Override application properties in Spring Boot integration tests - spring-boot

I'm preparing a Spring Boot starter (used in tests) and I want to override a specific application property. This particular case regards enabling / disabling cache based on a property (production-code starter uses #ConditionalOnProperty). When using the test starter I want to have the caching disabled by default.
Is there a way to do that except using #TestPropertySource? Hence it is not a repeatable annotation, I don't want to use it. It's a recommended way for end-user to add properties required for the test case.
Edit:
I'm providing more details on the specific use case
Production Starter auto-configuration:
#Configuration
#ConditionalOnWebApplication
#EnableConfigurationProperties(StarterCacheProperties.class)
#EnableCaching
public class StarterCacheAutoConfiguration {
..(omitted for clarity)..
#Configuration
#ConditionalOnProperty(prefix = "neostarter.saasmgr.cache", name = "enabled", havingValue = "false")
public static class SaasMgrNoCacheConfig extends CachingConfigurerSupport {
#Bean(name = SAAS_MGR_CACHE_MANAGER)
#Override
public CacheManager cacheManager() {
return new NoOpCacheManager();
}
}
#Configuration
#ConditionalOnProperty(prefix = "neostarter.saasmgr.cache", name = "enabled", matchIfMissing = true)
public static class SaasMgrCachingConfig extends CachingConfigurerSupport {
#Autowired
SaasMgrCacheProperties saasMgrCacheProperties;
#Bean(destroyMethod = "shutdown")
public net.sf.ehcache.CacheManager ehCacheManager() {
CacheConfiguration cacheConfiguration = new CacheConfiguration(SAAS_MGR_AUTH_CACHE, 1000)
.timeToLiveSeconds(saasMgrCacheProperties.getTimeToLiveSeconds());
net.sf.ehcache.config.Configuration config = new net.sf.ehcache.config.Configuration();
config.addCache(cacheConfiguration);
return net.sf.ehcache.CacheManager.newInstance(config);
}
#Bean(name = SAAS_MGR_CACHE_MANAGER)
#Override
public CacheManager cacheManager() {
return new EhCacheCacheManager(ehCacheManager());
}
}
This gives user a choice of caching particular calls in production (or not). However, I find it reasonable to disable caching in tests by default. I am fully aware that this can still be overridden by end-user. In a test starter (different dependency) I tried the following:
#Configuration
#AutoConfigureBefore({CacheAutoConfiguration.class, SaasMgrSecurityAutoConfiguration.class})
public class DisableCacheAutoConfiguration {
#Autowired
Environment environment;
#PostConstruct
public void disableCache() {
EnvironmentTestUtils.addEnvironment((ConfigurableEnvironment)environment, "neostarter.saasmgr.cache.enabled=false");
}
}
but no matter what StarterCacheAutoConfiguration with EhCacheCacheManager is always resolved before DisableCacheAutoConfiguration. In debug in AutoConfigurationSorter.java I see proper order of the configuration:
15 = "com.neoteric.starter.auth.saasmgr.test.DisableCacheAutoConfiguration"
16 = "com.neoteric.starter.auth.SaasMgrSecurityAutoConfiguration"
36 = "org.springframework.boot.autoconfigure.cache.CacheAutoConfiguration"

You can't change user's configuration in a starter, that would be way too opaque and invasive. You have no guarantee about it anyway: anybody could set the property value at a higher level (e.g. system property) and override what you're trying to change.
If you want to disable caching, the most easiest way to do it is to do what spring.cache.type=none does, that is providing an implementation of CacheManager that does not do anything.
Try to declare a bean of type org.springframework.cache.support.NoOpCacheManager

Related

where does spring boot set proxy as CGLib [duplicate]

