At the moment we store DNS state in the CWNP status. DNS state is required to quickly restore nftables rules in case of machine reboot. Here is a response about Status field limit, which is limited only by etcd entry size(by default 1.5MB). So the there shouldn't be any restrictions to store the state in the object's Status.
Still, there's question if it can be optimized in terms of user experience. Aforementioned response proposes to use ConfigMap for storing big statuses. Which(potentially) can lead to some performance improvements. Also, Status intention is to show the operational state. And in our case it's used for caching. Moving caching to ConfigMap looks like a more "clean" solution.
At the moment we store DNS state in the
CWNPstatus. DNS state is required to quickly restore nftables rules in case of machine reboot. Here is a response aboutStatusfield limit, which is limited only byetcdentry size(by default 1.5MB). So the there shouldn't be any restrictions to store the state in the object's Status.Still, there's question if it can be optimized in terms of user experience. Aforementioned response proposes to use
ConfigMapfor storing big statuses. Which(potentially) can lead to some performance improvements. Also,Statusintention is to show the operational state. And in our case it's used for caching. Moving caching toConfigMaplooks like a more "clean" solution.