Etherpad: https://autotest.etherpad.mozilla.org/1?
Ok guys, so I ended the 1st video by accident, sorry about that. We had
to restart and pick up from where we left. If you want to only see the
discussion of the RFCs, jump right to part 2 (part 1 was mostly
discussing minor, day-to-day issues with virt-test and autotest, so not
that interesting).
Youtube Video: (part 1): https://www.youtube.com/watch?v=OAbq_YDNCXs
Youtube Video: (part 2): https://www.youtube.com/watch?v=NJrH_z1ryr8
Highlights
Kudos to the virt-test project maintainers, we were able to shoot all
the bugs found last week in about 4 hours, pretty good!
Virt Test still keeps a good rate of reviews, and some mistakes prompted
an RFC to change the release management process used by autotest
projects. Let's discuss it during today's meeting
During the KVM Forum we had several discussions about the future of
autotest and virt-tests. A RFC with the ideas and a proposed backlog has
been sent to the mailing lists. We'll briefly discuss it in today's meeting.
Commits on next - autotest
Pull request 762 in fact proved to have a bug on automated regression
testing, and was reverted. Other than this hiccup, some unittests that
broke with the new utils.rmdir_safe were fixed. Nothing too big this week.
Feng Yang (1):
client.profilers.systemtap: Do not fail jobs when stap check
command fail.
Lucas Meneghel Rodrigues (6):
Merge pull request #763 from FengYang/systemtap
Merge pull request #762 from rbbratta/pep0263
Revert "autotest_remote: Fix BUG when control files require PEP
0263 source encoding"
Merge pull request #765 from rbbratta/install_fix1
client.job_unittest: Fix unittest broken with 3263bc8
client.setup_job_unittest: Fix unittest broken with 3263bc8
Ross Brattain (2):
autotest_remote: Fix BUG when control files require PEP 0263
source encoding
install-autotest-server.sh: relocate apache-web.conf to new $ATHOME
Commits on next - virt-test
This was a fairly low activity week, Xu Tian continues to produce some
good quality tests and libraries, such as the clock_offset tests for
qemu, and the lvm module.
Dongsheng Yang (2):
libvirt_vm: Add method to create install cmd with sec_label
guest-os: Add filter for svirt_install in guest-os.
Hao Liu (2):
virsh_change_media: Backup and restore original vm xml
libvirt: virsh_edit: Add maximum option for vcpucount
Lucas Meneghel Rodrigues (5):
Merge pull request #1049 from ruda/qemu-vm-monitor-logs
Merge pull request #908 from xutian/lvm_support
Merge pull request #974 from xutian/add_timedrift_case
Merge pull request #909 from yangdongsheng/svirt_install
Merge branch 'master' into next
Ruda Moura (1):
virttest.qemu_vm: Log information about the parent PID and the
monitor name.
Xu Tian (13):
qemu.tests: new case to check guest clock with syscall
qemu.tests: add timedrift_check_clock_offset cfg
qemu.tests: add check clock offset when vm crash/bsod
qemu.tests: add cfg for check clock offset when guest bsod/crash
qemu.tests: new case to check clock offset after hotplug vcpus
qemu.tests: cfg for checking clock offset after hotplug vcpu
qemu.tests: check kvm-clock using newer msr set
qemu.tests: cfg for kvm-clock newer msrs support check test
virttest: add interface to save lvm object to env file
virttest: LVM storage base module
virttest: add emulated lvm support
virttest: add lvm storage support for qemu tests
shared.cfg: add params for supporting lvm partiton
yangdongsheng (2):
Add test for svirt in virt-install.
Merge pull request #1055 from Hao-Liu/libvirt_ci
Pull requests still in review
Autotest
https://github.com/autotest/autotest/pull/764 - Pull request that adds
an extra param to software manager .upgrade() method. Ross was kind
enough to provide a review, I'm implementing his suggestions and this is
going to be pushed today.
https://github.com/autotest/autotest/pull/760 - Ruda's documentation
support to autotest. He'll rebase the PR and we'll push it today, hopefully.