Recently i found spring documentation page that says:
Spring AOP uses either JDK dynamic proxies or CGLIB to create the proxy for a given target object. (JDK dynamic proxies are preferred whenever you have a choice).
If the target object to be proxied implements at least one interface then a JDK dynamic proxy will be used.
But it doesn't seem to be the case in my application. I wanted to write a small benchmark to compare the performance of both types of proxying.
There are two similar classes. Both have one method annotated with the #Transactional annotation. One class implements the interface and the other does not:
#Service
public class Cglib {
#Transactional
public void method() {}
}
public interface Dynamic {
void method();
}
#Service
public class DynamicImpl implements Dynamic {
#Override
#Transactional
public void method() {}
}
And based on the documentation for the first class, a CGLIB proxy should be created, and for the second, a JDK dynamic proxy.
But in my case CGLIB was used for both classes:
#Component
#RequiredArgsConstructor
public class Runner implements ApplicationRunner {
private final Cglib cglib;
private final Dynamic dynamic;
#Override
public void run(ApplicationArguments args) {
System.out.println(cglib.getClass());
System.out.println(dynamic.getClass());
}
}
class com.example.demo.proxy.cglib.Cglib$$EnhancerBySpringCGLIB$$767ff22
class com.example.demo.proxy.dynamic.DynamicImpl$$EnhancerBySpringCGLIB$$20a564d6
There are no additional configurations in the application. Only #SpringBootApplication class generated via spring initializr
Am I doing something wrong? The code was run on Spring Boot 2.7.2 and JDK 17.
That is due to spring-boot autoconfiguation:
#Configuration(proxyBeanMethods = false)
#ConditionalOnBean(TransactionManager.class)
#ConditionalOnMissingBean(AbstractTransactionManagementConfiguration.class)
public static class EnableTransactionManagementConfiguration {
#Configuration(proxyBeanMethods = false)
#EnableTransactionManagement(proxyTargetClass = false)
#ConditionalOnProperty(prefix = "spring.aop", name = "proxy-target-class", havingValue = "false")
public static class JdkDynamicAutoProxyConfiguration {
}
#Configuration(proxyBeanMethods = false)
#EnableTransactionManagement(proxyTargetClass = true)
#ConditionalOnProperty(prefix = "spring.aop", name = "proxy-target-class", havingValue = "true",
matchIfMissing = true)
public static class CglibAutoProxyConfiguration {
}
}
it turns #EnableTransactionManagement(proxyTargetClass = true) (Indicate whether subclass-based (CGLIB) proxies are to be created (true) as opposed to standard Java interface-based proxies (false)) when the property spring.aop.proxy-target-class is not set (matchIfMissing = true)

How inject DataSource(HikariCP) in Auto Configuration class?

I'm trying to use Spring Boot 2 + Spring Data + Custom Auto Configuration Classes but for some reason can't inject DataSource(provided by HikariCP) in the third class.
#Configuration
#AutoConfigureOrder(Ordered.LOWEST_PRECEDENCE)
public class InitialAutoConfiguration {
//Beans to load in theory first.
}
#Configuration
#AutoConfigureOrder(Ordered.LOWEST_PRECEDENCE)
#AutoConfigureAfter(InitialAutoConfiguration.class)
#EntityScan(basePackageClasses = Asset.class)
#EnableJpaRepositories(basePackageClasses = AssetRepository.class,
repositoryBaseClass = BaseRepositoryImpl.class,
repositoryFactoryBeanClass = ExtendedJpaRepositoryFactoryBean.class)
public class JpaAutoConfiguration { //Load Jpa Classes
}
#Configuration
#AutoConfigureOrder(Ordered.HIGHEST_PRECEDENCE)
#AutoConfigureAfter(JpaAutoConfiguration.class)
#ComponentScan(basePackages = {"com.xxx"})
public class ServiceConfiguration {
#Inject
private DataSource dataSource; //Datasource is null
#Bean
public DbPropertySourcesPlaceholderConfigurer dbPropertySourcesPlaceholderConfigurer() {
DbPropertySourcesPlaceholderConfigurer placeholderConfigurer = new DbPropertySourcesPlaceholderConfigurer(dataSource);
placeholderConfigurer.setPlaceholderPrefix("%{");
placeholderConfigurer.setPlaceholderSuffix("}");
return placeholderConfigurer;
}
}
META-INF/spring.factories
org.springframework.boot.autoconfigure.EnableAutoConfiguration=com.xetec.autoconfigure.InitialAutoConfiguration,\
com.xetec.autoconfigure.JpaAutoConfiguration,\
com.xetec.autoconfigure.ServiceConfiguration
Looks like my classes are loading before the Spring Data Classes(DataSourceAutoConfiguration.Hikari).
Is there one way to first load the Spring Boot Starters Classes then after my custom ones please?
Thanks
I am no sure if the following change can solve your issue, but i meet with the similar problem
add static in your method
#Bean
public static DbPropertySourcesPlaceholderConfigurer dbPropertySourcesPlaceholderConfigurer() {
....
}
Your use of highest and lowest precedence is the wrong way round. Your ServiceConfiguration is ordered with highest precedence which means that it will be evaluated first.
Rather than using absolute ordering, I would use #AutoConfigureAfter(DataSourceAutoConfiguration.class)

