| Age | Commit message (Collapse) | Author |
|---|
|
Swap two unittest deprecated calls for their supported coutnerparts
* s/assertRaisesRegexp/assertRaisesRegex/
* s/assertEquals/assertEqual/
Also fix whitespace lint issues created from doing the above.
|
|
|
|
This reverts commit abebab1e1e30a16d4be891a051060e81254ff0c5.
|
|
|
|
(cherry picked from commit 7c7c29b008128ee5476cfa82b24762c18a0d2fcc)
|
|
(cherry picked from commit 5acd262e37cd3f2ae2b6134bf568806bd2736b0e)
|
|
(cherry picked from commit 48e5bc483e15d56b456fe629c8a9544cf61d8d64)
|
|
It's EOL, the required images are no longer in the MAAS image streams
and at least one vmtest fails if you manage to run it somehow.
|
|
* Use vmtest-{add,remove}-release to remove Groovy, and add Impish.
* vmtest-{add,remove}-release: set default path
|
|
Add a tool to generate new release classes based on a previous
release, for example:
./tools/vmtest-add-release -p tests/vmtests \
--previous-release=focal --distro-release=hirsute
User still needs to create the release class in tests/vmtests/releases.py
|
|
|
|
|
|
Special care and handling are needed for creating swap files on top
of btrfs filesystems. Curtin will attempt to disable btrfs CoW
on the target file before attempting to fallocate/dd the file.
- Btrfs swapfile requires target kernel 5.0+, older kernels cannot use.
https://btrfs.wiki.kernel.org/index.php/FAQ
- Query target fstype to check if we can proceed with swap file
btrfs requires newer kernel, xfs cannot use fallocate
- Update distro.parse_dpkg_version to handle additional formatting
including linux-image-generic, and versions with Epoch present.
- Adjust TestBasic,TestScsiBasic to add a swapfile to the /btrfs target
- Add the swap partition/file to TestBasic's fstab unittest
- Fix test_swaps_used to use fstab data instead of storage config
since file-based swaps are created via curtin 'swap' config not
storage-config.
- Add swap config key 'force', default is false
- Add curtin feature flag, BTRFS_SWAPFILE
LP: #1884161
|
|
- Drop skiptests for vlan issue, now fixed
- Drop skiptest on Ubuntu ReuseESP tests
- Add skiptest for Centos, new Bug #1881030
- Drop skip-by-date for CentosTestScsiBasic
Bug #1859858 is fixed
- Bump skip_by_date on tests with bcache by-uuid symlinks,
Bug #1861951 not fixed yet.
|
|
- For virtio-disks (TestBasic) use dname, except on centos66, use the byid
serial name.
- For scsi-disks (TestScsiBasic) use wwn everywhere
LP: #1874100
|
|
This handles unstable scsi device names (sdc/sdd) which shuffle around
from time to time; instead look up the assigned name via the dname
value. This should fix vmtest failures on BasicScsi on ppc64le/arm64.
|
|
The TestBasic currently fails on s390x due to usage of virtio-disks
and an "empty" virtio-scsi device controller. There is an existing
bug in s390x firmware that does not like an "empty" virtio-scsi
controller. Until that bug is fixed, skip TestBasic on s390x and
instead make TestScsiBasic work. This requires capturing output of
/dev/disk/by-path as s390x fstab entries prefer by-path as paths on
s390x are stable[1].
1. https://wiki.ubuntu.com/FSTAB
LP: #1866663
|
|
Remove disco tests, adding a new tool which can automate removal
of tests that match the release name.
|
|
Entries in /etc/fstab currently use UUID= in almost all cases. This
is not reliable in the face of LVM snapshots, multipath, and other
scenarios where multiple block devices may have the same UUID values.
Resolve this by following the recommendations in the Ubuntu Wiki[1]
where we will use specific udev persistent block paths which embed
a *device* uuid in the path (dm-uuid, md-uuid, etc). In addition
to modifying the fs_spec value we write, also include a comment
above each entry which includes a terse device path which may help
users recogize a disk where a UUID will not. This brings curtin
to parity with other Ubuntu installers which include:
# /mnt was mounted on /dev/sdb1 during installation
LP: #1851899
|
|
In Focal, there is a bug (#1859858) in udev which returns a
truncated value for ID_SERIAL on some scsi devices; this does
not break dname rules when the ephemeral environment matches
the target system, however if Focal deploys Bionic or in this
case Centos70, the udev in the target system produces a
different ID_SERIAL value which prevents the dname rules from
matching. Skip this test until we get a udev fix.
