Recent changes in handling Elasticsearch exceptions have led to 500s instead of 289s

/schema-worker workflow

Changes in this MR led to the fact that the AppException from mappingService.createMapping workflow are no longer intercepted as before and therefore does not turn the ElasticSearch exception with status 400 into AppException with status 289(SCHEMA_CONFLICT). This in turn leads to 500s and impact availability.

MR that resoves issue