How to configure springboot to wrap DataSource during integration tests?

My goal is to have a have integration tests that ensures that there isn't too many database queries happening during lookups. (This helps us catch n+1 queries due to incorrect JPA configuration)
I know that the database connection is correct because there is no configuration problems during the test run whenever MyDataSourceWrapperConfiguration is not included in the test. However, once it is added, the circular dependency happens. (see error below) I believe #Primary is necessary in order for the JPA/JDBC code to use the correct DataSource instance.
MyDataSourceWrapper is a custom class that tracks the number of queries that have happened for a given transaction, but it delegates the real database work to the DataSource passed in via constructor.
Error:
The dependencies of some of the beans in the application context form a cycle:
org.springframework.boot.autoconfigure.orm.jpa.HibernateJpaAutoConfiguration
┌─────┐
| databaseQueryCounterProxyDataSource defined in me.testsupport.database.MyDataSourceWrapperConfiguration
↑ ↓
| dataSource defined in org.springframework.boot.autoconfigure.jdbc.DataSourceConfiguration$Tomcat
↑ ↓
| dataSourceInitializer
└─────┘
My Configuration:
#Configuration
public class MyDataSourceWrapperConfiguration {
#Primary
#Bean
DataSource databaseQueryCounterProxyDataSource(final DataSource delegate) {
return MyDataSourceWrapper(delegate);
}
}
My Test:
#ActiveProfiles({ "it" })
#RunWith(SpringJUnit4ClassRunner.class)
#SpringApplicationConfiguration({ DatabaseConnectionConfiguration.class, DatabaseQueryCounterConfiguration.class })
#EnableAutoConfiguration
public class EngApplicationRepositoryIT {
#Rule
public MyDatabaseQueryCounter databaseQueryCounter = new MyDatabaseQueryCounter ();
#Rule
public ErrorCollector errorCollector = new ErrorCollector();
#Autowired
MyRepository repository;
#Test
public void test() {
this.repository.loadData();
this.errorCollector.checkThat(this.databaseQueryCounter.getSelectCounts(), is(lessThan(10)));
}
}
UPDATE: This original question was for springboot 1.5. The accepted answer reflects that, however, the answer from #rajadilipkolli works for springboot 2.x
In your case you will get 2 DataSource instances which is probably not what you want. Instead use BeanPostProcessor which is the component actually designed for this. See also the Spring Reference Guide.
Create and register a BeanPostProcessor which does the wrapping.
public class DataSourceWrapper implements BeanPostProcessor {
public Object postProcessBeforeInitialization(Object bean, String beanName) {
if (bean instanceof DataSource) {
return new MyDataSourceWrapper((DataSource)bean);
}
return bean;
}
public Object postProcessAfterInitialization(Object bean, String beanName) throws BeansException {
return bean;
}
}
Then just register that as a #Bean instead of your MyDataSourceWrapper.
Tip: Instead of rolling your own wrapping DataSource you might be interested in datasource-proxy combined with datasource-assert which has counter etc. support already (saves you maintaining your own components).
Starting from spring boot 2.0.0.M3 using BeanPostProcessor wont work.
As a work around create your own bean like below
#Bean
public DataSource customDataSource(DataSourceProperties properties) {
log.info("Inside Proxy Creation");
final HikariDataSource dataSource = (HikariDataSource) properties
.initializeDataSourceBuilder().type(HikariDataSource.class).build();
if (properties.getName() != null) {
dataSource.setPoolName(properties.getName());
}
return ProxyDataSourceBuilder.create(dataSource).countQuery().name("MyDS")
.logSlowQueryToSysOut(1, TimeUnit.MINUTES).build();
}
Another way is to use datasource-proxy version of datasource-decorator starter
Following solution works for me using Spring Boot 2.0.6.
It uses explicit binding instead of annotation #ConfigurationProperties(prefix = "spring.datasource.hikari").
#Configuration
public class DataSourceConfig {
private final Environment env;
#Autowired
public DataSourceConfig(Environment env) {
this.env = env;
}
#Primary
#Bean
public MyDataSourceWrapper primaryDataSource(DataSourceProperties properties) {
DataSource dataSource = properties.initializeDataSourceBuilder().build();
Binder binder = Binder.get(env);
binder.bind("spring.datasource.hikari", Bindable.ofInstance(dataSource).withExistingValue(dataSource));
return new MyDataSourceWrapper(dataSource);
}
}
You can actually still use BeanPostProcessor in Spring Boot 2, but it needs to return the correct type (the actual type of the declared Bean). To do this you need to create a proxy of the correct type which redirects DataSource methods to your interceptor and all the other methods to the original bean.
For example code see the Spring Boot issue and discussion at https://github.com/spring-projects/spring-boot/issues/12592.

