os-autoinst-distri-rocky/lib/utils.pm

351 lines
13 KiB
Perl
Raw Normal View History

package utils;
use strict;
use base 'Exporter';
use Exporter;
use lockapi;
use testapi;
our @EXPORT = qw/run_with_error_check type_safely type_very_safely desktop_vt boot_to_login_screen console_login console_switch_layout desktop_switch_layout console_loadkeys_us do_bootloader get_milestone boot_decrypt check_release menu_launch_type start_cockpit repo_setup/;
sub run_with_error_check {
my ($func, $error_screen) = @_;
die "Error screen appeared" if (check_screen $error_screen, 5);
$func->();
die "Error screen appeared" if (check_screen $error_screen, 5);
}
# high-level 'type this string quite safely but reasonably fast'
# function whose specific implementation may vary
sub type_safely {
my $string = shift;
type_string($string, wait_screen_change => 3, max_interval => 20);
wait_still_screen 2;
}
# high-level 'type this string extremely safely and rather slow'
# function whose specific implementation may vary
sub type_very_safely {
my $string = shift;
type_string($string, wait_screen_change => 1, max_interval => 1);
wait_still_screen 5;
}
# Figure out what tty the desktop is on, switch to it. Assumes we're
# at a root console
sub desktop_vt {
# use ps to find the tty of Xwayland or Xorg
my $xout;
# don't fail test if we don't find any process, just guess tty1
eval { $xout = script_output 'ps -C Xwayland,Xorg -o tty --no-headers'; };
my $tty = 1; # default
while ($xout =~ /tty(\d)/g) {
$tty = $1; # most recent match is probably best
}
send_key "ctrl-alt-f${tty}";
}
consolidate login waits, use postinstall not entrypoint for base Summary: I started out wanting to fix an issue I noticed today where graphical upgrade tests were failing because they didn't wait for the graphical login screen properly; the test was sitting at the 'full Fedora logo' state of plymouth for a long time, so the current boot_to_login_screen's wait_still_screen was triggered by it and the function wound up failing on the assert_screen, because it was still some time before the real login screen appeared. So I tweaked the boot_to_login_screen implementation to work slightly differently (look for a login screen match, *then* - if we're dealing with a graphical login - wait_still_screen to defeat the 'old GPU buffer showing login screen' problem and assert the login screen again). But while working on it, I figured we really should consolidate all the various places that handle the bootloader -> login, we were doing it quite differently in all sorts of different places. And as part of that, I converted the base tests to use POSTINSTALL (and thus go through the shared _wait_login tests) instead of handling boot themselves. As part of *that*, I tweaked main.pm to not require all POSTINSTALL tests have the _postinstall suffix on their names, as it really doesn't make sense, and renamed the tests. Test Plan: Run all tests, see if they work. Reviewers: jskladan, garretraziel Reviewed By: garretraziel Subscribers: tflink Differential Revision: https://phab.qadevel.cloud.fedoraproject.org/D1015