|
|
- Add the new release
- Switch some old Cosmic cases to Disco
- Add CentosXXFromFocal relbase
- Bump Skip-by dates for lvmroot, MTU and VLAN issues
|
|
The TestScsiBasic test has some assumptions on which SCSI device
will have expected content. On PPC64, the PCI device enumeration
is not as stable as it is on x86 resulting in test failures. This
patch resolves this by leveraging the dname values and looking up
the correct kernel name for a specific target dname.
|
|
* Call qemu with the correct -machine/-cpu options
* Default to UEFI boot
* Boot the Ubuntu "ga" arm64 kernel
* Disable the unsupported configurations (e.g. CentOS tests)
vmtest: bump the VM memory for Cosmic to 2GB
|
|
Changes:
- On ppc64el inject a PReP partition to the storage configuration
- Expect a GPT partition table and not a DOS one on ppc64el
- Skip XenialGAi386TestBasic if target_arch != amd64
- Skip XenialGAi386TestZfsRoot if target_arch != amd64
- Skip XenialGAi386TestZfsRootFsType if target_arch != amd64
- Skip test_partition_is_zero on ppc64el (LP: #1843288)
- Skip all the CentOS tests if target_arch = ppc64el
|
|
In tests 'arch' is now the arch of the host system, while target_arch
is the architecture the test targets. By default target_arch = arch.
Modify find_arches() to return all the target arches which not excluded
with arch_skip.
|
|
|
|
These are now passing, so we no longer need to skip them.
|
|
Newer MAAS will use Bionic to deploy Centos and curtin until now
did not test this path. Critically, Centos66 has issues with ext4
features that are enabled by default in Bionic and newer, namely
'64bit' and 'metadata_csum'. These features break centos66 grub
probing which results in failed deployment, LP: #1775424.
Add additional Centos66/70 tests to verify installing from Bionic
is successful. For Centos66, we test the known workaround which is
to use 'ext3' as the root filesystem, and we introduce a test-case
which recreates the failure (test_simple.py:Centos66BionicTestSimple)
but leave it disabled by default.
|
|
- Add support for detecting multipath devices and members
- Add handling to clear-holders for removing multipath partitions
- Update block.lookup_disk to return multipath member devices
- Add support to "hide" a multipath dm device while operating on
a single path member.
- vmtest: update xkvm to pass serial/wwn correctly for multipath
udev rules
- vmtest: update basic_scsi configuration and drop unused extra
disk.
LP: #1813228
|
|
|
|
Disco images now contain multipath packages and daemon enabled.
This affects how partitions are found when the underlying disk
is mpath enabled. This will be addressed in another branch, for
now disable the tests affected until it lands.
|
|
Trusty is EOL now, we can drop Trusty in vmtest to recover some
of that sweet, sweet CPU and IO time.
|
|
|
|
Currently /dev/disk/by-dname symlinks are not created
if a device does not have a partition table or a super block.
For devices which do have a serial or wwn number, curtin will
append an additional rule to the disk rules file which includes
both the partition table/superblock UUID as well as rule matching
the device provided serial or wwn. This ensures that even if
a block device is wiped that the dname will continue to function.
Other changes:
- add method to discover udev database value of a block device
ID_SERIAL value.
- refactor nvme vmtests to use 'serial' values
- don't hardcode nvme serials based on their enumerated index
- ensure we pass nvme serials to both install and boot phases
- update disks_to_check for basic and nvme to ensure we have a
dname for the disk itself
LP: #1735839
|
|
|
|
The partition handler in block_meta would wipe in "superblock" mode
before creating a partition but this only clears 1MB at the start
of the partition. Curtin needs to respect the wipe mode which may
wipe the location of the device differently.
Also:
- Document pre-partitioning superblock wiping
- Add PReP GPT config to basic/basic_scsi test
- Sync basic and basic_scsi partition config
- collect ls -al on /dev/disk/by-partuuid
LP: #1800153
|
|
If collect scripts exit non-zero, then cloud-init in the booted system
will complain that they have done so. This adds noise to logs.
At this point, the only command in a long list that affects the exit code
is the last one. At some point in the future we may have more systematic
collection. At that point it might make sense to exit if the collection
failed.
