WRED can be used to help avoid TCP slow-start synch problems.

Question

Options

• ATRUE
• BFALSE

• A
  90% (52)
• B
  10% (6)

Explanation

A (TRUE) is correct. WRED (Weighted Random Early Detection) addresses TCP global synchronization by proactively and randomly dropping packets from different flows before the queue fills up, rather than waiting until the queue is completely full and dropping everything at once (tail-drop behavior). Because drops are staggered across flows using different minimum/maximum threshold configurations, individual TCP connections enter congestion-avoidance mode at different times instead of all simultaneously triggering slow-start together - which is exactly the sync problem. Option B is wrong because tail-drop (not WRED) is the mechanism that causes synchronization by forcing all flows to back off simultaneously when the queue overflows.

Memory tip: Think of the "R" in WRED as "Randomized Relief" - by randomly relieving pressure on the queue early and unevenly, no two flows hit the brakes at the same moment, keeping traffic flowing smoothly instead of in synchronized waves.

No community discussion yet for this question.