"When we published Playground, the overwhelming need quickly exhausted the local lists for MMS nodes far quicker than the system could refresh them. Each node was operating to every other node to ask additional servers which just weren't there yet, or at least took a long time to buy Fortnite Items
pick out of the non-local lists.
The long compute times caused the CPU to end up with a backlog of pending orders, leading to a feedback loop which eventually resulted in the system to grind to a stop."
It took Epic longer than they expected to correct the problems, but eventually the manner returned on July 2. Epic states that the solution involved giving the Playground Mode its own service cluster and to then give that service cluster the capability to re-balance sessions from other nodes:"Once we identified the origin of the problem as the fatigue of sessions from local lists, the solution was to provide the cluster the capability to bulk rebalance sessions from different nodes to ensure repeated lookups were not necessary.
With the system constantly shifting regional capacity from nodes with too much nodes Which May be running low, the odds of some node running dry for fortnite weapons
a particular region and needing to seek out outside its Regional list have been radically reduced,