Open nivemaham opened 6 years ago
As far as i can tell, only one sequence generator is used. However it is used by all persistent entities which explains the gaps in ids, however the jump from 1000's range to 10000's range seems quite odd indeed. What are the performance tests actually doing? Perhaps they generate a lot of audit events? These are persistent entities as well and as a result also increment the sequence generator.
I don't think the ids are shared if the same sequence-generator is used by many persistent entities.
During performance testing i noticed the subject.id are not incremented properly with a uniform sequencer This will create failures in future when lower sequence reaches the higher. This should be investigated and fixed.