For now, just exit 0.
|
|
Add support for the majority of storage configurations including
partitioning, lvm, raid, iscsi and combinations of these. Some
storage configs are unsupported at this time:
Unsupported storage config options on Centos:
- bcache (no kernel support)
- zfs (no kernel support)
- jfs, ntfs, reiserfs (no kernel, userspace support)
Curtin's built-in curthooks now support Centos in addition
to Ubuntu. The built-in curthooks are now callable by
in-image curthooks. This feature is announced by the
presence of the feature flag, 'CENTOS_CURTHOOK_SUPPORT'
Other notable features added:
- tools/jenkins-runner now includes a test filtering
ability which enables generating the list of tests to
run by specifying attributes of the classes. For example
to run all centos70 tests append:
--filter=target_release=centos70
- curtin/distro.py includes distro specific methods, such as
package install and distro version detection
- util.target_path has now moved to curtin.paths module
|
|
Ubuntu Artful has reached EOL and as such this removes the Artful tests.
|
|
Simplify some code in vmtests by making sure that tests each have
release, distro, target_release, target_distro.
Then remove some redundant test cases for > artful and debug
use of 'btrfs inspect-internal'.
|
|
|
|
The zfs kernel module is not available for i386 so curtin cannot
expect that missing that module is fatal. Concentrate the various
checks for when we load the zfs kernel module into the zfs python
module. Add vmtest to exercise the i386 path.
LP: #1768709
|
|
bcache minor numbers are not stable across reboots which means
the dname symlinks pointing to a bcacheN isn't going to be stable either.
Instead, build a udev rule that links the backing device's bcache uuid
to the assigned dname. Additionally also update the backing device
'label' sysfs attribute with the dname value.
This does rely on changes to bcache-tools to trigger the rules, currently
there is a kernel bug (LP: #1729145) which is pending, and a user-space
workaround in bcache-tools (LP: #1728742). Either is sufficent for
the updated rules to be effective.
Some additional fixes found during this work
command/curthooks.py
- don't append .rule to rule files, they already have .rule
examples/tests/mdadm_bcache_complex.yaml
- removed, duplicate of mdadm_bcache.yaml
|
|
There is no actual code change here. The changes are
a.) move LICENSE file to LICENSE-AGPLv3
b.) put short form of AGPL-v3 license in LICENSE file,
and mention Copyright info there.
c.) remove license headers in python or shell files.
d.) put a single line header in all python or shell files.
e.) put a single footer line in all python or shell files.
There is a change here from "AGPLv3 or later" to just "AGPLv3".
|
|
Zesty is no longer supported as of 2018-01-13[1]
1. https://lists.ubuntu.com/archives/ubuntu-announce/2018-January/000227.html
|
|
Switching to v3 streams enables testing of Xenial HWE and HWE-edge
kernels and the squashfs root which MAAS 2.3 uses.
- Add v3 streams URL
- Introduce two attributes, ephemeral_ftype, target_ftype to determine
which files are needed for the ephemeral environment and which for the
target
- Add kflavor and subarch filters to streams query and release classes to
support kernel variants: ga, hwe, hwe-edge. We're currently only testing
subarch 'generic' and ignoring 'low-latency'
- When testing with hwe and hwe-edge, emit a kernel configuration to
curtin to ensure we install the matching kernel in the target image
- Make sure vmtests append collect scripts to the base class parameter
so all tests run the BaseVM collect_scripts (dpkg-query)
- Introduce new release classes, XenialGA, XenialHWE and XenialEdge
Note, XenialGA replaces the general Xenial release class, and that
not all test-cases run Xenial HWE/Edge kernels; most of the storage
ones do.
- Booting from a squashfs image requires some changes
- publish the squashfs image and boot VMs with root=URL (except trusty)
- utilize overlayfs to keep the ephemeral environment pristine and
use the squashfs image as a the curtin install source.
- Add some boot parameters to the ephemeral environment to enable
networking, including DNS for package installation.
- Use BOOTIF= parameter for multi-nic vmtests to ensure we only bring
up one nic for ephemeral networking. Sometimes multiple nics would
come online and then cause networking hiccups which prevented
installs from completing.
- Modify vmtest storage configurations for disks to all use a 'serial'
attribute. Currently we use 'disk-X' as the serial value where X
increments from 'a' -> 'z' as the test needs.
- jenkins-runner now prints additional information about what images
are found and available.
- tools/launch
- modify --append to allow multiple parameters to grow the array.
- fix bug in --publish which prevented more than a single file being
published
- allow PUBURL to be expanded in root-arg
- add a 'boot-image' parameter which is exported to the VM with a serial
number to provide a consistent path to the device,
/dev/disk/by-id/virtio-boot-disk
- drop '2G' boot size parameter to boot disks, instead allow qemu-img to
default to the size of backing image.
|
|
bzr-revno: 545
|
|
bzr-revno: 519.3.9
|
|
bzr-revno: 542.1.1
|
|
bzr-revno: 540.1.1
|