Virt Test
14 pull requests open, pretty good number overall, reflects a slowdown
on the pace of PRs.
https://github.com/autotest/virt-test/pull/1069 qemu.tests: Use qemu-kvm
-M ? to get machine type for physical resource ...
https://github.com/autotest/virt-test/pull/1068 virttest: Update
khugepaged test timeout
https://github.com/autotest/virt-test/pull/1067 Add option '--log-level'
to control log level for the top level log file
https://github.com/autotest/virt-test/pull/1064 libguestfs.tests:
Enhancements of guestmount test.
https://github.com/autotest/virt-test/pull/1063 libvirt.tests: Support
multiple devices for multifunction test.
https://github.com/autotest/virt-test/pull/1062 libvirt-bench: Add test
for domstate switching with iozone.
https://github.com/autotest/virt-test/pull/1060 Environment bugfix
https://github.com/autotest/virt-test/pull/1053 Mq change qnum
https://github.com/autotest/virt-test/pull/1052 Libguestfs: Enhancements
of guestfs_add test.
https://github.com/autotest/virt-test/pull/1051 Virsh connect back
https://github.com/autotest/virt-test/pull/1046 virt-libguestfs: Add
sparsify case for virt-sysprep test
https://github.com/autotest/virt-test/pull/1027 libvirt.tests: Move
remote about testcases to virsh.remote
https://github.com/autotest/virt-test/pull/991 Virsh update device
[working it...]
https://github.com/autotest/virt-test/pull/989 Libguestfs: Add test for
guestfish volume operations.
https://github.com/autotest/virt-test/pull/987 Add test for virsh
find-storage-pool-sources-as and bug fix
https://github.com/autotest/virt-test/pull/978 virt.virtnet : update
vm.virtnet when guest reboot
https://github.com/autotest/virt-test/pull/948 Virsh Snapshot-edit test
Grid maintenance
We had a bunch of problems with broken F19 install, forcing us to
disable all jobs that use this host profile. Currently investigating
what is wrong
New release management process
During the KVM forum, it was pointed out to us that the 'next' branch is
something that differs from the conventions used by other projects, and
therefore causes confusion. Another practice that is frowned upon is
non-ff updates to branches on a public, reference repo. So, the proposal
in a nutshell is:
* Remove 'next' branches, and use master as the dev branch. No stability
promised on master.
* Offer release tags for people wanting to use QA'ed code
I do understand people don't like the fast pace of the project and its
changes, but every "stakeholder" in the project has different
priorities, and it is always hard to juggle all those priorities.
The future of autotest and virt-test
See "RFC: Autotest and virt-test upcoming changes" e-mail thread on the
mailing lists
Libvirt updates
Ramping up for automated mass-testing
Lots of bug reporting & fixing
Questions Corner
TODO
lmr
* Test grid upgrades, stability work
* A lot of bugs to fix
* A lot less time dedicated to review
cevich
* 100% commited to https://github.com/autotest/virt-test/issues/1058
https://github.com/autotest/virt-test/pull/978
cleber
Work on arc + virt-test-job (for makefile target consumption)
Push installer + component isolation (pending bits) work upstream
jzupka
pull request work
fix bugs
glusterfs
rmoura
* Rebase API documentation for autotest
[https://github.com/ruda/autotest/commits/api-doc2]
* Anything else with logs in autotest or virt-test (create logs with
different log levels?)
* Follow up problem with qemu with sandbox ON and virt-test
[https://bugzilla.redhat.com/show_bug.cgi?id=1027421]
ldoktor
* Until the end of the year working only 2-3 days per week
* fix my bugs
* review
* in the mean time improve the pci_devices test
kvm-qe (BJ):
Prepare to update our testing tree. So will have more bug fix in the
following weeks.
fyang: Bug fix and device assignmnet
xutian: timer device
yuzheng: new case for zero copy and s3/s4 during and after network load
update patches for multi queue
ypu: Bug fix
_______________________________________________
Virt-test-devel mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/virt-test-devel