2016-09-27 18:48:15 +00:00
# Wait for login screen to appear. Handle the annoying GPU buffer
# problem where we see a stale copy of the login screen from the
# previous boot. Will suffer a ~30 second delay if there's a chance
# we're *already at* the expected login screen.
sub boot_to_login_screen {
my %args = @_;
$args{timeout} //= 300;
# we may start at a screen that matches one of the needles; if so,
# wait till we don't (e.g. when rebooting at end of live install,
# we match text_console_login until the console disappears)
my $count = 5;
while (check_screen("login_screen", 3) && $count > 0) {
sleep 5;
$count -= 1;
}
assert_screen "login_screen", $args{timeout};
if (match_has_tag "graphical_login") {
wait_still_screen 10, 30;
assert_screen "login_screen";
}
}
# Switch keyboard layouts at a console
sub console_switch_layout {
# switcher key combo differs between layouts, for console
if (get_var("LANGUAGE", "") eq "russian") {
send_key "ctrl-shift";
}
}
# switch to 'native' or 'ascii' input method in a graphical desktop
# usually switched configs have one mode for inputting ascii-ish
# characters (which may be 'us' keyboard layout, or a local layout for
# inputting ascii like 'jp') and one mode for inputting native
# characters (which may be another keyboard layout, like 'ru', or an
# input method for more complex languages)
# 'environment' can be a desktop name or 'anaconda' for anaconda
# if not set, will use get_var('DESKTOP') or default 'anaconda'
sub desktop_switch_layout {
my ($layout, $environment) = @_;
$layout //= 'ascii';
$environment //= get_var("DESKTOP", "anaconda");
# if already selected, we're good
return if (check_screen "${environment}_layout_${layout}", 3);
# otherwise we need to switch
my $switcher = "alt-shift"; # anaconda
$switcher = "super-spc" if $environment eq 'gnome';
# KDE? not used yet
send_key $switcher;
assert_screen "${environment}_layout_${layout}", 3;
}
# this subroutine handles logging in as a root/specified user into console
# it requires TTY to be already displayed (handled by the root_console()
# method of distribution classes)
sub console_login {
my %args = (
user => "root",
password => get_var("ROOT_PASSWORD", "weakpassword"),
@_);
# There's a timing problem when we switch from a logged-in console
# to a non-logged in console and immediately call this function;
# if the switch lags a bit, this function will match one of the
# logged-in needles for the console we switched from, and get out
# of sync (e.g. https://openqa.stg.fedoraproject.org/tests/1664 )
# To avoid this, we'll sleep a couple of seconds before starting
sleep 2;
my $good = "";
my $bad = "";
if ($args{user} eq "root") {
$good = "root_console";
$bad = "user_console";
}
else {
$good = "user_console";
$bad = "root_console";
}
if (check_screen $bad, 0) {
# we don't want to 'wait' for this as it won't return
script_run "exit", 0;
sleep 2;
}
check_screen [$good, 'text_console_login'], 10;
# if we're already logged in, all is good
return if (match_has_tag $good);
# if we see the login prompt, type the username
type_string("$args{user}\n") if (match_has_tag 'text_console_login');
check_screen [$good, 'console_password_required'], 30;
# on a live image, just the user name will be enough
return if (match_has_tag $good);
# otherwise, type the password if we see the prompt
if (match_has_tag 'console_password_required') {
type_string "$args{password}";
if (get_var("SWITCHED_LAYOUT") and $args{user} ne "root") {
# see _do_install_and_reboot; when layout is switched
# user password is doubled to contain both US and native
# chars
console_switch_layout;
type_string "$args{password}";
console_switch_layout;
}
send_key "ret";
}
# make sure we reached the console
assert_screen($good, 30);
}
# load US layout (from a root console)
sub console_loadkeys_us {
if (get_var('LANGUAGE') eq 'french') {
script_run "loqdkeys us", 0;
}
}
sub do_bootloader {
# Handle bootloader screen. 'bootloader' is syslinux or grub.
# 'uefi' is whether this is a UEFI install, will get_var UEFI if
# not explicitly set. 'postinstall' is whether we're on an
# installed system or at the installer (this matters for how many
# times we press 'down' to find the kernel line when typing args).
# 'args' is a string of extra kernel args, if desired. 'mutex' is
# a parallel test mutex lock to wait for before proceeding, if
# desired. 'first' is whether to hit 'up' a couple of times to
# make sure we boot the first menu entry. 'timeout' is how long to
# wait for the bootloader screen.
my %args = (
postinstall => 0,
params => "",
mutex => "",
first => 1,
timeout => 30,
uefi => get_var("UEFI"),
@_
);
# if not postinstall and not UEFI, syslinux
$args{bootloader} //= ($args{uefi} || $args{postinstall}) ? "grub" : "syslinux";
if ($args{uefi}) {
# we use the firmware-type specific tags because we want to be
# sure we actually did a UEFI boot
assert_screen "bootloader_uefi", $args{timeout};
} else {
assert_screen "bootloader_bios", $args{timeout};
}
if ($args{mutex}) {
# cancel countdown
send_key "left";
mutex_lock $args{mutex};
mutex_unlock $args{mutex};
}
if ($args{first}) {
# press up a couple of times to make sure we're at first entry
send_key "up";
send_key "up";
}
if ($args{params}) {
if ($args{bootloader} eq "syslinux") {
send_key "tab";
}
else {
send_key "e";
# ternary: 13 'downs' to reach the kernel line for installed
# system, 2 for UEFI installer
my $presses = $args{postinstall} ? 13 : 2;
foreach my $i (1..$presses) {
send_key "down";
}
send_key "end";
}
type_string " $args{params}";
}
# ctrl-X boots from grub editor mode
send_key "ctrl-x";
# return boots all other cases
send_key "ret";
