DB loading not all tokens, when there is larger set of tokens.

In the attached screenshot, all jobs should have 45k tokens, but none of them has