Reconciling Cluster every 5 sec

[nekufa]

We have simplest cartridge application that uses only built-in roles:

• vshard-router, vshard-storage
• crud-router, crud-storage
• migrator
• metrics

This application is deployed to kubernetes using standard tarantool-operator helm.
Everything looks okay, but every 5 seconds there are a lot of log messages about cluster reconcilling.

I collect all needed data at public gist page here:
https://gist.github.com/nekufa/bcf619ec14294211ad3d7ffe8eb4925b

Gist contains:

• cartridge application rockspec and init file (data-scm-1.rockspec, init.lua)
• helm values for deployment (queue-data.yaml)
• all service logs (tarantool-operator-*.log, queue-data-instance-*.log)