COLLATION 'utf8_bin' is not valid for CHARACTER SET 'utf8mb4'

Getting the title error on my linux machine. Works great with the windows test-server on the same database (cloned) and the same code.

I am using utf8mb4_general_ci as standard collation, but utf8mb4_bin at column “token”. I think the part (`lime`.`user`.`token` = ? COLLATE utf8_bin) should use utf8mb4_bin instead? Interestingly enough: It does not help when I convert the column token to utf8_bin!

Stacktrace

2017-06-08T18:10:42.336Z INFO  [http-nio-8080-exec-1] (#APPLICATION_BUILDER) - Speedment (Open Source) version 3.0.10 by Speedment, Inc. Specification version 3.0
2017-06-08T18:10:42.336Z INFO  [http-nio-8080-exec-1] (#APPLICATION_BUILDER) - Java Platform API Specification 1.8 by Oracle Corporation. Implementation N/A 1.8.0_131 by N/A
2017-06-08T18:10:42.655Z INFO  [http-nio-8080-exec-1] (#APPLICATION_BUILDER) - MySQL, 10.0.29-MariaDB-0ubuntu0.16.10.1, MariaDB connector/J 2.0.2, JDBC version 4.2
2017-06-08T18:10:42.852Z ERROR [http-nio-8080-exec-3] (c.s.r.c.i.d.AsynchronousQueryResultImpl) - Error executing SELECT `id`,`token`,`firebase`,`initial`,`timestamp_location`,`distance`,`longitude`,`latitude` FROM `lime`.`user` WHERE (`lime`.`user`.`token`  = ? COLLATE utf8_bin), values=[sometoken]
java.sql.SQLSyntaxErrorException: (conn:18774) COLLATION 'utf8_bin' is not valid for CHARACTER SET 'utf8mb4'
	at org.mariadb.jdbc.internal.util.exceptions.ExceptionMapper.get(ExceptionMapper.java:158)
	at org.mariadb.jdbc.internal.util.exceptions.ExceptionMapper.getException(ExceptionMapper.java:118)
	at org.mariadb.jdbc.MariaDbStatement.executeExceptionEpilogue(MariaDbStatement.java:245)
	at org.mariadb.jdbc.MariaDbPreparedStatementClient.executeInternal(MariaDbPreparedStatementClient.java:218)
	at org.mariadb.jdbc.MariaDbPreparedStatementClient.execute(MariaDbPreparedStatementClient.java:153)
	at org.mariadb.jdbc.MariaDbPreparedStatementClient.executeQuery(MariaDbPreparedStatementClient.java:167)
	at com.speedment.runtime.core.internal.db.AsynchronousQueryResultImpl.stream(AsynchronousQueryResultImpl.java:95)
	at com.speedment.runtime.core.internal.component.sql.SqlStreamSupplierImpl.lambda$stream$6(SqlStreamSupplierImpl.java:184)
	at com.speedment.runtime.core.internal.stream.builder.pipeline.PipelineImpl.getStream(PipelineImpl.java:85)
	at com.speedment.runtime.core.internal.stream.builder.pipeline.PipelineImpl.getAsReferenceStream(PipelineImpl.java:63)
	at com.speedment.runtime.core.internal.component.sql.override.def.reference.DefaultFindAnyTerminator.apply(DefaultFindAnyTerminator.java:45)
	at com.speedment.runtime.core.internal.manager.sql.SqlStreamTerminator.findAny(SqlStreamTerminator.java:176)
	at com.speedment.runtime.core.internal.stream.builder.ReferenceStreamBuilder.lambda$findAny$16(ReferenceStreamBuilder.java:406)
	at com.speedment.runtime.core.internal.stream.builder.AbstractStreamBuilder.finallyCloseReference(AbstractStreamBuilder.java:189)
	at com.speedment.runtime.core.internal.stream.builder.ReferenceStreamBuilder.findAny(ReferenceStreamBuilder.java:406)

Database

[fablue]

Fixed by adding the following to my.ini

character-set-server  = utf8mb4
collation-server      = utf8mb4_general_ci

[minborg]

Komisch, nicht wahr? It might be the case that there are different defaults for the table collation on the two platforms. Try to re-create your table using “utf8” . We do our integration tests on Linux and here is one example of a table created for MariaDB:

CREATE TABLE `user` (
  `id` int(10) unsigned NOT NULL AUTO_INCREMENT,
  `name` varchar(45) NOT NULL,
  `surname` varchar(45) NOT NULL,
  `email` varchar(45) NOT NULL,
  `password` varchar(45) NOT NULL,
  PRIMARY KEY (`id`),
  UNIQUE KEY `Index_email` (`email`),
  KEY `Index_name` (`name`)
) ENGINE=InnoDB AUTO_INCREMENT=101 DEFAULT CHARSET=utf8;