STP port states still confusing me after 3 weeks – RSTP vs classic won't click
I've been studying for my networking cert for about 5 weeks now and Spanning Tree Protocol is the one topic I keep fumbling on practice exams. I understand the basic concept – loop prevention, root bridge election – but the moment they ask about specific port states and how RSTP handles convergence differently, I blank out. I'm getting around 62% on STP questions when I need 75%+ to feel confident going into the real exam.
The thing tripping me up most is the blocking vs discarding state distinction between classic STP and RSTP, and I can't lock in why RSTP converges so much faster. I've watched probably four different YouTube videos and read the chapter twice but it's not clicking the way other topics have.
I'm also confused about PortFast and BPDU Guard – when exactly do you use them and what's the risk if you accidentally enable PortFast on a trunk port? Any analogies or memory tricks would genuinely help.
I was at 60% on STP questions for two weeks straight. What broke it open for me was drawing out a topology and manually going through root bridge election step by step – lowest bridge ID wins, then root ports, then designated ports. Do it 10 times with different topologies and it becomes automatic.
The exam loves to test edge cases like equal-cost paths where tiebreakers kick in. Make sure you know the tiebreaker order: root path cost, then sender bridge ID, then sender port ID. That sequence shows up on almost every STP question set I've seen.
PortFast on a trunk port is a disaster waiting to happen – it skips listening and learning states, so if that port somehow creates a loop, STP won't catch it in time. Only use PortFast on access ports connected to end hosts. BPDU Guard is what you pair with it so that if a switch gets plugged in, the port shuts down immediately.
Think of classic STP port states as a slow approval process – blocking, listening, learning, forwarding, each step waits on a timer. RSTP tosses most of those timers and uses a proposal/agreement handshake instead, so convergence drops from 30–50 seconds to under a second. That mental model finally made it click for me.
Discarding in RSTP is basically blocking and listening combined into one state, which is why you only see discarding/learning/forwarding in RSTP rather than the five classic states.