}
sub get_milestone {
# FIXME: we don't know how to do this with Pungi 4 yet.
return '';
}
sub boot_decrypt {
# decrypt storage during boot; arg is timeout (in seconds)
my $timeout = shift || 60;
assert_screen "boot_enter_passphrase", $timeout; #
type_string get_var("ENCRYPT_PASSWORD");
send_key "ret";
}
sub check_release {
# Checks whether the installed release matches a given value. E.g.
# `check_release(23)` checks whether the installed system is
# Fedora 23. The value can be 'Rawhide' or a Fedora release
# number; often you will want to use `get_var('VERSION')`. Expects
# a console prompt to be active when it is called.
my $release = shift;
my $check_command = "grep SUPPORT_PRODUCT_VERSION /usr/lib/os.release.d/os-release-fedora";
validate_script_output $check_command, sub { $_ =~ m/REDHAT_SUPPORT_PRODUCT_VERSION=$release/ };
}
sub menu_launch_type {
# Launch an application in a graphical environment, by opening a
# launcher, typing the specified string and hitting enter. Pass
# the string to be typed to launch whatever it is you want.
my $app = shift;
# super does not work on KDE, because fml
send_key 'alt-f1';
# srsly KDE y u so slo
wait_still_screen 3;
type_very_safely $app;
send_key 'ret';
}
sub start_cockpit {
# Starting from a console, get to a browser with Cockpit (running
# on localhost) shown. If $login is truth-y, also log in. Assumes
# X and Firefox are installed.
my $login = shift || 0;
# run firefox directly in X as root. never do this, kids!
type_string "startx /usr/bin/firefox -width 1024 -height 768 http://localhost:9090\n";
assert_screen "cockpit_login";
wait_still_screen 5;
if ($login) {
type_safely "root";
wait_screen_change { send_key "tab"; };
type_safely get_var("ROOT_PASSWORD", "weakpassword");
send_key "ret";
assert_screen "cockpit_main";
# wait for any animation or other weirdness
# can't use wait_still_screen because of that damn graph
sleep 3;
}
}
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
sub _repo_setup_compose {
# Appropriate repo setup steps for testing a compose
# disable updates-testing and updates and use the compose location
# as the target for fedora and rawhide rather than mirrorlist, so
# tools see only packages from the compose under test
my $location = get_var("LOCATION");
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
return unless $location;
assert_script_run 'dnf config-manager --set-disabled updates-testing updates';
# we use script_run here as the rawhide repo file won't always exist
# and we don't want to bother testing or predicting its existence;
# assert_script_run doesn't buy you much with sed anyway as it'll
# return 0 even if it replaced nothing
script_run "sed -i -e 's,^metalink,#metalink,g' -e 's,^#baseurl.*basearch,baseurl=${location}/Everything/\$basearch,g' /etc/yum.repos.d/{fedora,fedora-rawhide}.repo", 0;
script_run "cat /etc/yum.repos.d/{fedora,fedora-rawhide}.repo", 0;
}
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
sub _repo_setup_updates {
# Appropriate repo setup steps for testing a Bodhi update
# Check if we already ran, bail if so
return unless script_run "test -f /etc/yum.repos.d/advisory.repo";
# Use baseurl not metalink so we don't wind up timing out getting
# metadata from a slow source...baseurl will always give us dl
# in infra
assert_script_run "sed -i -e 's,^metalink,#metalink,g' -e 's,^#baseurl,baseurl,g' /etc/yum.repos.d/fedora*.repo";
