summaryrefslogtreecommitdiff
path: root/client
AgeCommit message (Collapse)AuthorFilesLines
2011-05-20KVM test: physical_resources_check: Fixing error messagelmr1-2/+2
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5364 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-17KVM Test: Make import file_transfer works in nic_bonding.pylmr2-1/+1
Add __init__.py in virt/tests/ and correct import link for file_transfer. Signed-off-by: Feng Yang <fyang@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5363 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-06mirror.config-sample: Fixing undefined variable errorlmr1-0/+1
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5359 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-06perf: fix think-o with perf trace outputlmr1-1/+1
The command to generate perf trace output is "perf script", not what I had in the previous patch which was a think-o during cleanup for upstream. Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5355 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-06wb_kupdate: Adding missing __init__.py filelmr1-0/+0
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5354 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-06Mass permission and spacing fixeslmr5-2/+1
All across the autotest tree. Signed-off-by: Lucas Menghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5353 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-05KVM test: Improve get_started.py scriptlmr1-14/+48
* Space steps so it's easier to spot the blocks of info * Use ask to make the questions look better * Give the user the chance to not check SHA1 sum of the isos * Print download URLs if asked to download * Print CD SHA1 sum hashes You can see how a typical execution looks like here: http://www.linux-kvm.org/page/KVM-Autotest/Client_Install#For_the_impatient Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5351 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-05Utils: Moving function ask to common lib utilslmr1-0/+14
This way we can use it in other interactive scripts in autotest. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5350 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-04client job HTML report: Play safe when generating the reportlmr1-1/+4
It is bad to possibly jeopardize a job result for some bug on the HTML report code, so wrap its execution and log errors in case of exceptions. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5349 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-04HTML report: Handle errors when trying to determine test lengthlmr1-4/+7
ome test like operations (like a kernel booting), the status log does not contain test lenght info. So handle accordingly and mark on report as 'Unknown'. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5348 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-04KVM test: installer: Fix KojiInstaller buglmr1-11/+52
That code was not updated during the last version of the refactor patches, so it was referencing the old KojiDownloader class. Let's fix this. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5347 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-04KVM test: Update ctcs run on guestslmr3-22/+21
Just update the config file and the control file that is supposed to run on guests. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5346 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-04client.tests: Move cerberus to ctcs, use new source repolmr7-454/+40
Since CTCS2 was abandoned, I've created a new git repo with it and plan on maintaining CTCS code. Modify the autotest wrapper accordingly, renaming it to 'ctcs'. The prior patches that had to be applied were incorporated to the code base, so some cleanup could be made on the wrapper itself. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5345 592f7852-d20e-0410-864c-8624ca9c26a4
2011-05-02virt.virt_utils: Fix hardcoded test URL in run_tests()lmr1-2/+4
Now that we aim to have different implementations of virtualization tests, the utility function run_tests() can't have the 'kvm' url hardcoded in the function. Rather, it needs to pick up 'vm_type' from the test params. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5344 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29client.virt.kvm_vm: Remove shebang from source filelmr1-1/+0
It is not necessary as this file is not a stand alone program. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5343 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29KVM test: Repair bug of creating kvm guest machinelmr1-0/+1
Signed-off-by: Jiri Zupka <jzupka@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5342 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29KVM test: Fix problem when unexpected strings appear on SSH promptlmr1-2/+5
In some conditions, ssh may respond things like: Warning: Permanently added localhost' (RSA) to the list of known hosts. Take that into account on the remote login function. Signed-off-by: Jason D. Gaston <jason.d.gaston@intel.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5341 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29KVM test: control: Get rid of report generation codelmr1-3/+0
As this is now handled by autotest and control file writers don't have to worry about it anymore. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5340 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29virt.virt_utils: Get rid of create_report functionlmr1-12/+0
As it has been transfered to the html_report module itself. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5339 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29job: Write a job HTML report for every autotest client joblmr1-0/+4
We have a tool that can generate such a file and it makes it easier for people who don't have access to the autotest web interface to analyze job results. With this, all client jobs write such a file, so test writers don't have to worry about it. This change does not regress the job unittests. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5338 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-29tools/html_report: Make html report generation autotest genericlmr1-54/+104
The html report tool was a bit too tied to KVM autotest, and that was not necessary. Turn it into a generic module that can be used by any autotest job, encapsulate the main worker function into autotest API so we don't need to exec it in a subshell, fix naming of the report and CSS details. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5337 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-28client.tools: Add __init__.py file to that dirlmr1-0/+0
