community.mysql/Makefile
Laurent Indermühle dd7e297d50
Add support for MariaDB 11.4 (#703)
* fix missing symlink to mysql binaries for MariaDB 11+

* update tested version of MariaDB 11.4 instead of 10.5

* add changelog fragment

* [CI] add way to trigger workflow manually

Useful in the case we don't modifiy any files in the paths: sections of the push event.

* add version check for mariadb < 10.4.6 without mariadb* binaries

* Use same concatenation method between functions to avoid future confusion

I didn't notice that db_dump and db_import were different, thus I introduced a bug with the initialization of the variable cmd. This commit fixes that.
2025-03-10 18:55:42 +01:00

116 lines
4.6 KiB
Makefile

SHELL := /bin/bash
# To tell ansible-test and Make to not kill the containers on failure or
# end of tests. Disabled by default.
ifdef keep_containers_alive
_keep_containers_alive = --docker-terminate never
endif
# This match what GitHub Action will do. Disabled by default.
ifdef continue_on_errors
_continue_on_errors = --continue-on-error
endif
# Set command variables based on database engine
# Required for MariaDB 11+ which no longer includes mysql named compatible
# executable symlinks
ifeq ($(db_engine_name),mysql)
_command = mysqld
_health_cmd = mysqladmin
else
_command = mariadbd
_health_cmd = mariadb-admin
endif
.PHONY: test-integration
test-integration:
@echo -n $(db_engine_name) > tests/integration/db_engine_name
@echo -n $(db_engine_version) > tests/integration/db_engine_version
@echo -n $(connector_name) > tests/integration/connector_name
@echo -n $(connector_version) > tests/integration/connector_version
@echo -n $(ansible) > tests/integration/ansible
# Create podman network for systems missing it. Error can be ignored
podman network create podman || true
podman run \
--detach \
--replace \
--name primary \
--env MARIADB_ROOT_PASSWORD=msandbox \
--env MYSQL_ROOT_PASSWORD=msandbox \
--network podman \
--publish 3307:3306 \
--health-cmd '$(_health_cmd) ping -P 3306 -pmsandbox | grep alive || exit 1' \
docker.io/library/$(db_engine_name):$(db_engine_version) \
$(_command)
podman run \
--detach \
--replace \
--name replica1 \
--env MARIADB_ROOT_PASSWORD=msandbox \
--env MYSQL_ROOT_PASSWORD=msandbox \
--network podman \
--publish 3308:3306 \
--health-cmd '$(_health_cmd) ping -P 3306 -pmsandbox | grep alive || exit 1' \
docker.io/library/$(db_engine_name):$(db_engine_version) \
$(_command)
podman run \
--detach \
--replace \
--name replica2 \
--env MARIADB_ROOT_PASSWORD=msandbox \
--env MYSQL_ROOT_PASSWORD=msandbox \
--network podman \
--publish 3309:3306 \
--health-cmd '$(_health_cmd) ping -P 3306 -pmsandbox | grep alive || exit 1' \
docker.io/library/$(db_engine_name):$(db_engine_version) \
$(_command)
# Setup replication and restart containers using the same subshell to keep variables alive
db_ver=$(db_engine_version); \
maj="$${db_ver%.*.*}"; \
maj_min="$${db_ver%.*}"; \
min="$${maj_min#*.}"; \
if [[ "$(db_engine_name)" == "mysql" && "$$maj" -eq 8 && "$$min" -ge 2 ]]; then \
prima_conf='[mysqld]\\nserver-id=1\\nlog-bin=/var/lib/mysql/primary-bin\\nmysql-native-password=1'; \
repl1_conf='[mysqld]\\nserver-id=2\\nlog-bin=/var/lib/mysql/replica1-bin\\nmysql-native-password=1'; \
repl2_conf='[mysqld]\\nserver-id=3\\nlog-bin=/var/lib/mysql/replica2-bin\\nmysql-native-password=1'; \
else \
prima_conf='[mysqld]\\nserver-id=1\\nlog-bin=/var/lib/mysql/primary-bin'; \
repl1_conf='[mysqld]\\nserver-id=2\\nlog-bin=/var/lib/mysql/replica1-bin'; \
repl2_conf='[mysqld]\\nserver-id=3\\nlog-bin=/var/lib/mysql/replica2-bin'; \
fi; \
podman exec -e cnf="$$prima_conf" primary bash -c 'echo -e "$${cnf//\\n/\n}" > /etc/mysql/conf.d/replication.cnf'; \
podman exec -e cnf="$$repl1_conf" replica1 bash -c 'echo -e "$${cnf//\\n/\n}" > /etc/mysql/conf.d/replication.cnf'; \
podman exec -e cnf="$$repl2_conf" replica2 bash -c 'echo -e "$${cnf//\\n/\n}" > /etc/mysql/conf.d/replication.cnf'
# Don't restart a container unless it is healthy
while ! podman healthcheck run primary && [[ "$$SECONDS" -lt 120 ]]; do sleep 1; done
podman restart -t 30 primary
while ! podman healthcheck run replica1 && [[ "$$SECONDS" -lt 120 ]]; do sleep 1; done
podman restart -t 30 replica1
while ! podman healthcheck run replica2 && [[ "$$SECONDS" -lt 120 ]]; do sleep 1; done
podman restart -t 30 replica2
while ! podman healthcheck run primary && [[ "$$SECONDS" -lt 120 ]]; do sleep 1; done
mkdir -p .venv/$(ansible)
python$(local_python_version) -m venv .venv/$(ansible)
# Start venv (use `; \` to keep the same shell)
source .venv/$(ansible)/bin/activate; \
python$(local_python_version) -m ensurepip; \
python$(local_python_version) -m pip install --disable-pip-version-check \
https://github.com/ansible/ansible/archive/$(ansible).tar.gz; \
set -x; \
ansible-test integration $(target) -v --color --coverage --diff \
--docker ubuntu2204 \
--docker-network podman $(_continue_on_errors) $(_keep_containers_alive); \
set +x
# End of venv
rm tests/integration/db_engine_name
rm tests/integration/db_engine_version
rm tests/integration/connector_name
rm tests/integration/connector_version
rm tests/integration/ansible
ifndef keep_containers_alive
podman stop --time 0 --ignore primary replica1 replica2
podman rm --ignore --volumes primary replica1 replica2
endif