Hangfire.Pro multiple continuations - batches can be stuck in the Started state

I have some strange issue; after some uptime hangfire do not picking up the jobs.

Server have the X workers but the amount of processing jobs is 0 or lower than X.

In the same time I see Batch is pending because it's dependent on other Job that is in Enqueued status, when I requeue this job then it unstuck it.

I also see that in the Queues view, for example, some queue have Fetched Jobs = 3, but when I inspect it all the 3 jobs are Succeeded like 4 days ago.

I can't pinpoint this to a single change on our side, we are using Hangfire 1.7.30 and 1.8 newest RC in various combinations (SQL Server or Memory Storage) and the issues are the same.

We have Startup subscription.

No files, tests, or code entry points are named. Start by reproducing the stalled batch and queue state with Hangfire 1.7.30 or 1.8 RC using SQL Server or Memory Storage; done means continuations leave Started or pending states and fetched-job counts no longer remain stale.

Written by the indexing model from the issue text.