You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The current implementation keeps pool entry in the store even if the pool has no active IP allocations. This is needed to preserve information about the latest allocated IP address. We need this info to support round-robin IP address allocation for use-case when the pool has few users on the node. The downside of the current implementation is that we may have a lot of stale pool entries in the store.
The proposed solution is to remove stale pool entries from the store within the cleaner loop (which cleans stale IP address allocations). The logic is quite simple: we need to remove the pool data from the store if the pool has no allocations and if information about the pool is unavailable in the Kubernetes API.
The text was updated successfully, but these errors were encountered:
The current implementation keeps pool entry in the store even if the pool has no active IP allocations. This is needed to preserve information about the latest allocated IP address. We need this info to support round-robin IP address allocation for use-case when the pool has few users on the node. The downside of the current implementation is that we may have a lot of stale pool entries in the store.
The proposed solution is to remove stale pool entries from the store within the cleaner loop (which cleans stale IP address allocations). The logic is quite simple: we need to remove the pool data from the store if the pool has no allocations and if information about the pool is unavailable in the Kubernetes API.
The text was updated successfully, but these errors were encountered: