Version
main
Describe the problem you're encountering
When we enabled serialize_worker_startup=true on a cluster we noticed user could not update their _replicator docs. Replication would finish, replicator updated the completed state. User grabs the rev (say 2-y) and then tries to use the rev and it gets the 409. Retrying to get the revs doesn't make any difference: user still gets the same rev (2-y) and still gets a 409s when trying to use it. There is not 3-z rev to wait on, 2-y is the latest rev. Eventually when the internal replicator catches up, the user can finally make progress and update the doc (use 2-y to delete it)
Expected Behaviour
User can grab the latest rev then use the latest revs to update the docs in the db
Steps to Reproduce
No response
Your Environment
Debian Trixie, Intel, Erlang 27
Additional Context
No response
Version
main
Describe the problem you're encountering
When we enabled serialize_worker_startup=true on a cluster we noticed user could not update their _replicator docs. Replication would finish, replicator updated the
completedstate. User grabs the rev (say 2-y) and then tries to use the rev and it gets the 409. Retrying to get the revs doesn't make any difference: user still gets the same rev (2-y) and still gets a 409s when trying to use it. There is not 3-z rev to wait on, 2-y is the latest rev. Eventually when the internal replicator catches up, the user can finally make progress and update the doc (use 2-y to delete it)Expected Behaviour
User can grab the latest rev then use the latest revs to update the docs in the db
Steps to Reproduce
No response
Your Environment
Debian Trixie, Intel, Erlang 27
Additional Context
No response