Forum Discussion

1 Reply

  • Arie's avatar
    Arie
    Icon for Altostratus rankAltostratus

    Queuing works in conjunction with the Ram Cache. Normally all parallel (i.e. simultaneous) requests for the same object would be sent to the origin server (pool) if the object does not exist in Ram Cache (or is expired), which negates the benefits of using Ram Cache.

     

    Queuing detects multiple concurrent requests for the same object and causes the system to hold those requests (if the object is not cached) while it retrieves the object from the origin node. When that object is in cache the BIG-IP will then respond to the queued requests by sending the object from cache.