Best way to test JPA Spring 3.1

I have written several JPA Repositories, Services and support classes for my Spring 3.1.1/JPA/Hibernate 4 web app. However, I really want to write some unit and integration tests for it (I know, you are supposed to write those first). I am using JavaConfig rather than XML, so I am wondering the best way to test. Here is the particular problem I am trying to solve:
I have a #Configuration that declares DataSource, JpaTransactionManager, LocalContainerEntityManagerFactoryBean, and all my respositories. Obviously I don't want to start all that up for an integration test, so I thought I could use the EmbeddedDatabase and H2 to create an in memory database, populate with values, and then use my Repositories against it. However, the documentation I have seen hasn't helped me put this together. I have this:
#RunWith( SpringJUnit4ClassRunner.class )
public class TestMenuService {
private EmbeddedDatabase database;
#Autowired
private MenuRepository menuRepository;
#Before
public void setUp() throws Exception {
database = new EmbeddedDatabaseBuilder().setType(EmbeddedDatabaseType.H2).setName("myschema")
.addScript("classpath:schema.sql").build();
menuRepository = new MenuRepository();
Assert.assertNotNull(database);
}
But the menuRepository does not get instantiated, so I tried creating a test version of my #Configuration
#Configuration } )
#ComponentScan( basePackages = { "com.mycompany.service"} )
#EnableTransactionManagement
#ImportResource( "classpath:applicationContext.xml" )
#PropertySource( "classpath:test-application.properties" )
public class TestEdmConfiguration {
#Bean
MenuRepository menuRepository() {
return new MenuRepository();
}
My test-applicationContext.xml
<jpa:repositories base-package="com.mycompany.servce.repository"/>
My test-application.properties:
db.driver=org.h2.Driver
db.username=sa
db.password=
db.url=jdbc:h2:mem:myschema
hibernate.dialect=org.hibernate.dialect.H2Dialect
hibernate.format_sql=true
hibernate.hbm2ddl.auto=create
hibernate.ejb.naming_strategy=org.hibernate.cfg.ImprovedNamingStrategy
hibernate.show_sql=true
But this requires that I create all the datasources, etc mentioned above. It seems like I am just duplicating all the support beans for this one.
Is there a way to have the reposository and embeddeddatabase isolated to my test without all the other dependencies?
If you want to test your repository in a full integration test I would image that you need everything else to be setup i.e. EntityManagerFactory, PlatformTransactionManager etc.
Since you are using Spring 3.1 I would suggest that you achieve this using bean profiles.
I would create two profiles one for tests and one for the application, each of which supplies a datasource.
#Configuration
#Profile("test")
public class EmbeddedDataSource {
#Bean
public DataSource dataSource() {
// Return a H2 datasource
}
}
#Configuration
#Profile("application")
public class ApplicationDataSource {
#Bean
public DataSource dataSource() {
// Return a normal datasource
}
}
The you can create a test which starts up the spring context as follows:
#RunWith(SpringJUnit4ClassRunner.class)
#ContextConfiguration(loader = AnnotationConfigContextLoader.class, classes = { MyConfigClass.class })
#ActiveProfiles(profiles = {"test"})
#TransactionConfiguration(transactionManager = "transactionManager", defaultRollback = true)
public class TestRepository {
}
In here you can specify the profiles where are active for the test.

