A systems administrator manages a pool of application servers behind a load balancer. Each time a server process starts, it takes several minutes to warm up, consuming CPU resources and delaying user requests if a cold server is selected. The administrator wants the load balancer to favor servers that already have active connections so that new-or "cold"-server processes are started only when necessary. Which load-balancing algorithm should be configured to meet this requirement?
The Most Recently Used (MRU) algorithm satisfies the requirement. MRU always selects the backend that most recently handled a connection, reusing an already-running server process before directing traffic to an idle one. This minimizes warm-up delays and reduces the overhead of starting new server instances.
Round robin cycles through the server list without regard to whether a server is warm or cold, so users could be sent to a process that is still starting.
Active-active describes a clustering role in which all nodes are online; it is not itself a request-distribution algorithm.
Link aggregation combines multiple NICs for bandwidth or redundancy but does not choose among application servers.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
What is the Most Recently Used (MRU) algorithm in load balancing?
Open an interactive chat with Bash
Why is round robin not suitable for managing cold server processes?
Open an interactive chat with Bash
What is the difference between MRU and active-active configurations?