# Set up an additional repo containing the update packages. We do
# this rather than simply running a one-time update because it may
# be the case that a package from the update isn't installed *now*
# but will be installed by one of the tests; by setting up a repo
# containing the update and enabling it here, we ensure all later
# 'dnf install' calls will get the packages from the update.
assert_script_run "mkdir -p /opt/update_repo";
assert_script_run "cd /opt/update_repo";
assert_script_run "dnf -y install bodhi-client git createrepo", 300;
# download the packages
my $version = lc(get_var("VERSION"));
if ($version eq 'rawhide' || $version > 25) {
# bodhi client 2.x
assert_script_run "bodhi updates download --updateid " . get_var("ADVISORY"), 300;
}
else {
# bodhi client 0.9
# use git python-fedora for
# https://github.com/fedora-infra/python-fedora/pull/192
# until packages with that fix are pushed stable
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
assert_script_run "git clone https://github.com/fedora-infra/python-fedora.git";
assert_script_run "PYTHONPATH=python-fedora/ bodhi -D " . get_var("ADVISORY"), 300;
}
# log the exact packages in the update at test time, with their
# source packages and epochs. log is uploaded by _advisory_update
# and used for later comparison by _advisory_post
assert_script_run 'rpm -qp *.rpm --qf "%{SOURCERPM} %{EPOCH} %{NAME}-%{VERSION}-%{RELEASE}\n" | sort -u > /var/log/updatepkgs.txt';
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
# create the repo metadata
assert_script_run "createrepo .";
# write a repo config file
assert_script_run 'printf "[advisory]\nname=Advisory repo\nbaseurl=file:///opt/update_repo\nenabled=1\nmetadata_expire=3600\ngpgcheck=0" > /etc/yum.repos.d/advisory.repo';
# run an update now
script_run "dnf -y update", 600;
Add support for testing updates Summary: This adds an entirely new workflow for testing distribution updates. The `ADVISORY` variable is introduced: when set, `main.pm` will load an early post-install test that sets up a repository containing the packages from the specified update, runs `dnf -y update`, and reboots. A new templates file is added, `templates-updates`, which adds two new flavors called `updates-server` and `updates-workstation`, each containing job templates for appropriate post-install tests. Scheduler is expected to post `ADVISORY=(update ID) HDD_1=(base image) FLAVOR=updates-(server|workstation)`, where (base image) is one of the stable release base disk images produced by `createhdds` and usually used for upgrade testing. This will result in the appropriate job templates being loaded. We rejig postinstall test loading and static network config a bit so that this works for both the 'compose' and 'updates' test flows: we have to ensure we bring up networking for the tap tests before we try and install the updates, but still allow later adjustment of the configuration. We take advantage of the openQA feature that was added a few months back to run the same module multiple times, so the `_advisory_update` module can reboot after installing the updates and the modules that take care of bootloader, encryption and login get run again. This looks slightly wacky in the web UI, though - it doesn't show the later runs of each module. We also use the recently added feature to specify `+HDD_1` in the test suites which use a disk image uploaded by an earlier post-install test, so the test suite value will take priority over the value POSTed by the scheduler for those tests, and we will use the uploaded disk image (and not the clean base image POSTed by the scheduler) for those tests. My intent here is to enhance the scheduler, adding a consumer which listens out for critpath updates, and runs this test flow for each one, then reports the results to ResultsDB where Bodhi could query and display them. We could also add a list of other packages to have one or both sets of update tests run on it, I guess. Test Plan: Try a post something like: HDD_1=disk_f25_server_3_x86_64.img DISTRI=fedora VERSION=25 FLAVOR=updates-server ARCH=x86_64 BUILD=FEDORA-2017-376ae2b92c ADVISORY=FEDORA-2017-376ae2b92c CURRREL=25 PREVREL=24 Pick an appropriate `ADVISORY` (ideally, one containing some packages which might actually be involved in the tests), and matching `FLAVOR` and `HDD_1`. The appropriate tests should run, a repo with the update packages should be created and enabled (and dnf update run), and the tests should work properly. Also test a regular compose run to make sure I didn't break anything. Reviewers: jskladan, jsedlak Reviewed By: jsedlak Subscribers: tflink Differential Revision: https://phab.qa.fedoraproject.org/D1143
2017-01-25 16:16:12 +00:00
}
sub repo_setup {
# Run the appropriate sub-function for the job
get_var("ADVISORY") ? _repo_setup_updates : _repo_setup_compose;
}