning; CSTL, IBM Corp. Tel: 86-10-82451626 Tieline: 9051626 E-mail: shuming(a)cn.ibm.com or shuming(a)linux.vnet.ibm.com Address: 3/F Ring Building, ZhongGuanCun Software Park, Haidian District, Beijing 100193, PRC

11 years, 2 months

3
2
0 / 0

sphinx doc generation

by Giuseppe Vallarelli

Dear vdsm developers I'm Giuseppe I've recently pushed a little work I've done to improve vdsm's sphinx documentation. I've moved doc structure to the project root folder and updated references in the different *.rst files used. The way the import statements are used in vdsm doesn't allow sphinx to generate documentation especially when auto directives are used, so to try out my change set, is important to update the python path env variable by making it pointing also to the vdsm python package. Unluckily the modules clientIf and vm still cause troubles when sphinx wants to extract documentation (see import error: from vdsm import vdscli). More work needs to be done to clean up documentation, this is just a starting point. In the commit messages I've linked this bug (https://bugzilla.redhat.com/show_bug.cgi?id=491467) cause I thought it was the most appropriate. I do apologize for creating more commits than necessary. I usually split my work in small chunks, but with gerrit it seems that reviews are done on a per commit basis (I'm used to github's pull requests). Cheers Giuseppe -- Giuseppe Vallarelli `All code is guilty until proven innocent.` www.giuseppevallarelli.com (http://www.giuseppevallarelli.com)

11 years, 2 months

1
1
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

vdsm-devel February 2013