Hi, As you can see on other topics, I'm facing some difficulties with MultiTenancy in JPA. I haven't diagnose yet why settings are not applied but I translated the JDO TestCase to JPA. These tests

Ok, thanks Andy! Will change the local type as suggested.

"Actually JDBC metadata is not a reliable source of information in all or almost all drivers, don't trust it." Says much, about the attitude in the "database community" to a standard that was intended

Using properties.setProperty(PropertyNames.PROPERTY_MAPPING_TENANT_PROVIDER,MyTenantProvider.class.getName()) raises an error WARN Error accessing property datanucleus.TenantProvider; should be an

Interesting outcome in the discussion with Evgenij, H2 accepts NVARCHAR but doesn't directly support it. You can use it in H2 DDL though, will just be converted to VARCHAR. Wondering if it would be

Thanks for your support, below works for us. SQLTypeInfo sqlType = new H2TypeInfo( "NVARCHAR", // String typeName (short)Types.NVARCHAR, // short dataType 2147483647, // int

A committer on the H2 project suggested that "Actually JDBC metadata is not a reliable source of information in all or almost all drivers, don't trust it.". [1] Maybe the person is right and we should

Unable to set the multitenant properties (on current source): props.put(PropertyNames.PROPERTY_MAPPING_TENANT_READ_IDS, tenant); props.put(PropertyNames.PROPERTY_MAPPING_TENANT_ID, tenant); final

Yes H2 does indeed claim support for NVARCHAR: https://h2database.com/html/datatypes.html?highlight=NVARCHAR&search=NVARCHAR#varchar_type But it fails to expose it via it's JDBC driver as supported

So does H2 actually claim to support "NVARCHAR" ? By that I mean in its response to DatabaseMetadata.getTypeInfo()? You can easily run DataNucleus SchemaTool in "dbinfo" mode, and get output like the

Hi, Define what you had before and what you have now. Was the String field stored in a CLOB column (defined in the generated DDL as "CLOB") ? and presumably it was previously read/written using

On Sat, Feb 27, 2021 at 01:22 AM, <passignat@...> wrote: multiTenancyProvider Case: https://github.com/datanucleus/datanucleus-core/issues/365 I think it's available in this fix DN Version:

When mapping a column with java type String and using jdbcType="NVARCHAR" in the @Column annotation we receive an error that the type is not supported by our H2 datastore. We looked at the H2Adapter

Hi, I'm using Dependency-Track with PostgreSQL, which is built on DN. Dependency-Track recently switched from DN 5.2.2 to DN 5.2.6, making data in String fields annotated as CLOB inaccessible when

User makes changes when an object is detached. The persistence provider is not involved in the process when the object is detached. The dirty flag is then set on ALL fields that have their setters

I'm trying to use datanucleus.TenantReadIds at the EntityManager level, setting this property entityManager.setProperty(PropertyNames.PROPERTY_MAPPING_TENANT_READ_IDS, tenant); As it doesn't seems

Hi, I'm testing datanucleus.TenantReadIds and receive a NullPointerException PersistenceNucleusContextImpl:getMultiTenancyReadIds line 1928 return new String[]

Please use this attachment. Failing tests: testUpdateNoChangeDetach testUpdateNoChangeCopy Workaround for simple situation: testUpdateNoChangeApply -- Stephane

Hi Andy, my issue seems related to the merge operation. I attached a junit testcase. thanks -- Stephane

An object is not "made dirty" when setting a field to the same value ... except if the value is a floating point / double (aka imprecise) value. NucleusJPAHelper.getObjectState(obj) defines the state