Inserts are not properly ordered when populating number of cells

[arteymix]

I have hibernate.order_inserts=true configured, and I did check in the batch logging and we're getting this by default:

2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,740 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,741 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1
2025-12-15 11:53:47,742 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR (ID, BIO_ASSAY_DIMENSION_FK, DESIGN_ELEMENT_FK, DATA, QUANTITATION_TYPE_FK, EXPRESSION_EXPERIMENT_FK) values (null, ?, ?, ?, ?, ?)
2025-12-15 11:53:47,742 DEBUG 210056 [main] o.h.e.j.s.SqlStatementLogger.logStatement(104) | insert into RAW_EXPRESSION_DATA_VECTOR_NUMBER_OF_CELLS (NUMBER_OF_CELLS, ID) values (?, ?)
2025-12-15 11:53:47,742 DEBUG 210056 [main] o.h.e.j.b.i.BatchingBatch.doExecuteBatch(100) | Executing batch size: 1

I have a workaround for this that consists of flushing the vectors without the NOCs populated and then adding those back.