QA (public) » openQA Project (public) » openQA Tests (public) » qe-yam

Created WIP PR#20303

Merged: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/20338

Remaining part is:
To install Tumbleweed with the (current production) installer and compare the result with a installation with agama, both should contain the root and the swap partitions encrypted and similar characteristics of LUKSs, etc., if they don't then we stop here and file a bug (or discuss with squad in our channel if in doubt).
If those are similar, then let's add YAML_TEST_DATA to pass validation in validate_encrypt test module.

Repeat the same process to enable the test suite in Leap 16 and SLES 16 job groups.

Closed PR#20303

The current installation contains 2 encrypted device, verified also by local VM test.
But the test data eg

• test_data/yam/agama_tumbleweed_full_disk_encryption.yaml
• test_data/yam/agama_full_disk_encryption.yaml

contains:

---
crypttab:
  num_devices_encrypted: 1

The proposal is to create a new yaml test data for YAML_TEST_DATA that fit the actual number of devices

Could be enough to add the section test_data into yaml schedule file schedule/yam/agama_full_disk_encryption.yaml:

test_data:
  crypttab:
    num_devices_encrypted: 2
  <<: !include test_data/yast/encryption/default_enc_luks2.yaml

as in the Tumbleweed test case crypt_no_lvm

Created PR#20351

Merged PR#20351

Verified in production with build 25.64: agama_full_disk_encryption

Following the scope described in the ticket:

• Could you take a look to the failures in Leap: https://openqa.opensuse.org/tests/overview?distri=opensuse&version=agama-9.0&build=25.71&groupid=116
• S390x I don't think we have check at least what happen there (it might worth another ticket if requires big changes)
• Add the test suite to Sles 16: https://openqa.suse.de/group_overview/607?limit_builds=50

I think that for USB installation (agama_full_disk_encryption@uefi-usb-4G) we need a different YAML_SCHEDULE because after agama perl module has completed,
it shown again the same windows as in boot_agama perl module during the run of boot_full_disk_encryption, like the system is still need to install.
See:
https://openqa.opensuse.org/tests/overview?build=build-issues-167470
https://openqa.opensuse.org/tests/4549174#step/boot_full_disk_encryption/3
https://openqa.opensuse.org/tests/4549174#step/boot_agama/1

lmanfredi wrote in #note-12:

I think that for USB installation (agama_full_disk_encryption@uefi-usb-4G) we need a different YAML_SCHEDULE because after agama perl module has completed,
it shown again the same windows as in boot_agama perl module during the run of boot_full_disk_encryption, like the system is still need to install.
See:
https://openqa.opensuse.org/tests/overview?build=build-issues-167470
https://openqa.opensuse.org/tests/4549174#step/boot_full_disk_encryption/3
https://openqa.opensuse.org/tests/4549174#step/boot_agama/1

yes, that will be in another ticket to tackle, requires a workaround to go to hard disk, I was expecting to see it in some of the new scenarios we are adding.

Created needles with tag:

• enter-passphrase-for-root-partition
• enter-passphrase-for-swap-partition

on repo os-autoinst-needles-sles

enter-passphrase-for-root-partition-20241015.json:    "enter-passphrase-for-root-partition"
enter-passphrase-for-root-partition-x86_64-20241015.json:    "enter-passphrase-for-root-partition"
enter-passphrase-for-swap-partition-20241015.json:    "enter-passphrase-for-swap-partition"

for SLES.

Created os-autoinst-distri-opensuse PR#20399

Skipped test modules validate_encrypt in job template EXCLUDE_MODULES: validate_encrypt

Merged PR#20399