Can #PropertySources be chosen by Spring profile?

I have a Spring 3.1 #Configuration that needs a property foo to build a bean. The property is defined in defaults.properties but may be overridden by the property in overrides.properties if the application has an active override Spring profile.
Without the override, the code would look like this, and work...
#Configuration
#PropertySource("classpath:defaults.properties")
public class MyConfiguration {
#Autowired
private Environment environment;
#Bean
public Bean bean() {
...
// this.environment.getRequiredProperty("foo");
...
}
}
I would like a #PropertySource for classpath:overrides.properties contingent on #Profile("overrides"). Does anyone have any ideas on how this could be achieved? Some options I've considered are a duplicate #Configuration, but that would violate DRY, or programmatic manipulation of the ConfigurableEnvironment, but I'm not sure where the environment.getPropertySources.addFirst() call would go.
Placing the following in an XML configuration works if I inject the property directly with #Value, but not when I use Environment and the getRequiredProperty() method.
<context:property-placeholder ignore-unresolvable="true" location="classpath:defaults.properties"/>
<beans profile="overrides">
<context:property-placeholder ignore-unresolvable="true" order="0"
location="classpath:overrides.properties"/>
</beans>
Update
If you're trying to do this now, check out Spring Boot's YAML support, particularly the 'Using YAML instead of Properties' section. The profile support there would make this question moot, but there isn't #PropertySource support yet.
Add the overriding #PropertySource in a static inner class. Unfortunately, you must specify all property sources together which means creating a "default" profile as the alternative to "override".
#Configuration
public class MyConfiguration
{
#Configuration
#Profile("default")
#PropertySource("classpath:defaults.properties")
static class Defaults
{ }
#Configuration
#Profile("override")
#PropertySource({"classpath:defaults.properties", "classpath:overrides.properties"})
static class Overrides
{
// nothing needed here if you are only overriding property values
}
#Autowired
private Environment environment;
#Bean
public Bean bean() {
...
// this.environment.getRequiredProperty("foo");
...
}
}
I suggest, defining two files, where the second is optional with the profile as suffix:
#Configuration
#PropertySources({
#PropertySource("classpath:/myconfig.properties"),
#PropertySource(value = "classpath:/myconfig-${spring.profiles.active}.properties", ignoreResourceNotFound = true)
})
public class MyConfigurationFile {
#Value("${my.prop1}")
private String prop1;
#Value("${my.prop2}")
private String prop2;
}
You can do:
<context:property-placeholder location="classpath:${spring.profiles.active}.properties" />
Edit: if you need something more advanced, you can register your PropertySources on application startup.
web.xml
<context-param>
<param-name>contextInitializerClasses</param-name>
<param-value>com.xxx.core.spring.properties.PropertySourcesApplicationContextInitializer</param-value>
</context-param>
file you create:
public class PropertySourcesApplicationContextInitializer implements ApplicationContextInitializer<ConfigurableApplicationContext> {
private static final Logger LOGGER = LoggerFactory.getLogger(PropertySourcesApplicationContextInitializer.class);
#Override
public void initialize(ConfigurableApplicationContext applicationContext) {
LOGGER.info("Adding some additional property sources");
String[] profiles = applicationContext.getEnvironment().getActiveProfiles()
// ... Add property sources according to selected spring profile
// (note there already are some property sources registered, system properties etc)
applicationContext.getEnvironment().getPropertySources().addLast(myPropertySource);
}
}
Once you've done it you just need to add in your context:
<context:property-placeholder/>
I can't really answer to your question about multiple profiles but I guess you activate them on such an initializer, and you could register the appropriate PropertySource items during profile activations.
I can't think of any other way than one you have suggested Emerson, which is to define this bean in a separate #Configuration file with an #Profile annotation:
#Configuration
#Profile("override")
#PropertySource("classpath:override.properties")
public class OverriddenConfig {