So it is importable Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5336 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-28Virt module: Adding missing __init__.py filelmr1-0/+0
While applying the patches for the kvm autotest refactor, patch did not create the 0 sized file __init__.py. Fixing it. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5335 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Moving generic tests to common tests arealmr27-0/+0
So other virt tests can benefit from them. The tests that don't depend on any KVM VM specific features (ie, monitors) were moved. As soon as we abstract some of the monitor functionality, the KVM specific tests can be rewritten on a virt tech agnostic way and then be moved. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5334 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Try to load subtests from a shared tests locationlmr1-5/+13
As we have several subtests that can be shared among different virtualization tests (kvm, xen), manipulate kvm.py to try loading subtests from the kvm area (client/tests/kvm/tests) first, then falling back to the specific test area (planned to be client/virt/tests). Changes from v1: * Martin Jenner has suggested inverting the lookup order. Looking first on the specific kvm subtest area *then* common allows people to override the commom implementation of a test that might be better for that particular virt technology. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> Signed-off-by: Martin Jenner <mjenner@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5333 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Removing the old libraries and programslmr16-13399/+0
As they've been replaced by the new virt_namespace, as well as the new program versions moved to tools. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5332 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Adapt the test code to use the new virt namespacelmr54-271/+269
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5331 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: tests_base.cfg: Introduce parameter 'vm_type'lmr3-0/+4
In order to allow the shared infrastructure to select the correct vm class to instantiate a VM, introduce the parameter vm_type, which for kvm based VMs is, not surprisingly, 'kvm'. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5330 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Move test utilities to client/toolslmr4-0/+1880
The programs cd_hash, html_report, scan_results can be used by other users of autotest, so move them to the tools directory inside the client directory. This patch contains bugfixes suggested by Amos Kong. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> Signed-off-by: Amos Kong <akong@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5329 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-27KVM test: Create autotest_lib.client.virt namespacelmr16-0/+11762
This patch moves all libraries from the KVM test to autotest_lib.client.virt, and makes the needed adaptations and abstracts the KVM implementation from the generic infrastructure. The patch contains bugfixes pointed by Ray Chauvin, and also patches from Cleber Rosa that isolate base vm functionality into a BaseVM class, with a collection of public methods, intended to be used in tests Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> Signed-off-by: Ray Chauvin <rchauvin@redhat.com> Signed-off-by: Cleber Rosa <crosa@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5328 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-26perf profiler: support trace outputlmr1-1/+14
As well as reports, one of the other most useful things you can generate from perf is trace output. Support generating that with trace=True added to the profiler, e.g.: job.profilers.add('perf', events=['fs:*'], trace=True) the output is a raw trace file in the profiling directory. Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5327 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-26perf profiler: include call graph if "parent" report to be generatedlmr1-0/+2
The "parent" sort to perf report is most useful when a call graph is included in the trace, so if we see parent in the sort keys, pass the -g option to perf record to get that output in the trace. Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5326 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-26perf profiler: exec perf record so it is stoppablelmr1-1/+1
perf was being executed via a shell, this means that the pid in self._process.pid will be the pid of the shell and the actual pid of perf will be a child of that process. This would be fine in practice except that the signal being used to stop perf is SIGINT, and this is being sent to the shell. The shell won't necessarily deliver this to its child process (no controlling tty) so what happens is that the shell ignores the signal and we get stuck in self._process.wait() There's no reason to keep the intermediate shell around, so prefix the command with "exec" so that the shell replaces itself with perf and the pid we know is perf, and SIGINT is sent to perf. Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5325 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-25perf profiler: replace strip with replacelmr1-1/+1
strip() strips all characters listed in the given string, replace() simply replaces the string with another (in this case the empty string), the latter is clearly what was intended here. This fixes the sort last option being mangled Change-Id: Ic491f159339637cc2959e19c9c37be1cc31e4833 Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5324 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-19KVM test: Introduce OpenSUSE 11.4 guest definitionslmr1-0/+27
For both 32 and 64 bit. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5322 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-19KVM test: Make OpenSUSE and SLES to use unattended_cdromlmr1-28/+28
We had disabled unattended_cdrom usage with SLES and OpenSUSE due to the fact that autoyast=cdrom wasn't working properly to specify a profile. However, the interaction of the linux floppy driver with qemu-kvm and qemu tends to generate some pretty bad guest kernel panics, reason why we came up with unattended_cdrom for Fedora and RHEL in the first place. Thinkering a bit more with autoyast documentation, I've discovered that specifying autoyast=device://scd0/autoinst.xml works fine and so I've changed the default for both distros to use the CD again. Now, people will have a much smoother experience installing those distros under KVM autotest. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5321 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-19KVM test: unattended_install: Introduce wait_no_ack paramlmr4-30/+23
Some distros such as OpenSUSE won't let the install test to communicate the guest ACK for some weird reason [1], this commit modifies the unattended_install test introducing a wait_no_ack param that will skip the ACK check, and break the unattended install loop when the machine turns off itself (directive halt on OpenSUSE and SLES unattended file). Other distros shouldn't be affected by this change, their install behavior was kept the same. [1] Tried several ways of transmitting the ACK by means of chroot scripts, no way. In the end of the day, I was short on time to spend too much of it figuring out a way to do it, hence the wait_no_ack param was born. I really could use some help from the guys that introduced OpenSUSE and SLES support to kvm autotest. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5320 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-19KVM test: Unattended install Give Linux VMs time to shutdown cleanlylmr11-9/+21
During unattended install, right after we receive the ACK from the guest the test is deemed to be finished, and as shutdown_vm = yes, it'll try to end the vm issuing a shutdown command to it. However, on virtually all Linux guests an SSH server is not available at the end of install, so KVM autotest will end the VM forcefully, which is not really safe, although it has served us well so far. We did not fix this 'problem' so far because on RHEL3, a supported guest, the anaconda syntax does not support the 'poweroff' directive, only 'reboot', so if we don't finish the VM right after the ACK from guest we really can't prevent it from starting the install again, getting an infinite loop. Well then, let's restrict this behavior only to RHEL 3, and fix this properly for all other systems, by introducing a 'shutdown_cleanly' param and setting it to 'yes' to everybody but RHEL 3. Windows doesn't need this at all, as when the ACK is received the system is fully booted up and functional, so the postprocessing will take care of shutting down the VM cleanly. I've fixed all Fedora and RHEL unattended files as part of this change. With this change, we can get rid of another parameter, post_install_delay, that was created with the only purpose of handling the fact that on autoyast install, when the ACK is sent to host, the system is not fully done with install. So we end up fixing yet another problem :) Changes from v1: * Properly convert the post ack timeout to an int * Don't fix SLES unattended files, they'll be fixed in posterior patches Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5319 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-19KVM test: Fix incorrect CdromDisk instantiation when distro is OpenSUSElmr1-1/+1
We need to pass an additional tmpdir param to the class initialization. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5318 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-12perf profiler: allow events to be specified as an iterablelmr1-4/+9
Limiting the events list to be a comma-separated string doesn't work for the more complex syntaxes perf now supports for things like kprobes; allow the list to be given as an iterable (while still supporting a string) and pass each using a separate -e argument Signed-off-by: Scott James Remnant <keybuk@chromium.org> git-svn-id: svn://test.kernel.org/autotest/trunk@5316 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Turning off debug output for frequently called monitor commandslmr4-13/+13
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5315 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Introducing monitor commands and response logginglmr1-2/+21
In several KVM tests, it is very useful to log commands sent to the monitor and their responses, but most of the time output from such commands in verify_responsive and screendump threads are not necessary. So, introduce the param debug in monitor.cmd(), that defaults to True, and then later turn off selectively outputs from the parts of the framework where monitor commands are excessively called and their output is not so interesting. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5314 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Introduce ntpdate in the package list of all RH-like guestslmr11-0/+11
As this utility is needed to execute the ntp time drift variants successfuly. I would need a little help making the same maintenance on SUSE-based guests, as I don't have those guests handy here. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5313 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11Removing block_scsi tests from RHEL variantlmr1-0/+1
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5312 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: iofuzz: Fix exception formatting buglmr1-2/+2
Show both operation and operand when VM quits abnormally. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5311 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Fixing dependencies of ksm_overcommit testlmr1-1/+1
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5310 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: installer: Fix bug on patch applying codelmr1-3/+3
Patch command was missing <, let's fix it. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5309 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Cleaning unused imports on test_setuplmr1-3/+1
Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5308 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Moving unattended_install setup to test codelmr4-498/+498
Also, set start_vm = no Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5307 592f7852-d20e-0410-864c-8624ca9c26a4
2011-04-11KVM test: Moving enospc test setup to test codelmr4-104/+99
Michael Goldish has pointed out that some of the test setup code contained inside test_setup could be easily moved to the tests themselves, so we avoid some 'hardcoding' present on kvm_preprocessing. So let's start by moving enospc test setup to the test code itself. The only caveat here is that we need to specify start_vm = no so test setup can happen before the vm is started. Signed-off-by: Lucas Meneghel Rodrigues <lmr@redhat.com> git-svn-id: svn://test.kernel.org/autotest/trunk@5306 592f7852-d20e-0410-864c-8624ca9c26a4
generated by cgit v1.2.3 (git 2.39.1) at 2025-09-21 04:39:29 +0000