#Autowired
private Environment environment;
#Bean
public Bean bean() {
//if..
}
}
In case you need to support multiple profiles you could do something like this:
#Configuration
public class Config {
#Configuration
#Profile("default")
#PropertySource("classpath:application.properties")
static class DefaultProperties {
}
#Configuration
#Profile("!default")
#PropertySource({"classpath:application.properties", "classpath:application-${spring.profiles.active}.properties"})
static class NonDefaultProperties {
}
}
That way you don't need to define a static configuration class for each profile.
Thanks David Harkness for putting me into the right direction.
Note: This answer provides an alternate solution to using properties files with #PropertySource. I went this route because it was too cumbersome trying to work with multiple properties files that may each have overrides while avoiding repetitive code.
Create a POJO interface for each related set of properties to define their names and types.
public interface DataSourceProperties
{
String driverClassName();
String url();
String user();
String password();
}
Implement to return the default values.
public class DefaultDataSourceProperties implements DataSourceProperties
{
public String driverClassName() { return "com.mysql.jdbc.Driver"; }
...
}
Subclass for each profile (e.g. development, production) and override any values that differ from the default. This requires a set of mutually-exclusive profiles, but you can easily add "default" as the alternative to "overrides".
#Profile("production")
#Configuration
public class ProductionDataSourceProperties extends DefaultDataSourceProperties
{
// nothing to override as defaults are for production
}
#Profile("development")
#Configuration
public class DevelopmentDataSourceProperties extends DefaultDataSourceProperties
{
public String user() { return "dev"; }
public String password() { return "dev"; }
}
Finally, autowire the properties configurations into the other configurations that need them. The advantage here is that you don't repeat any #Bean creation code.
#Configuration
public class DataSourceConfig
{
#Autowired
private DataSourceProperties properties;
#Bean
public DataSource dataSource() {
BoneCPDataSource source = new BoneCPDataSource();
source.setJdbcUrl(properties.url());
...
return source;
}
}
I am still not convinced I'll stick with this over manually configuring properties files based on the active profiles in a servlet context initializer. My thought was that doing manual configuration would not be as amenable to unit testing, but I'm not so sure now. I really prefer reading properties files to a list of property accessors.
All mentioned here solutions are a bit awkward, work only with one profile preset, and they won't work with more/other profiles. Currently a Spring team refuses to introduce this feature. But here's the working workaround I've found:
package com.example;
public class MyPropertySourceFactory implements PropertySourceFactory, SpringApplicationRunListener {
public static final Logger logger = LoggerFactory.getLogger(MyPropertySourceFactory.class);
#NonNull private static String[] activeProfiles = new String[0];
// this constructor is used for PropertySourceFactory
public MyPropertySourceFactory() {
}
// this constructor is used for SpringApplicationRunListener
public MyPropertySourceFactory(SpringApplication app, String[] params) {
}
#Override
public void environmentPrepared(ConfigurableBootstrapContext bootstrapContext, ConfigurableEnvironment environment) {
activeProfiles = environment.getActiveProfiles();
}
#Override
public PropertySource<?> createPropertySource(String name, EncodedResource encodedResource) throws IOException {
logger.info("Loading: {} with profiles: {}", encodedResource.toString(), activeProfiles);
// here you know all profiles and have the source Resource with main
// properties, just try to load other resoures in the same path with different
// profile names and return them as a CompositePropertySource
}
}
To make it working you have to have src/main/resources/META-INF/spring.factories with the following content:
org.springframework.boot.SpringApplicationRunListener=com.example.MyPropertySourceFactory
Now you can put your custom properties file somewhere and load it with #PropertySources:
#Configuration
#PropertySource(value = "classpath:lib.yml", factory = MyPropertySourceFactory.class)
public class PropertyLoader {
}

Resources